Dynamics of Biomass and Carbon Stocks during Reforestation on Abandoned Agricultural Lands in Southern Ural Region

Abstract

Due to the global increase in CO₂ in the atmosphere, studies focusing on the carbon balance in forest ecosystems are currently particularly relevant. Abandoned agricultural lands could provide an important contribution to carbon sequestration in many parts of the world. In the broad-leaved forest zone of the Cis-Ural (Southern Ural region, Russia), the carbon sequestration dynamics in the biomass of woody and herbaceous plants, as well as in the litter and soil on abandoned arable lands repopulated with silver birch (Betula pendula), was studied. The data were collected on 35 round (with diameter of 30 m) sample plots located within communities representing the different stages of reforestation with tree stands aged 3 to 30 years. It was found that the carbon content of the stem wood and herbaceous understory did not depend on the succession stages, which largely corresponds to the literature data. The carbon content in root biomass and soil organic matter increased along with the growth of tree stands. While the forest stand grew, the carbon content in the grey forest soil increased from 2.5 to 4.4%, and in the more fertile dark grey forest soil it changed only slightly. The carbon deposition by the forest stands on the sample plots located on the dark grey forest soils was higher than on grey forest soils. The average rate of carbon sequestration in the tree stand was 2.7 t/ha/year. Most mature, 25–30-years-old silver birch tree stands provided the highest average annual increase in tree biomass and the rate of carbon sequestration evaluated was 9 t/ha/year. Also, the carbon pool in the 30 cm soil layer was 2.7 times greater than in the tree stand. It was concluded that abandoned agricultural lands overgrowing by forest in the Cis-Ural are promising for carbon sequestration.

1. Introduction

Modern agricultural production worldwide has been shaped by the expansion of arable lands and the intensification of crop production on existing tilled lands [1]. However, the change in environmental, socioeconomic, and political conditions has caused the abandonment of much of the croplands [2,3]. The withdrawal of large agricultural areas has led to changes in the structure and hydrology regime of the landscape and changes in biodiversity, fire frequency, etc., as well as an increase in the environmental and economic costs of their reclamation [3,4,5,6,7,8,9]. According to various sources, there is from 1.7 [10] to 2.23 [11] million km² of abandoned agricultural land in the world. Impressive examples of this transformation of abandoned agricultural lands can also be seen after the breakup of the Soviet Union and the transition to a market system of management, which, in turn, led to a significant reduction in agricultural production [12,13]. According to the official data, in Russia, in 1990–2010, the total area of abandoned agricultural lands (croplands, hayfields, pastures) amounted to 56.8 million ha [14]. Moreover, according to the 2016 agricultural census, the total area of unutilized agricultural lands was already 97.2 million ha, or 44% of all agricultural lands in the country [15]. Much of these lands were overgrown with woody and shrub vegetation [16,17].

The Republic of Bashkortostan (RB) is among the regions in Russia with the highest percentage of abandoned croplands overgrown with forest vegetation, which cover over 4 million ha [18]. According to data of the global forest cover monitoring from Landsat satellite remote sensing, fully fledged forest plantations were formed on many plots of fallow lands in 25–30 years [19]. They are particularly widespread in the zone of broad-leaved forests [20,21,22]. Most of these abandoned lands cannot be used for agriculture in the near future, but they are quite suitable for forest farming, plantation forestry, protective afforestation, agroforestry, etc. [17]. One of the positive aspects of reforestation on fallow lands is carbon sequestration in phytomass and soil, especially at the stage of tree stand formation [23,24,25].

Fast-growing tree species can facilitate both high biomass production and an increase in the soil organic carbon pool within a relatively short period of time, especially in areas of natural reforestation [26,27,28,29,30,31,32,33,34,35]. Factors affecting the rate of carbon sequestration include composition and age of tree stands, soil fertility, and forest management [36,37]. Even the understory layer contribution to carbon sequestration may be significant [38].

The silver birch (Betula pendula Roth) grows in a wide range across Eurasia from the Atlantic to the Far East and takes a special place among the fast-growing trees that occupy the fallow lands [39,40]. In the Baltic States and Northern Europe, birch is the most important commercial wood species and contributes from 11 to 28% to the total volume of growing stock [41,42]. In the RB, birch contributes about 20% to the total volume of growing stock and actively expands to the unutilized arable lands. In this connection, it is important to study the features of carbon sequestration by vegetation and soil in areas with plant communities representing different successional stages of birch regrowth on the abandoned croplands.

In the Southern Ural region, the carbon sequestration potential of silver birch and the lands covered/overgrown by it has been poorly studied. Prior to this study, only the stages of overgrowth of abandoned agricultural lands, as well as general patterns of changes in floristic composition of successional plant communities, were briefly described in the region [43]. In order to objectively evaluate the carbon stock of overgrown abandoned agricultural lands, comprehensive studies of the carbon sequestration potential of trees, understory vegetation, litter, and soil at different stages of stand formation are required. In the Cis-Ural, these studies were initiated in the framework of the state program for the creation and functioning of carbon plots in the RB, as part of the Eurasian carbon polygon. The aim of this study is to analyze the carbon sequestration dynamics in the biomass of woody and herbaceous plants, as well as in the litter and soil in abandoned croplands overgrown with birch in the broad-leaved forest zone of the Cis-Ural by undertaking a case study in Mishkinsky District of RB.

2. Materials and Methods

2.1. Site Description

The study area sized 10 × 10 km is located in the broad-leaved forest zone of the Cis-Ural (Figure 1). The climate of the study area is moderately continental. The average annual air temperature is +3.5 °C. Over the past decades, the average monthly temperature of the coldest month (January) is −15.2 °C and that of the warmest month (June) is +18.7 °C. The average annual precipitation is 600 mm (https://rp5.ru/, accessed on 1 May 2023).

The relief of study area is hilly plain. The soil cover consists mostly of grey forest soil and less often of dark grey forest soils, and according to the WRB classification [44], both soil types are Luvic Greyzemic Phaeozems. The forest vegetation belongs to the alliance Aconito lycoctoni-Tilion cordatae Solomeshch et Grigoriev in Willner et al. 2016 of the class Carpino-Fagetea sylvaticae Jakucs ex Passarge 1968, and is represented by secondary forb-grass broad-leaved forests with Tilia cordata Mill., Ulmus glabra Huds., Acer platanoides L., and Quercus robur L., and rarely by birch forests with a predominance of Betula pendula Roth.

The broad-leaved forest species (Aegopodium podagraria L., Dryopteris filix-mas (L.) Schott, Galium odoratum (L.) Scop., Lathyrus vernus (L.) Bernh., Pulmonaria obscura Dumort., Stellaria holostea L., Viola mirabilis L., etc.) are predominant in the herb layer.

In the vegetation of study area, the plant communities representing the different successional stages of reforestation in the abandoned agricultural lands (croplands) are most common. The fallow lands are overgrowing mainly with silver birch (Betula pendula), and rarely with aspen (Populus tremula L.) or goat willow (Salix caprea L.). The sites situated near the pine stands may also be overgrown with pine (Pinus sylvestris L.). The herb layer is dominated by forb species typical for meadow and forest edge communities belonging to the class Molinio-Arrhenatheretea (Galium album Mill., Leucanthemum vulgare Lam., Festuca pratensis Huds., Trifolium medium L., Agrostis diluta Kurczenko, Bromopsis inermis (Leyss.) Holub, Pimpinella saxifraga L., Achillea millefolium L., etc.), as well as for ruderal and synanthropic vegetation of the classes Artemisietea vulgaris and Sisymbrietea (Potentilla argentea L., Convolvulus arvensis L., Sonchus arvensis L., Potentilla anserina L., Taraxacum officinale F.H. Wigg., Cirsium setosum (Willd.) Besser, Picris hieracioides L., Artemisia vulgaris L., etc.).

The process of natural regeneration of vegetation begins after the withdrawal of agricultural lands from the crop rotation.

Regardless of the proximity of the fallow lands to the broad-leaved forest stands, the regeneration of lime, maple, elm, or oak in understory layers has not been observed. The overgrowth of these areas with birch is caused by high seed production, large numbers and the rapid growth of seedlings of this pioneer species, as well as the ability of its seeds to spread over a large distance [45,46].

According to estimates based on satellite images taken in 1985 at the current model/studied site, the restoration of forest vegetation occurs on 2445 ha of abandoned agricultural lands, 51% of which was previously ploughed. At present, the birch forests with a canopy area fraction of 50% or more and an average age of 15–25 years dominate [22].

2.2. Successional Stages of Reforestation on Fallow Lands

The chronosequence method is widely used to study the dynamics of various aspects of tree stand formation and to elucidate how the tree stand age impacts on stocks and carbon content of the biomass. However, the use of the chronosequence method for birch stands is still rare [47,48]. The use of this approach for the analysis of natural reforestation on abandoned agricultural lands is affected by several problems related to the high inter-site variability of biotic and abiotic habitat conditions and the influence of the history of agricultural use of fallow lands [42]. This leads to a large variation in timber stock and carbon content between sites representing similar stages of reforestation. Still, it is one of the main tools for such studies. In order to reduce the impact of inter-site variability within a succession stage, the sample plots with repetitions were selected to minimize the anthropogenic impact. For instance, the sites located near settlements or forest edges, as well as those grazed or logged, were rejected. The observed spatial and temporal heterogeneity of the overgrowth process on the basic model plot was formalized into 5 main stages of natural reforestation, within which two variants with different levels of tree stand cover density were distinguished (variant 1 with a low projective/density cover and variant 2 with a higher projective cover of the trees) (Figure 2 and Figure 3 and

Table 1. Characteristics of plant communities representing the different stages and variants of overgrowth with Betula pendula on fallow lands in the broad-leaved forest zone in the Cis-Ural.

Stage of Overgrowth	I	II	III	IV	V
Height of trees, m	0.5–1.5	2–3	5–8	9–14	15–18
Age of trees, years	3–8	9–14	15-20	20–25	25–30
Diameter of trunks, cm	–	1–3	6–8	10–14	16–20
Variant 1 (tree layer PC, %)	1–5	10–20	30–50	50–60	50–60
Variant 2 (tree layer PC, %)	7–10	30–50	60–80	75–90	75–90

Note: PC—Projective cover.

).

Within the plant communities of each successional stage, circular plots of 30 m diameter were selected to study in detail the species composition, timber stock, and sampling of biomass of woody and herbaceous plants, roots, litter, and soil.

Stage I is represented with the vegetation of fallow lands with a predominance of species typical for meadows, forest edges, and ruderal communities. At this stage, the renewal of tree species is represented by 3–8-years-old birch seedlings with a height of 0.5–1.5 m and projective cover up to 5% (rarely up to 10%). The projective cover of the herb layer varies from 35 to 90%. At this stage, six round sample plots with a diameter of 30 m were established, i.e., five plots representing variant 1 (three on grey forest soils and two on dark grey forest soils) and one plot representing variant 2 (on grey forest soil).

Stage II refers to the more developed tree stands of 2–3 m in height. The projective cover of the birch tree layer is 20–45%. The average age of trees is 9–14 years. The projective cover of the herb layer is 55–65%. At this stage, ten round sample plots were established, i.e., five plots representing variant 1 (four on grey forest soil and one on dark grey forest soil) and five plots representing variant 2 (two on grey forest soils and three on dark grey forest soils).

Stage III designates the plant communities with a well-developed tree layer of 5–8 m in height and a crown density of 40–65%. The average age of trees is 15–20 years. In comparison with stage II, the herb layer is sparser. The forests representing this stage are rare in the study area. At this stage, four sample plots were laid, including three plots representing variant 1 (two on grey forest soils and one on dark grey forest soil) and one plot representing variant 2 (on dark grey forest soil).

Stage IV refers to young 20–25-years-old birch forests of 9 to 14 m in height with a dense tree canopy and projective cover of 70–90%. The herb layer is not developed and its projective cover is 1–5% (rarely–10%). At this stage, seven sample plots were laid, including four plots representing variant 1 (three on grey forest soils and one on dark grey forest soil) and three plots representing variant 2 (on dark grey forest soil).

Stage V represents middle-aged birch stands of 15–18 m in height, forming a dense tree canopy with a crown density of 75–90%. The average age of trees is 25–30 years. The communities of this stage are similar to the previous ones in tree stand density and herb layer composition, but differ in age and height of the stand. Broad-leaved tree species may be sporadically present in the understory layer. In the study area, there were no birch forests belonging to variant 1 with a sparse tree layer. Therefore, eight sampling plots of variant 2 were laid (four on grey forest soils and four on dark grey forest soils).

2.3. Partitioning of Aboveground Biomass and Estimation of Production

Assessments of biomass and carbon stocks on fallow cropland overgrown with birch were conducted on 35 circular plots of 30 m in diameter. The above-ground biomass and the production of tree stands were estimated at the end of August, when the biomass formation processes were completed. The model tree method was used to estimate the timber stock. The mass of living trees and deadwood was estimated using the number of trees, the data on tree diameter at breast height (DBH, i.e., 1.3 m height above the ground), and the selection of model trees within each sample plot. In all cases, model trees were chosen and felled near the middle of the sample plot to avoid edge effects [42]. The samples of stem wood, branches, and leaves were dried at 70 °C to constant weight and weighed to the nearest 0.01 g.

2.4. Estimation of the Biomass of Herbaceous Plants and Litter Flux

To analyze the productivity of above-ground biomass of understory herbaceous plants during the period of their maximum development (at the end of July), we randomly set five square plots sized 50 × 50 cm on each of the 35 round sample plots. Within each square sample plot, the above-ground parts of herbaceous plants were cut and the dead biomass was sampled. The dead biomass included both plant litter (dead decomposing plants) and dead parts of herbs that have not yet lost their connection with living plants. In the communities of the later successional stages, dead biomass also included branches and partly decomposed small trunks of trees, that were deposited during the process of self-thinning of the stand. All samples were dried to an air-dry state and weighed with an accuracy of 0.01 g.

2.5. Estimation Root Biomass

The soil monolith method [49,50,51] was used to determine the root biomass. Root biomass was determined in the 0–30 cm soil layer. Five duplicated monoliths (10 × 10 × 30 cm) for each of the 35 studied plots were obtained. A total of 175 monoliths were taken. Before analysis, soil and non-organic material were carefully washed away from roots by running tap water and were then manually separated from organic debris. Because it is difficult to identify the species to which specific roots in the samples belong, roots were sorted only by diameter without considering plant species features. Roots were separated into the classes of fine (diameter < 1 mm) and coarse (diameter > 1 mm) roots. All root samples were oven-dried at 60 °C to a constant mass, and their weights were measured using analytical scales (VLTE-150, Gosmeter, Russia) with an accuracy of 0.001 g. For each monolith, the total root biomass (TRB) as well as the fine root biomass (FRB) were calculated with subsequent averaging for each stage of succession. The root biomass (TRB and FRB) was recalculated per ha (g per ha). Root sampling campaigns were carried out in September 2022 (completion of the growing season).

2.6. Soil Characteristic, Sampling, and Laboratory Analysis

For soil type determination and sampling, profile pits (3 m in length, 1 m in width, and 0.5 m in depth) were excavated in each plot. Soil samples were taken from every 0–10 cm layer until 50 cm, where, in most cases, the illuvial horizon (B) is formed. In order to increase the statistical reliability, the extra soil core samples (5 pcs within plot) were taken from a depth of 0–30 cm using a hand sampler (JMC, Newton, MA, USA; inner diameter: 4.5 cm) via a stratified random scheme. Generally, the soil organic carbon (SOC) balance is recommended for estimation in humus-accumulative horizons (A + AB). Olson and Al-Kaisi [52] and Qin et al. [53] suggested analyzing the soil layer to a deeper depth (≥1 m) for better understanding the SOC vertical distribution; however, Jobbagi and Jackson [54] noted that SOC mostly presented in topsoil (0–30 cm), and in this layer, the largest changes in SOC content are observed as a result of afforestation [33,34]. The soil samples (~300 g) were collected in a plastic bag and then delivered to a laboratory. The stones and tree/plant roots were removed from the samples, then samples were air-dried to constant weight, ground in a mortar, and sieved through a 2 mm sieve for further analysis. The SOC in the soil samples was determined by the wet-combustion method according to Tyurin [55] (direct analog of Walkley–Black method [56]) using a Specord M40 spectrophotometer (VEB Carl Zeiss, Jena, Germany). The SOC stocks were calculated based on SOC content and soil bulk density, particularly using the following formula: SOC stocks (t/ha) = SOC content (%) × L (soil layer, cm) × bulk density (g/cm³). The determination/calculation of bulk density (mass of oven-dried soil ÷ total soil volume) was made in an established manner [57,58]. For this, the undisturbed soil samples were taken from cross-sectional profiles using metal cylinders (10 cm height and diameter), which were hammered in every 10 cm (until 50 cm).

Based on morphological properties, the soil cover of study site is presented by grey forest and dark grey forest soils, with a “medium” thickness of humus-accumulative horizons (40–60 cm) and a clay loam texture of the topsoil.

2.7. Analysis of the Carbon Content of the Samples

Samples of live and dead biomass were ground with Vilitek cutting mills (VLM series) to a particle size of less than 0.5 mm. The finest parts of the roots were ground to a powder in porcelain mortars with liquid nitrogen. The carbon content in the samples was determined using a CHNS EA-3100 elemental analyzer (Eurovector, Pavia, Italy) in the Laboratory of physical and chemical methods of analysis (PCMA) at the Ufa Institute of Chemistry of UFRC RAS. The calculations of the quantitative content were provided by a special software package Weaver.

Based on the carbon content in plant and soil samples, conversion coefficients were calculated, which were compared with the conversion coefficients used to calculate the carbon content in the tree and herb layers of studied birch stands.

3. Results

The main carbon sequestration on fallow lands overgrowing by forest occurs in the growing tree stand and soil. The average carbon content in the trunk wood is 48.4%, and those in leaves and branches are 49.2% and 49.7%, respectively (

Table 2. Content of carbon accumulated by plant matter and the soil at different stages of overgrowth with Betula pendula on fallow lands in the broad-leaved forest zone in the Cis-Ural.

Stage (S) and Variant (V) of Overgrowth	Carbon Content, %
	Trunk Wood	Branches of Trees	Leaves of Trees	Above-Ground Part of Herbs	Mortmass	Roots of Trees and Herbs	Grey Forest Soil (0–30 cm)	Dark Grey Forest Soil (0–30 cm)
S I V 1	49.36 ± 0.22	51.55 ± 0.28	50.07 ± 0.16	42.41 ± 0.27	40.24 ± 0.73	36.81 ± 0.59	2.5 ± 0.09	5.32 ± 0.15
S I V 2	49.08 ± 0.10	51.87 ± 0.54	49.92 ± 0.02	42.42 ± 0.21	43.42 ± 0.34	34.59 ± 1.00	2.21 ± 0.08	-
S II V 1	48.39 ± 0.17	49.14 ± 0.12	49.15 ± 0.69	43.26 ± 0.20	43.4 ± 0.28	36.52 ± 0.59	2.22 ± 0.05	5.73 ± 0.35
S II V 2	48.70 ± 0.12	50.03 ± 0.34	48.87 ± 0.69	42.91 ± 0.15	43.32 ± 0.70	38.35 ± 0.78	3.01 ± 0.20	4.75 ± 0.20
S III V 1	48.10 ± 0.07	49.67 ± 0.27	48.87 ± 1.05	42.55 ± 0.15	45.43 ± 0.65	39.52 ± 0.67	3.57 ± 0.24	6.02 ± 0.19
S III V 2	48.04	48.83 ± 0.16	48.90 ± 0.20	39.19 ± 1.08	47.91 ± 0.54	39.88 ± 0.68	-	5.12 ± 0.18
S IV V 1	47.98 ± 0.06	49.06 ± 0.11	48.45 ± 0.53	41.88 ± 0.46	46.93 ± 0.43	42.09 ± 1.19	3.76 ± 0.35	6.22 ± 0.29
S IV V 2	47.90 ± 0.20	48.73 ± 0.23	48.48 ± 0.60	40.68 ± 0.65	47.25 ± 0.50	44.23 ± 0.85	-	4.61 ± 0.21
S V V 2	48.10 ± 0.12	48.82 ± 0.10	49.63 ± 0.56	42.7 ± 0.26	46.49 ± 0.47	44.82 ± 0.39	4.43 ± 0.20	5.89 ± 0.17

Note: ±—standard error.

). The dependence of the carbon content in the stand fractions and the age of trees was not established. We distinguished two variants in each stage of fallow land overgrowth—variant 1 with a sparser stand and variant 2 with a denser stand. At the time of the research, variants with sparse stands were identified only at stages I–IV of overgrowth and were absent in the communities of stage V. The carbon content in the above-ground parts of the herbs varied from 39.2 to 43.3% of dry organic matter with an average content of 42.0%. This value was lower than the carbon content in the above-ground biomass of birch and did not depend on the stage of succession or soil type.

The carbon content in the dead biomass varied more greatly than that in the herb layer. The percentage of carbon in the dead biomass was lowest in the first two successional stages (40.2–43.4%), and close to that in wood (46.5–47.3%) in stages IV and V (Figure 4).

The carbon content in the combined root biomass of trees and herbs varied from 34.6 to 44.8% and it increased together with the formation of the tree stand and the increase in the proportion of woody species in the root biomass during succession.

The carbon content in the grey forest soil increased almost twice during succession (comparing stages I and V), reaching 4.4% in stage V. In all successional stages, the carbon content in the dark grey forest soil was 1.5–2.5 times higher than that in grey forest soil, and was not increased during the reforestation. Thus, the increase in carbon stocks sequestrated in vegetation matter of plant communities during the studied succession (from stages I to V) was mainly caused by an increase in the biomass of the tree stand (Figure 5).

Table 3. Mean dry biomass at different stages of overgrowth with Betula pendula on fallow lands in the broad-leaved forest zone in the Cis-Ural.

Reforestation Stage	Soil	Variant	Biomass, kg/ha
			Leaves of Trees	Branches of Trees	Trunk Wood	Tree Layer	Above-Ground Biomass of Herb Layer	Mortmass	Underground Biomass
I	Grey forest soil	1	5.1 ± 1.2	2.3 ± 0.8	4.9 ± 1.2	12.2 ± 3.2	3498.5 ± 228.9	367.2 ± 49.4	2213.0 ± 223.2
		2	37.9	19.7	51.6	109.2	3235.6 ± 218.2	264.3 ± 52.7	2510.4 ± 50.4
	Dark grey forest soil	1	17.5 ± 10.6	9.7 ± 6.0	22.6 ± 14.1	49.7 ± 30.6	3428.0 ± 143.3	752.3 ± 53.7	2669.5 ± 3.2
Average at stage I			14.7 ± 6.0	7.7 ± 3.3	18.5 ± 8.3	40.9 ± 17.6	3403.2 ± 134.3	447.8 ± 48.7	2428.4 ± 121.0
II	Grey forest soil	1	139.3 ± 66.1	180.7 ± 91.6	294.5 ± 124.2	614.5 ± 280.2	3326,4 ± 186.2	900.7 ± 78.1	2452.4 ± 237.8
		2	553.2 ± 233.4	770.6 ± 332.1	2252.6 ± 1272.2	3576.4 ± 1837.7	2843.8 ± 359.2	1334.6 ± 288.7	2452.4 ± 529.8
	Dark grey forest soil	1	194.6	349.8	383.2	927.6	2995.5 ± 611.3	1261.4 ± 380.8	2063.0 ± 0.1
		2	396.3 ± 84.2	500.0 ± 99.5	1528.2 ± 323.1	2424.5 ± 152.3	1732.4 ± 115.6	2744.9 ± 458.3	3500.6 ± 503.8
Average at stage II			304.7 ± 78.9	411.4 ± 107.8	1065.1 ± 373.7	1781.2 ± 537.0	2718.6 ± 157.0	1576.8 ± 192.2	2727.9 ± 264.6
III	Grey forest soil	1	2281.5 ± 1140.5	6207.5 ± 2717.3	18,526.2 ± 8753.5	27,015.2 ± 12,611.2	1036.3 ± 210.4	4246.0 ± 462.9	4878.9 ± 1245.3
	Dark grey forest soil	1	842.2	2870.7	7506.2	11,219.1	3150.2 ± 599.2	3175.0 ± 503.9	6177.6 ± 0.1
		2	2602.3	6668.4	31,030.7	40,301.4	127.2 ± 34.0	6704.8 ± 626.3	5946.8 ± 0.1
Average at stage III			2001.9 ± 664.5	5488.5 ± 1557.5	18,897.3 ± 6040.2	26,387.7 ± 8141.9	1095.4 ± 265.4	5015.3 ± 435.2	5565.8 ± 559.0
IV	Grey forest soil	1	2632.1 ± 424.2	8408.8 ± 1565.5	34,854.1 ± 8328.2	45,895.1 ± 9428.6	762.9 ± 165.6	4297.4 ± 978.7	3248.9 ± 553.3
	Dark grey forest soil	1	1763.3	11,943.8	70,384.0	84,091.2	222.0 ± 103.6	9607.7 ± 1047.2	2444.4 ± 0.1
		2	2485.6 ± 484.9	5573.0 ± 690.2	62,281.5 ± 3662.0	70,340.0 ± 4818.0	141.4 ± 44.8	7039.5 ± 424.9	4202.6 ± 257.8
Average at stage IV			2445.2 ± 296.6	7698.5 ± 1101.4	51,684.4 ± 6823.2	61,828.0 ± 7119.4	419.3 ± 90.4	6231.2 ± 576.0	3542.7 ± 353.5
V	Grey forest soil	2	3346.3 ± 266.8	6990.0 ± 201.1	104,752.8 ± 6064.8	115,089.1 ± 6342.6	200.1 ± 37.5	9734.0 ± 843.0	3746.9 ± 168.5
	Dark grey forest soil	2	6752.0 ± 1035.3	12,907.8 ± 1780.5	166,034.3 ± 12870.4	185,694.2 ± 15367.9	335.6 ± 45.7	12,126.7 ± 948.8	4674.9 ± 153.9
Average at stage V			5292.5 ± 876.8	10,371.6 ± 1505.9	139,770.8 ± 13,864.7	155,434.9 ± 16,090.7	284.8 ± 33.5	11,229.5 ± 696.5	4326.9 ± 196.2

Note: ±—standard error.

shows the data on the biomass of trees, understory vegetation, and plant mortmass, based on which the carbon stocks accumulated by plant matter were calculated for the communities of different stages of overgrowth of Betula pendula on fallow lands in the broad-leaved forest zone in the Cis-Ural.

In the plant communities of the first three successional stages growing on both grey and dark grey forest soils, the stock of carbon sequestrated in the stand was characterized by significant variability between sample plots within each stage of overgrowth (

Table 4. Stocks of carbon accumulated by plant–root–soil system at different stages of overgrowth with Betula pendula on fallow lands in the broad-leaved forest zone in the Cis-Ural.

Reforestation Stage	Soil	Variant	Carbon Stocks, kg/ha
			Leaves of Trees	Branches of Trees	Trunk Wood	Tree Layer	Aboveground Biomass of Herb Layer	Mortmass	Underground Biomass	Soil (0–30 cm), t/ha
I	Grey forest soil	1	2.5 ± 0.5	1.2 ± 0.3	2.4 ± 0.4	6.1 ± 1.2	1492.2 ± 98.8	149.7 ± 20.3	810.1 ± 47.1	97.6 ± 3.5
		2	18.9	10.2	25.3	54.5	1373.1 ± 93.2	114.8 ± 22.7	868.4 ± 25.2	86.3 ± 3.3
	Dark grey forest soil	1	8.8 ± 5.3	4.9 ± 3.0	11.1 ± 6.9	24.8 ± 15.2	1443.0 ± 62.3	296.6 ± 23.4	1000.4 ± 16.3	207.4 ± 5.9
Average at stage I			7.3 ± 3.0	3.9 ± 1.7	9.1 ± 4.1	20.4 ± 8.8	1444.1 ± 58.0	181.7 ± 19.7	881.1 ± 26.3	126.7 ± 8.7
II	Grey forest soil	1	67.2 ± 31.9	88.6 ± 45.0	143.4 ± 60.6	299.2 ± 136.5	1441.5 ± 80.9	391.7 ± 33.0	889.7 ± 41.7	86.5 ± 2.1
		2	268.0 ± 109.2	374.7 ± 160.5	1101.3 ± 622.5	1744.1 ± 892.2	1201.7 ± 146.8	590.8 ± 130.4	915.7 ± 94.3	117.3 ± 7.7
	Dark grey forest soil	1	98.0	173.9	184.7	456.5	1285.8 ± 260.1	535.3 ± 158.7	766.5 ± 11.0	223.4 ± 13.5
		2	196.0 ± 42.4	249.3 ± 49.5	742.8 ± 156.7	1188.1 ± 69.3	747.8 ± 49.0	1210.7 ± 205.4	1387.7 ± 113.5	185.3 ± 8.0
Average at stage II			149.1 ± 38.0	202.6 ± 52.5	518.9 ± 182.6	870.6 ± 261.4	1169.9 ± 67.0	691.6 ± 85.6	1032.0 ± 53.9	136.0 ± 7.3
III	Grey forest soil	1	1147.5 ± 586.2	3100.0 ± 1345.4	8896.5 ± 4181.4	13,144.0 ± 6113.1	441.0 ± 90.3	1745.5 ± 285.3	1912.6 ± 211.1	139.3 ± 9.4
	Dark grey forest soil	1	419.6	1411.0	3619.5	5450.1	1344.7 ± 257.2	1407.6 ± 281.2	2441.6 ± 60.5	237.9 ± 7.4
		2	1269.6	3254.7	14,907.1	19,431.4	47.4 ± 11.6	3195.0 ± 279.7	2371.5 ± 40.5	199.7 ± 6.8
Average at stage III			996.1 ± 338.0	2716.4 ± 771.7	9080.0 ± 2891.6	12,792.4 ± 3934.7	464.3 ± 113.8	2257.7 ± 229.7	2202.0 ± 99.0	182.0 ± 8.0
IV	Grey forest soil	1	1307.3 ± 229.3	4135.1 ± 768.7	16751.7 ± 4007.8	22194.1 ± 4556.4	323.5 ± 70.3	2049.4 ± 475.5	1277.0 ± 72.5	146.7 ± 13.6
	Dark grey forest soil	1	837.5	5857.2	33,629.5	40,324.1	87.3 ± 40.5	4500.9 ± 561.9	1142.8 ± 31.9	242.5 ± 11.2
		2	1228.5 ± 253.5	2733.0 ± 355.2	29,819.8 ± 1689.0	33,781.2 ± 2291.9	57.5 ± 18.4	3009.2 ± 269.8	1858.8 ± 61.5	179.8 ± 8.0
Average at stage IV			1206.4 ± 157.8	3780.2 ± 542.8	24,763.4 ± 3251.7	29,750.0 ± 3398.6	175.8 ± 38.4	2811.0 ± 283.9	1507.2 ± 66.2	174.6 ± 7.9
V	Grey forest soil	2	1707.9 ± 132.9	3395.0 ± 88.0	50,185.5 ± 2934.4	55,288.4 ± 3064.4	84.3 ± 15.9	4379.2 ± 502.0	1666.8 ± 34.0	173.0 ± 7.8
	Dark grey forest soil	2	3298.9 ± 441.9	6340.9 ± 875.2	79,882.5 ± 6192.2	89,522.3 ± 7371.1	143.5 ± 19.2	5519.1 ± 455.9	2104.1 ± 41.7	229.9 ± 6.4
Average at stage V			2617.0 ± 394.4	5078.4 ± 745.1	67,155.2 ± 6704.9	74,850.6 ± 7776.1	121.3 ± 14.2	5091.6 ± 352.5	1940.1 ± 44.3	201.4 ± 6.5

Note: ±—standard error.

). Starting from the IV successional stage, higher rates of carbon sequestrated by the tree layer in the communities growing on the dark grey forest soils were observed. At stage V of reforestation, the average total carbon stock in the above-ground part of woody plants (including leaves, branches, and trunk wood) was estimated as 55.3 t/ha in the stands formed on grey forest soils and 89.5 t/ha in the forests on more fertile dark grey forest soils. During the course of reforestation, the proportion of carbon sequestrated in leaves and branches in the sum of carbon sequestrated by the above-ground part of the tree stand uniformly decreased (in leaves from 22.2 to 3.1%, in branches from 32 to 5.8%), while the proportion of carbon sequestrated in trunk wood significantly increased (from 45.9 to 91.2%).

The maximum stocks of carbon accumulated by above-ground phytomass of the herb layer (1.4 t/ha) were typical for stage I of overgrowth, when there is no coenotic pressure of the tree layer on herbs. The minimum stocks of carbon in the above-ground phytomass of the herb layer were observed at stages IV and V of overgrowth in the communities with a dense crown canopy (0.18 and 0.12 t/ha, respectively). At the same time, there was no dependence on soil types. Root biomass stocks increased depending on the stage of overgrowth, which may be explained by the increasing participation of roots of woody plants. During the course of overgrowth, the proportion of carbon stocks of the above-ground part of the herb layer in the total amount of carbon deposited by plant matter decreased from 60.7% at stage I to 0.14% at stage V.

In the communities of successional stage I, the proportion of carbon stocks in the mortmass represented 6% of the total carbon stock in plant matter. During the course of succession, this value increased to 18.3% at the second stage, and then decreased to 6.2% at stage V. Minimum stocks of carbon accumulated in mortmass, mainly consisting of dead decomposing herbs, were typical for stage I of overgrowth (0.18 t/ha). The maximum stocks of carbon accumulated in mortmass were revealed in the communities at stage V of overgrowth (5.09 t/ha). Thus, there is a clear tendency of an increase in stocks in mortmass depending on the stage of overgrowth of a fallow lands.

At all successional stages, the carbon stocks in dark grey forest soil were on average twice higher than those in the grey forest soil. At the same time, the noticeable dependence on the successional stage changes of the carbon stocks in the upper 0–30 cm layer of grey forest soil was revealed only from stage III of overgrowth with an increase of 66% at stage V of overgrowth. In the more fertile and humus-rich dark grey forest soil, this process was less pronounced, and the content of organic carbon increased by 10% and its stocks increased by 7% by stage V of overgrowth (Figure 6).

4. Discussion

4.1. Percentage of Carbon in Plant Substance and Soil

In the study area, the main carbon deposition in the above-ground parts of plants in communities at the middle and late stages of forest stand formation occurs in stem wood, which is consistent with the literature data [42]. The carbon content in birch stem wood (on average 48.5%) is close to similar data obtained in the hemiboreal zone of southeastern Estonia (47.0 to 49.3%) [42] and is somewhat higher than in the middle taiga subzone of the Komi Republic, Russia (45.6%) [59]. The calculation of carbon stocks in wood using an average conversion coefficient of 0.5 may give somewhat overestimated values [60,61]. The conversion coefficient calculated to estimate the carbon stock in birch forests on abandoned croplands of the Cis-Ural (0.485) exceeds the similar conversion coefficient recommended for young stands (0.437) and middle-aged stands (0.396) of silver birch [62]. In our studies, the decrease in the carbon content in stem wood is not so significant (from 49.4% in small undergrowth at stage I to 48.1% in the middle-aged forest stand). Therefore, to calculate the carbon stock in silver birch stem wood in the study area, we decide to use the average conversion coefficient. This allows us to conclude that the use of the recommended conversion coefficients in the “Methodological guidelines…” [62] may lead to an underestimation of carbon stocks in young and middle-aged birch forests on abandoned farmlands. In addition, this indicates the need to improve the system of regional conversion coefficients for calculating carbon stocks in birch stands, especially when studying birch forests on abandoned agricultural lands, which are highly productive and fast growing, in comparison with birch stands within forest areas [63,64,65].

In the study area, the carbon content in the above-ground part of the herb layer is similar to analogical data on birch stands growing on abandoned agricultural lands in Estonia [42]. The lack of interrelation between the carbon content in the above-ground biomass of the herb layer and the successional stages makes it possible to use the conversion coefficient of 0.42 to calculate the carbon content in the herb layer. In the plant communities of the later successional stages, the biomass of the herb layer is usually negligible, so in some cases, it can be neglected [66].

The carbon content in the mortmass seems to be dependent on the differences in the mortmass composition at different successional stages. In the first two stages of reforestation, the mortmass is mainly represented by the dead parts of herbaceous plants, whereas at stages IV and V with self-thinning of the stand, it mainly consists of birch woody residues and leaves, whose percentage of carbon is higher and close to that in wood.

In the early two successional stages, the carbon content in the underground phytomass is 34–38% and is represented largely by the roots of herbaceous plants. These values are lower than the percentage of carbon in the above-ground part of the herb layer. In the communities of later successional stages, the proportion of woody roots in the composition of the underground biomass becomes large, and the carbon content increases significantly (up to 44.8%). Thus, several conversion coefficients should be used to estimate carbon deposition by underground phytomass at different stages of reforestation, which coincides with the literature data [67].

Different literature sources provide quite contradictory data on the carbon content in soils. In some cases, there are no statistically reliable differences in the soil carbon content in the areas with stands of different ages [68,69,70]. In contrast, another study noted significant changes in the top 10 cm of soil due to additional carbon sequestration from decomposed roots and litter and an increase in the percentage of carbon in soils as stands formed [42].

Some studies have reported increases in SOC content on the areas with initial stages of tree stand formation [71], and several other investigations have reported slight increases in SOC content with increasing stand age [72,73,74]. In our case, the increase in stand age is accompanied by an increase in percentage of carbon in the grey forest soil from 2.5% at successional stage I to 4.4% at stage V, whereas the percentage of carbon content in the more fertile dark grey forest soil is almost unchanged at the different stages.

4.2. Phytomass and Carbon Stocks in Plant Matter and Soil

In the communities of the first three successional stages, the significant role in calculating the above-ground phytomass and stand carbon stocks belongs to the mosaic distribution and age differences of stands of the sample plots. These differences have a stronger effect on the variability in stand carbon stocks than soil conditions. In the communities of stages IV and V, the dense stand makes for a stronger competition for mineral nutrition among the plants. In this regard, the role of soil differences increases, and the phytomass and carbon stock in the tree stands in the communities on dark grey forest soils are higher than in those growing on less fertile grey forest soils. The content of carbon accumulated in the above-ground part of trees in the forest stands of successional stage V (25–30 years after beginning of reforestation) growing on the dark grey forest soils (89.5 t/ha) is approximately equal to the carbon stock in silver birch mature stands (60 years) on overgrown arable land in Estonia (89.2 t/ha) [42]. Several publications contain detailed information on tree biomass, but there are no available data on carbon stocks. The use of an average conversion coefficient of 0.5 [42] allows a comparison of our results with the data on carbon stocks in European birch forests [75,76,77,78,79,80]. The average carbon stocks in the 25–30-year birch tree stand on abandoned agricultural land (74.9 t/ha) in the study area are lower than those in Sweden, where above-ground tree biomass of the 32-year stand reaches 175.3 t/ha, which approximately corresponds to 87.7 t/ha of deposited carbon [75]. In other studies, the values of tree biomass are higher than those in the Cis-Urals. For instance, in the study area, the value of average tree biomass in 15-20-years-old forests on dark grey forest soils (40.3 t/ha) are lower than similar data for the Czech Republic (75 t/ha for 15-year birch stands) [76] and Poland (166.5 t/ha for 19-year birch stands) [77]. In Sweden, the value of above-ground tree biomass of 22–23-year stands (106.1–108.8 t/ha) [78] is higher than tree biomass in the tree stands of the same age growing on the dark grey forest soils in the Cis-Ural (70.3–84.1 t/ha). However, in the study area, the growth rate of the tree biomass in 25–30-years-old forest stands increases significantly, and the above-ground biomass of stand doubles and reaches 155 t/ha.

The average rate of carbon sequestration in above-ground biomass of trees in the studied area is 2.7 t/ha/year, which is 35% higher than the rate of carbon sequestration in Estonian birch forests [42]. At the same time, the maximum carbon sequestration rate in Estonian birch forests is 5.9 t/ha/year [42]. In the Czech Republic, the average annual growth of biomass (MAI) of silver birch trees is highest in the 15–25-years-old stand and reaches 5.0–6.5 t/ha of dry biomass, which is approximately equal to 2.5–3.2 t/ha of deposited carbon per year [78]. The older stands were not considered in the study mentioned above, so it is possible that the MAI of trees could be revealed in stands of older age. In Sweden, the average annual growth of biomass of 8–30-year stands ranges from 0.71 to 8.44 t/ha, with a peak growth at the stand age of 10 to 30 years [75]. In the Cis-Ural, it is highest at the age of 25–30 years and amounts to 9 t/ha of deposited carbon per year.

The stock of organic carbon in soils is a result of the difference between litter input and the decomposition of soil carbon over time [81]. Forest soils store significant amounts of carbon, often more than forest vegetation [72]. According to the EC/UN-ECE (2003-1) report, forest soils in Europe contain about 1.5 times more carbon than trees. At the landscape level in Finland, the potential carbon sink of soils was estimated to be 30–70% of the sink of trees, depending on forest management [82]. Based on the carbon balance of European forests, it was shown that the carbon flux into the soil is about two-thirds of the carbon sink of trees found [81]. The birch forests formed on abandoned arable lands are characterized by a high proportion of annual litter biomass, which includes branches and tree trunks dying off during the process of self-thinning. According to our data, in the communities of stage V of the overgrowth of fallow lands, the carbon stocks in the 30 cm soil layer are 2.7 times greater than the carbon stocks in the stand. At the same time, in the course of reforestation succession, the stock of organic carbon in the grey forest soil after 30 years of reforestation increases by 80.6 t/ha, while the increase in carbon stock in the biomass of these forests is only 59.1 t/ha. Phytomass and carbon stocks in 25–30-years-old birch forests on dark grey soil richer in nutrients are significantly higher than those in the forests on grey forest soils, while organic carbon stock remains practically unchanged. This coincides with the data on postagrarian birch forests on rich soils studied in Estonia [42]. Thus, there is no significant increase in the organic carbon pool in the richer soils under forests, due to the higher growth rates of trees.

5. Conclusions and Implications

On abandoned agricultural lands overgrown with trees, the main carbon sequestration occurs in soils. At all stages of restorative succession, the carbon stock in the 30 cm soil layer significantly exceeded the carbon stock in the wood. In the communities of middle and late stages of succession, among different fractions of the above-ground part of plants, the main carbon deposition occurs in stem wood. To avoid an underestimation of carbon stock value in the trunk wood, caused by the use of averaged conversion coefficients recommended for Russia, it is necessary to develop a system of regional conversion coefficients for birch postagrogenic forests/lands. The absence of dependence of the carbon content in the above-ground herb biomass on the succession stages allows the use of a single conversion coefficient. As the carbon content in mortmass and under-ground phytomass depends on the succession stage, it is required to use different conversion coefficients for the communities of different successional stages. With the increase in tree stand age, the carbon content increases in the grey forest soil (depleted by nutrients), and it does not change in the richer dark grey forest soil.

The conducted studies confirm the possibility of exploiting abandoned arable lands overgrowing by forest of the Cis-Ural in terms of creating of carbon farms. Taking into account the highest rate of carbon sequestration by trees at successional stage V, the carbon farms may be created on the areas with birch stands of 20–30-years-old. Climate change trends should be considered in the long-term prediction for tens of years, as they affect the growth rate and phenological development of trees [83,84,85,86]. In the Southern Ural, a decrease in summer precipitation and an increase in winter precipitation are predicted [87]. These changes will be favorable for the expansion of woody vegetation but may reduce the carbon sequestration by trees under hotter and drier growing seasons. On the other hand, an increase in forest areas may reduce the negative impacts of current and projected climate changes [88].