Total Soil CO2 Efflux from Drained Terric Histosols

Histosols cover about 8–10% of Lithuania’s territory and most of this area is covered with nutrient-rich organic soils (Terric Histosols). Greenhouse gas (GHG) emissions from drained Histosols contribute more than 25% of emissions from the Land Use, Land Use Change and Forestry (LULUCF) sector. In this study, as the first step of examining the carbon dioxide (CO2) fluxes in these soils, total soil CO2 efflux and several environmental parameters (temperature of air and topsoil, soil chemical composition, soil moisture, and water table level) were measured in drained Terric Histosols under three native forest stands and perennial grasslands in the growing seasons of 2020 and 2021. The drained nutrient-rich organic soils differed in terms of concentrations of soil organic carbon and total nitrogen, as well as soil organic carbon and total nitrogen ratio. The highest rate of total soil CO2 efflux was found in the summer months. Overall, the rate was statistically significant and strongly correlated only with soil and air temperature. A trend emerged that total soil CO2 efflux was 30% higher in perennial grassland than in forested land. Additional work is still needed to estimate the net CO2 balance of these soils.


Introduction
Globally, peat soils (Histosols) cover approximately 3% of the land area and have accumulated about one third of total soil organic carbon stocks [1][2][3].One third of all organic soils are found in Europe [4].Peatlands have an extremely important role in the global carbon cycle.Globally, the European Union (EU) is the second largest source of greenhouse gas (GHG) emissions from drained peatlands, with 220 Mt CO 2 -C annual emissions from peatlands.Lithuania is recognized as one of 11 EU countries in which GHG emissions from drained Histosols contribute more than 25% of Land Use, Land Use Change and Forestry (LULUCF) emissions [5].
Like other EU member states, Lithuania is obligated by the Kyoto Protocol to report annual GHG emissions from LULUCF in its national GHG inventory report.Currently, the State Forest Service uses default CO 2 -C annual emission factors (EF) to represent drained peatlands when reporting GHG inventories in Lithuania [13].These Tier 1 EFs are determined by using relatively wide categories for averaging annual soil carbon balance values from peer-reviewed studies performed in boreal and cold temperate regions [14].More recently, the CO 2 -C EFs were significantly increased from 0.68 to 2.6 t C ha −1 y −1 for forest land and from 0.25 to 6.1 t C ha −1 y −1 for fertile grasslands on drained organic soils in temperate regions [15].
Total soil CO 2 efflux is the second largest carbon flux (after photosynthesis) in terrestrial ecosystems [16] and is a result of autotrophic (CO 2 produced by vegetation living roots) and heterotrophic (CO 2 produced by soil microbes in the decomposition of organic matter deposited into soil and on the soil surface) soil respiration [17][18][19].The total soil respiration rate depends on many variables such as local climate, land use, nutrient availability, soil temperature, soil moisture, and water table level (WTL) [20][21][22][23].Generally, lower WTL in drained areas has a positive effect on oxygen availability, which can enhance oxidation and decomposition at greater peat layer depth from the soil surface [24,25].Furthermore, soil CO 2 efflux strongly depends on soil organic matter (SOM) decomposition due to increased biomicrobial biomass and activity [26,27].It is well documented that soil organic carbon (SOC), total nitrogen (TN), and the soil organic carbon/total nitrogen (C/N) ratio are the main indicators of SOM quality [28,29], and these parameters usually differ between forest and grassland ecosystems [30].
Despite numerous studies investigating autotrophic and heterotrophic soil CO 2 efflux from drained Histosols, studies on soil CO 2 emission trends in hemi-boreal regions are still scarce and form an urgent need for data applicable to regional or local conditions [4].
In this study, we monitored total soil CO 2 efflux and environmental factors such as air temperature, soil physical parameters (temperature and moisture of the topsoil), soil chemical parameters (SOC, TN, and C/N in the topsoil), and soil water level table in three forest types and perennial grassland.These types of land cover are commonly found in drained nutrient-rich organic soils, more specifically Terric Histosols, in Lithuania.The objectives were: (1) to determine the concentrations of SOC and TN in the topsoil of drained Terric Histosols under stands of different tree species and perennial grassland; (2) to investigate the variation in total soil CO 2 efflux across different months and between stands of different tree species and perennial grassland; and (3) to assess the environmental factors that may have the greatest influence on total soil CO 2 efflux in drained Terric Histosols under perennial grassland and forest stands.The hypothesis of this study is that the concentrations of SOC and TN, along with the C/N ratio, are the key factors influencing total soil CO 2 efflux in forest and grassland biotopes.

Concentrations of Soil Organic Carbon and Total Nitrogen in Drained Terric Histosols
The variation between the study sites in the concentrations of soil organic carbon (SOC) and total nitrogen (TN) as well as the SOC/TN (C/N) ratio in the drained Terric Histosols was statistically significant (Figure 1).
The highest mean concentration of SOC (46%), highest C/N ratio (25), and the lowest concentration of TN (1.7%) were in the topsoil of the Norway spruce stand (Figure 1).A high mean concentration of SOC (43%) was also found in the topsoil of the silver birch stand, while the C/N ratio was intermediate (19).The concentration of TN in all study sites, except the spruce stand, was similar and ranged between 2.0% and 2.3%.Significantly lower concentrations of SOC (33% in perennial grassland and 31.6% in the black alder stand) and lower C/N ratios (14.4 in grassland and 16.8 in the black alder stand) were observed in the topsoil of the perennial grassland and the black alder stand.We interpreted the relatively Plants 2024, 13, 139 3 of 13 high concentration of SOC and high C/N ratio as indicating a lower decomposition rate of peat in the silver birch and, especially, the Norway spruce stand, while the opposite results in black alder and perennial grassland were taken to indicate a high peat decomposition rate and high nitrogen release during the decay process [31].The highest mean concentration of SOC (46%), highest C/N ratio (25), and the low concentration of TN (1.7%) were in the topsoil of the Norway spruce stand (Figure 1) high mean concentration of SOC (43%) was also found in the topsoil of the silver bi stand, while the C/N ratio was intermediate (19).The concentration of TN in all stu sites, except the spruce stand, was similar and ranged between 2.0% and 2.3%.Sign cantly lower concentrations of SOC (33% in perennial grassland and 31.6% in the bl alder stand) and lower C/N ratios (14.4 in grassland and 16.8 in the black alder stand) w observed in the topsoil of the perennial grassland and the black alder stand.We in preted the relatively high concentration of SOC and high C/N ratio as indicating a low decomposition rate of peat in the silver birch and, especially, the Norway spruce sta while the opposite results in black alder and perennial grassland were taken to indica high peat decomposition rate and high nitrogen release during the decay process [31].

Total Soil CO2 Efflux under Forest and Grassland Sites
In 2021, total soil CO2 efflux in the grassland site and the three forest sites varied fr 0.12 g CO2 m −2 h −1 to 1.5 g CO2 m −2 h −1 (Figure 2).The lowest values for total soil CO2 effl (0.12-0.45 g CO2 m −2 h −1 ) occurred in the spring and in the autumn.Higher values of m total soil CO2 efflux were detected in the summer, ranging from 0.5 g CO2 m −2 h −1 to 1. CO2 m −2 h −1 .Furthermore, in summer the differences between the sites became appar and the highest total soil CO2 efflux was observed in the perennial grassland site, rang on average from 0.8 g CO2 m −2 h −1 to 1.5 g CO2 m −2 h −1 , while the effluxes in the forest s were lower with averages of between 0.5 and 0.9 g CO2 m −2 h −1 .

Total Soil CO 2 Efflux under Forest and Grassland Sites
In 2021, total soil CO 2 efflux in the grassland site and the three forest sites varied from 0.12 g CO 2 m −2 h −1 to 1.5 g CO 2 m −2 h −1 (Figure 2).The lowest values for total soil CO 2 efflux (0.12-0.45 g CO 2 m −2 h −1 ) occurred in the spring and in the autumn.Higher values of mean total soil CO 2 efflux were detected in the summer, ranging from 0.5 g CO 2 m −2 h −1 to 1.5 g CO 2 m −2 h −1 .Furthermore, in summer the differences between the sites became apparent and the highest total soil CO 2 efflux was observed in the perennial grassland site, ranging on average from 0.8 g CO 2 m −2 h −1 to 1.5 g CO 2 m −2 h −1 , while the effluxes in the forest sites were lower with averages of between 0.5 and 0.9 g CO 2 m −2 h −1 .
Plants 2024, 13, x FOR PEER REVIEW 4 of 13 The results obtained in 2022 indicated that total soil CO2 efflux increased at the beginning of the vegetative period, in a similar pattern as was found in 2021 (Figure 2).In spring (April and May), total soil CO2 efflux varied on average between 0.37 and 0.51 g CO2 m −2 h −1 in the perennial grassland and in the silver birch stand, and on average be- The results obtained in 2022 indicated that total soil CO 2 efflux increased at the beginning of the vegetative period, in a similar pattern as was found in 2021 (Figure 2).In spring (April and May), total soil CO 2 efflux varied on average between 0.37 and 0.51 g CO 2 m −2 h −1 in the perennial grassland and in the silver birch stand, and on average between 0.27 and 0.29 g CO 2 m −2 h −1 in the Norway spruce stand.In June, total soil CO 2 efflux increased on average by 40% in the Norway spruce stand, from 40% to 50% in the perennial grassland, and from 40% to 55% in the silver birch stand, as compared with the total soil CO 2 efflux values found in May and April, respectively.Furthermore, the highest total soil CO 2 efflux (an average of 0.8 g CO 2 m −2 h −1 ) was determined in the black alder stand in June, possibly for the reasons discussed in Section 4.2.

The Influence of Environmental Factors on Total Soil CO 2 Efflux
The variations of environmental factors such as air and soil temperature, soil moisture, and WTL are presented in Table 1.The data obtained in our study led us to evaluate the effect of environmental factors on total soil CO 2 efflux (Figure 3, Table 2).The studied environmental factors such as soil chemical properties (SOC, TN, and the C/N ratio in the topsoil), WTL, soil and air temperature (in the topsoil), different forms of land use (forest and grassland sites), and the depth of drainage ditches were included in principal components analysis (PCA) (Figure 3).
The data obtained in our study led us to evaluate the effect of environmental factors on total soil CO2 efflux (Figure 3, Table 2).The studied environmental factors such as soil chemical properties (SOC, TN, and the C/N ratio in the topsoil), WTL, soil and air temperature (in the topsoil), different forms of land use (forest and grassland sites), and the depth of drainage ditches were included in principal components analysis (PCA) (Figure 3).PCA of soil chemical properties (SOC, TN, and the C/N ratio in the topsoil), water table level, soil and air temperature (in the topsoil), drainage (the depth of drainage ditches), and total soil CO2 efflux.The x-axis represents principal component 1, and the y-axis component 2. Abbreviations: CO2_efflux-soil total CO2 efflux; Tair-air temperature; Tsoil-soil temperature at 10 cm depth; WTL-water table level; Moisture-soil moisture at 5 cm depth, SOC-soil organic carbon determined at 0-10 cm soil depth soil; TN-total nitrogen determined at 0-10 cm soil depth; CN_ratio-SOC/TN ratio determined at 0-10 cm soil depth; Drainage-depth of the drainage ditches.
PCA indicated a strong positive relationship between air and soil temperatures and total soil CO2 efflux (Figure 3, Component 2).The efflux further showed a negative relationship to wetness (variables WTL and moisture).The depth of drainage ditches was associated with the SOC concentration and C/N ratio in the topsoil (Figure 3, Component 1) of drained Terric Histosols.TN was not included in any of the groups, because TN concentrations were similar in all studied forest and grassland sites (Figure 1).WTL-water table level; Moisture-soil moisture at 5 cm depth, SOC-soil organic carbon determined at 0-10 cm soil depth soil; TN-total nitrogen determined at 0-10 cm soil depth; CN_ratio-SOC/TN ratio determined at 0-10 cm soil depth; Drainage-depth of the drainage ditches.PCA indicated a strong positive relationship between air and soil temperatures and total soil CO 2 efflux (Figure 3, Component 2).The efflux further showed a negative relationship to wetness (variables WTL and moisture).The depth of drainage ditches was associated with the SOC concentration and C/N ratio in the topsoil (Figure 3, Component 1) of drained Terric Histosols.TN was not included in any of the groups, because TN concentrations were similar in all studied forest and grassland sites (Figure 1).
The correlation analysis also indicated that the strongest significant and positive linear relation (r = 0.7, p < 0.05) was found between air and soil temperature and total soil CO 2 efflux (Table 2).Total soil CO 2 efflux strongly depended on air and soil (at 10 cm depth) Plants 2024, 13, 139 6 of 13 temperatures and total soil CO 2 efflux tended to increase with increasing air and soil temperatures (Table 1; Figure 2).
Soil moisture (r = −0.4,p < 0.05) and WTL (r = −0.3,p < 0.05) were moderately and negatively correlated with total soil CO 2 efflux, indicating that increasing moisture and higher WTL resulted in lower efflux.Meanwhile, a weak positive relation (r = −0.2,p > 0.05) was observed between total nitrogen (TN) concentration in the topsoil and total soil CO 2 efflux and, conversely, a weak negative (r = 0.2, p > 0.05) relation was found between SOC and SOC/TN (C/N) ratio.

Discussion
Numerous studies investigating soil CO 2 efflux from drained peatlands have been conducted in Fennoscandia and in the Baltic countries.This may be due to the fact that peatlands cover 31-34% of Sweden and Finland, around 20% of Estonia, around 10% of Latvia and Lithuania, and about 7.7% of Norway's territory [32][33][34][35].Several studies in which total soil CO 2 efflux measurements were performed using a closed chamber system were selected for comparison (Table 3).A study conducted in the boreal region of Russia reported that the total soil CO 2 efflux released from Histosols in black alder forests could be approximately 0.33 g CO 2 m −2 h −1 [36].Another study performed in Finland reported that the annual total CO 2 efflux in less fertile forests was an average of 2000 g CO 2 m −2 y −1 , and an average of 3000 g CO 2 m −2 y −1 in fertile forests [37].Meanwhile, total soil CO 2 effluxes in perennial grasslands and agricultural croplands have been found to have similar values [38], with annual emissions were approximately 5-7 t CO 2 ha −1 y −1 [39].The total soil CO 2 efflux in grasslands can be about two times higher during the growing season in the hemi-boreal zone than in the boreal zone (Table 3).In this study, the highest total soil CO 2 efflux released from perennial grassland in summer (average of 1.16 g CO 2 m −2 h −1 ) is well in line with a recent study evaluating total soil CO 2 efflux in Histosols under grasslands in Sweden and Latvia [40,41].Also, similar results of total soil CO 2 efflux from forest sites were obtained in the study conducted in Latvia [41].Studies investigating annual CO 2 emissions have reported that total soil CO 2 efflux in winter ranged up to 5-30% of the total CO 2 efflux measured during the growing season [39,42].Oertel et al. [23] summarised studies conducted in the last two decades and reported that local climate conditions, land use and land cover, soil temperature and humidity, and soil fertility are the key drivers for GHG emissions from various types of soils.In general terms, the findings of the present study reveal that forest and grassland biotopes differ by SOC, TN concentrations, and C/N ratio in the topsoil of drained Terric Histosols.However, SOC, TN, and C/N ratios did not exhibit the expected relationship with total soil CO 2 efflux (Table 2).In addition, the mean total soil CO 2 efflux calculated over the entire vegetation period during the study did not show significant (p > 0.05) differences among the Norway spruce, black alder, and silver birch stands and perennial grassland (Figure 4).The obtained data only show that total soil CO 2 efflux was on average 30% higher in grassland than in forested Terric Histosols.
Oertel et al. [23] summarised studies conducted in the last two decades and reported that local climate conditions, land use and land cover, soil temperature and humidity, and soil fertility are the key drivers for GHG emissions from various types of soils.In general terms, the findings of the present study reveal that forest and grassland biotopes differ by SOC, TN concentrations, and C/N ratio in the topsoil of drained Terric Histosols.However, SOC, TN, and C/N ratios did not exhibit the expected relationship with total soil CO2 efflux (Table 2).In addition, the mean total soil CO2 efflux calculated over the entire vegetation period during the study did not show significant (p > 0.05) differences among the Norway spruce, black alder, and silver birch stands and perennial grassland (Figure 4).The obtained data only show that total soil CO2 efflux was on average 30% higher in grassland than in forested Terric Histosols.The higher total soil CO 2 efflux found in perennial grassland could be related to a huge alteration in the local mix of herbaceous species during the growing season (i.e., when one species dies, others grow).Other studies have reported that higher total soil CO 2 efflux in perennial grasslands may be related to the higher fine root biomass (to which autotrophic soil CO 2 efflux is attributed) in grasslands compared to forest ecosystems [43,44].
Furthermore, the Norway spruce, black alder, and silver birch forest stands differed from each other in terms of both dendrometric parameters (Supplementary Table S1) and the species composition of their understory vegetation cover (see Section 4.1).Such differences could have an impact on the amount of solar radiation entering the studied forest stands and their understory vegetation, and could, therefore, accelerate or reduce autotrophic total soil CO 2 efflux.
The soil WTL fluctuated across the entire study period (Table 1), which could be related to the relative depth of the drainage systems (using channelized forest streams) present in the studied forest and perennial grassland sites (see Section 4.1).Also, precipitation distribution over time and rain volumes could have an impact on the soil water table level.It is important to note that WTL was at least 40 cm below the soil surface at all studied sites (Table 1).The topsoil of the studied drained Terric Histosols had good aeration conditions.This indicates that neither the relative depth of drainage systems nor fluctuations in the WTL had significant effects on total CO 2 efflux (Figure 3; Table 1).
In addition, since total soil CO 2 efflux is the sum of several processes taking place in the soil, including heterotrophic respiration from the decomposition of organic matter, autotrophic respiration by live roots, and possibly oxidation of methane produced in deep anoxic soil layers, it is challenging to interpret its variation when detailed results on the actual constituent processes are not available.On the other hand, the results obtained by this study confirm the findings of previous studies [45][46][47] and indicate that fluctuations in soil and air temperatures are the most important factors determining the seasonal (spring, summer, and autumn) variations of total soil CO 2 efflux.Several studies have reported that higher soil temperature has a positive influence on the activity of soil microbiota and promotes CO 2 production by both autotrophic and heterotrophic soil respiration [48,49].

Study Sites
This study was performed in the southern part of Lithuania (Figure 5).The study area belongs to the hemi-boreal forest zone, with an annual average temperature of 6.5 • C and mean annual rainfall of 650 mm [50].The study sites (four in total) were located on drained nutrient-rich organic soils classified as Sapric [10] or Terric [9] Histosols, in Norway spruce (Picea abies (L.) H. Karst.), black alder (Alnus glutinosa (L.) Gaertn.), and silver birch (Betula pendula Roth) stands, and in a perennial grassland (Figure 5, Supplementary Table S1).Forests containing the three studied tree species, along with perennial grassland, represent 60% of all drained Terric Histosols found in Lithuania [8,53].All four study sites were drained using channelized forest streams (existing forest streams that were straightened, widened, and deepened) for the first time in 1958.The peat layer was 50 cm thick in the perennial grassland site and >100 cm thick in all three studied forest stands.
The first study site was in a pure 70-year-old Norway spruce stand.This site was drained for forestry purposes using channelized forest streams of 1.6 m depth.Compared with the other studied forest stands, the Norway spruce stand was the oldest and had the highest total stem volume and height (Supplementary Table S1).The ground vegetation was dominated by moss red-stemmed feathermoss (Pleurozium schreberi (Brid.)Mitt.) and dwarf shrub blueberry (Vaccinium myrtillus L.).
The second study site was in a 43-year-old silver birch forest stand with black alder admixture.This site was drained using a channelized forest stream of 1.2 m depth.The ditch was cleaned by removing a layer of organic sediment in 2015.Nevertheless, the stud- Forests containing the three studied tree species, along with perennial grassland, represent 60% of all drained Terric Histosols found in Lithuania [8,53].All four study sites were drained using channelized forest streams (existing forest streams that were straightened, widened, and deepened) for the first time in 1958.The peat layer was 50 cm thick in the perennial grassland site and >100 cm thick in all three studied forest stands.
The first study site was in a pure 70-year-old Norway spruce stand.This site was drained for forestry purposes using channelized forest streams of 1.6 m depth.Compared with the other studied forest stands, the Norway spruce stand was the oldest and had the highest total stem volume and height (Supplementary Table S1).The ground vegetation was dominated by moss red-stemmed feathermoss (Pleurozium schreberi (Brid.)Mitt.) and dwarf shrub blueberry (Vaccinium myrtillus L.).
The second study site was in a 43-year-old silver birch forest stand with black alder admixture.This site was drained using a channelized forest stream of 1.2 m depth.The ditch was cleaned by removing a layer of organic sediment in 2015.Nevertheless, the studied stand had the lowest stem volume of the three forest sites (Supplementary Table S1).The ground vegetation was dominated by herbs such as wood-sorrel (Oxalis acetosella L.), northern beech fern (Phegopteris connectilis (Michx.)Watt), and may lily (Maianthemum bifolium (L.) F.W. Schmidt).
The third study site was in a pure 30-year-old planted black alder stand.This stand was drained using channelized forest streams of 0.6 m depth.The black alder stand had the lowest DBH and, conversely, the highest tree density (Supplementary Table S1).The ground vegetation was dominated by raspberry (Rubus idaeus L.) shrubs.

Measurements of Total Soil CO 2 Efflux and Environmental Factors
Measurements were carried out once per month from the beginning of May to the end of October 2021 and from April to June 2022.Three subplots for total soil CO 2 efflux measurements were established in each study site.The distance between subplots was 30 m.Total soil CO 2 efflux was measured in each subplot, and with at least three replicates, using a closed chamber system that includes a portable CO 2 analyzer (EGM-4, PP systems, Amesbury, MA, USA) with a chamber (diameter 32 cm; height 18 cm).The chamber was installed on the sample plot directly over a 2 cm deep groove in the ground.The CO 2 concentration was measured for a 120 s period.Total soil CO 2 was measured only during the daytime when there was no precipitation, from 11 am to 2 pm.No data were collected during the cold period due to snow or ice cover on the soil.It was also impossible to measure the total soil CO 2 efflux in the black alder stand in April and May 2022 because the study area was destroyed by wild boars in spring, and soil CO 2 efflux values in disturbed peat can be up to three times higher than in undisturbed peat [54].
In total, 324 measurements were performed during the entire study period.CO 2 efflux (g CO 2 m −2 h −1 ) was calculated from the linear change of CO 2 concentration (ppm) in time as a function of chamber volume, air temperature, and air pressure [55].Also, the rate was calculated from the linear change (R 2 ≥ 0.95) in CO 2 concentration during the measurement period.
Alongside the CO 2 efflux measurements, soil and air temperature, soil moisture, and soil water table level were simultaneously measured in each subplot of each study site (Table 1).Air temperature was measured in the shade using a thermometer (Comet system, Rožnov pod Radhoštěm, Czech Republic), soil temperature (at a depth of 0-10 cm) was measured using soil temperature probe (Comet system, Czech Republic), soil water table level measured using a water level probe (Odyssey ® , Christchurch, New Zealand), and soil moisture was determined (at a depth of 0-10 cm) using a soil moisture meter (HH2, Delta-T Devices, Burwell, UK).

Soil Sampling and Chemical Analysis
Soil sampling was conducted in October 2021 after the growing season.Composite soil samples were taken from three systematically distributed points 28 m apart from each other.Three square-shaped 40 cm × 40 cm pits were dug at each sampling point.Samples of topsoil (0-10 cm depth) were taken from all sides of the pit in three replicates using a 100 cm 3 cylinder.In total, 108 topsoil samples were taken for chemical analysis at the Agrochemical Research Laboratory of the Lithuanian Research Centre for Agriculture and Forestry.Soil organic carbon (SOC) [56] and total nitrogen (TN) [57] were determined by dry combustion at 900 • C with a CNS analyzer (Elementar Analy-sensysteme GmbH, Langenselbold, Germany).

Statistical Analyses
The normality of the data was checked with Kolmogorov-Smirnov and Shapiro-Wilk tests, and all the data were found to be normally distributed.The impact of different forms of land use and different months, both treated as categorical variables, on soil CO 2 efflux was determined using univariate analysis of variance with Tukey's HSD test with significance level α = 0.05.Univariate analysis of variance with Tukey's HSD test was also used to compare SOC, TN, and the C/N ratio between the study sites.Pearson correlation analysis was used as a measure of the strength of the linear relation between CO 2 efflux and abiotic factors such as soil moisture, soil water table level (WTL), soil temperature at 10 cm depth (T soil 10 cm), and air temperature (T air).
Principal component analysis (PCA) was used to determine whether there were any significant interactions between total soil CO 2 efflux and several environmental factors, including soil physical parameters (soil moisture and soil temperature at 10 cm depth), soil chemical composition (SOC, TN, and C/N ratio at 10 cm depth), air temperature, WTL, and the depth of drainage ditches.PCA was performed using Varimax rotation.All statistical analyses were conducted using IBM SPSS 22.0 (IBM, New York, NY, USA).

Conclusions
This study was performed in drained Terric Histosols under Norway spruce, silver birch, and black alder stands and perennial grassland with the aim of measuring total soil CO 2 efflux.The nutrient-rich organic soils under the Norway spruce and silver birch stands were distinguished by significantly higher C/N ratios and concentrations of SOC and TN, indicating lower peat decomposition rates.In contrast, the opposite results found in the soils under the black alder stand and perennial grassland were interpreted to indicate a high rate of peat decomposition and nitrogen release during the peat decay process.However, the hypothesis that SOC, TN, and C/N ratio are the key factors influencing total soil CO 2 efflux from drained Terric Histosols under forest and grassland biotopes was not supported.The results indicate a trend that total CO 2 efflux was 30% higher in drained Terric Histosols under perennial grassland than it was under forest stands.However, air and topsoil temperatures were the most important of all studied environmental parameters (soil chemistry, land use, WTL, and soil moisture), explaining the variation of the total CO 2 efflux from drained Terric Histosols.In order to form national emission factors, it may be necessary to carry out flux monitoring throughout the year (also in winter).Based on soil total CO 2 efflux measurements, from a climate change mitigation perspective, forest stands are a better land-use option for reducing CO 2 emissions than perennial grasslands.However, quantifying the full net CO 2 balance of these soils requires further measurements.

Figure 2 .
Figure 2. The distribution of the mean monthly total soil CO2 efflux in perennial grassland and Norway spruce, silver birch, and black alder stands during the vegetative period (2021-2022).The data represent mean ± SE (n = 324).The different lower-case letters indicate statistically significant differences between the sites, p ˂ 0.05 (Tukey's HSD test).

Figure 2 .
Figure 2.The distribution of the mean monthly total soil CO 2 efflux in perennial grassland and Norway spruce, silver birch, and black alder stands during the vegetative period (2021-2022).The data represent mean ± SE (n = 324).The different lower-case letters indicate statistically significant differences between the sites, p < 0.05 (Tukey's HSD test).

Figure 3 .
Figure3.PCA of soil chemical properties (SOC, TN, and the C/N ratio in the topsoil), water table level, soil and air temperature (in the topsoil), drainage (the depth of drainage ditches), and total soil CO2 efflux.The x-axis represents principal component 1, and the y-axis component 2. Abbreviations: CO2_efflux-soil total CO2 efflux; Tair-air temperature; Tsoil-soil temperature at 10 cm depth; WTL-water table level; Moisture-soil moisture at 5 cm depth, SOC-soil organic carbon determined at 0-10 cm soil depth soil; TN-total nitrogen determined at 0-10 cm soil depth; CN_ratio-SOC/TN ratio determined at 0-10 cm soil depth; Drainage-depth of the drainage ditches.

Figure 3 .
Figure3.PCA of soil chemical properties (SOC, TN, and the C/N ratio in the topsoil), water table level, soil and air temperature (in the topsoil), drainage (the depth of drainage ditches), and total soil CO 2 efflux.The x-axis represents principal component 1, and the y-axis component 2. Abbreviations: CO 2 _efflux-soil total CO 2 efflux; Tair-air temperature; Tsoil-soil temperature at 10 cm depth; WTL-water table level; Moisture-soil moisture at 5 cm depth, SOC-soil organic carbon determined at 0-10 cm soil depth soil; TN-total nitrogen determined at 0-10 cm soil depth; CN_ratio-SOC/TN ratio determined at 0-10 cm soil depth; Drainage-depth of the drainage ditches.

Figure 4 .Figure 4 .
Figure 4. Distribution of mean total soil CO2 efflux over the entire study period (from May 2021 to June 2022) among the three studied forest sites and the perennial grassland site.The data represent mean ± SE.

Table 1 .
Seasonal variation of air and soil temperatures, soil moisture, and soil water table level (WTL) in the studied Norway spruce, silver birch, and black alder stands and perennial grassland.Notes: The data represent mean ± SE.Soil temperature and soil moisture were measured in the topsoil (10 cm depth).The data of environmental variables was collected in: a May 2021; b June, July, and August 2021; c September and October 2021; d April and May 2022; e June 2022.

Table 2 .
Pearson's correlation matrix showing relationships between environmental variables and total soil CO 2 efflux in drained Terric Histosols in Norway spruce, silver birch, and black alder forest stands and in perennial grassland (n = 153).

Table 3 .
Total soil CO 2 efflux released from drained Histosols in boreal and hemi-boreal regions.