The influence of the hydrometeorological factors on the CO2 fluxes from the oligotrophic bog surface.

Global climate change is one of the most important and promising phenomena to study in actual time. One of the key causes of global climate change is increasing the greenhouse gas (GHG) concentrations in the atmosphere [IPCC, 2023]. The main greenhouse gases are methane, carbon dioxides and nitric oxide, which contribute to the greenhouse effect and global warming [Lashof, Ahuja, 1990]. Carbon dioxide (CO₂) is one of the most significant and widespread gases involved in the planet's global carbon cycle [Lashof, Ahuja. 1990]. At the same time, living organisms play a key role in creation of atmosphere composition. Autotrophic organisms use a carbon dioxide to build their body structures, including complex organic compounds. During ecosystem functioning, the part of the carbon dioxide is released into the atmosphere through organism respiration, while another part is released through the decomposition of dead organic matter. Carbon dioxide may also be produced through natural and anthropogenic processes.

Peatland ecosystems play a significant role in the planet's carbon cycle, both locally and globally. Peatlands in their natural undisturbed state are a significant long-term carbon sink¹. However, the process of carbon deposition is not constant – in different years, peatlands may serve either as carbon sink or source². The main factor stimulating the carbon sequestration by peatland ecosystems is climatic conditions [Harenda et al., 2018; Bond-Lamberty et al., 2018]. Peatlands are the second most significant carbon stock on Earth and the largest on land. Despite covering only 2.84% of the Earth's land surface, the amount of soil organic carbon stored in them accounts for about one-third of all soil organic carbon on Earth. Peatlands in the northern hemisphere play a particularly important role in carbon sequestration, with an estimated accumulated carbon quantity of ~473–621 Gt of carbon [Yu et al., 2010].

The largest area of peatlands in Russia is located in Western Siberia, estimated at ~42% of the total Russian area [Vomperskiy et al., 1994; Sheng et al., 2004]. The territory of Western Siberia is featured to a high share of peatlands in original undisturbed state, making them an ideal location to study the impact of global changes on peatland biogeochemical functioning worldwide.

The carbon balance of peatlands is mainly determined by two processes: photosynthesis and respiration [Harenda et al., 2018]. The main factors influencing the CO₂ flux from peatlands are photosynthetically active radiation, atmospheric air temperature (T_avg), soil temperature (T_soil), and water table level (WTL) [Miao et al., 2013; Juszczak et al., 2013; Dyukarev et al., 2019]. At the same time, the level of mutual influence and the degree of determination have not yet been fully determined.

To study the carbon balance of terrestrial ecosystems, the chamber method [Davidson et al., 2002] is widely used. The chamber method allows to estimate the CO₂ flux from the surface of the ecosystem. At the same time, the use of the modern automatic system LI-COR LI-8100A (LI-COR, USA) provides high-frequency continuous data on carbon dioxide fluxes over a long period of time, which makes it possible to assess the total accumulation of carbon and significantly improve the reliability of the identified relationships with environmental factors [Zarov et al., 2022].

The purpose of this study was to assess carbon dioxide flluxes and discover the main hydrometeorological parameters that influence the flow in the hollows of the Mukhrino raised bog.

MATERIALS AND METHODS

The research was carried out at the «Mukhrino» field station [Dyukarev et al., 2021], located in the central part of Western Siberia, 30 km southwest of the city of Khanty-Mansiysk. The climate is featured by high repeatability of anticyclonic conditions, rapid changes in weather conditions, a humid, moderately warm summer, and a fairly harsh, snowy winter. The chamber system was installed in a homogeneous area of the peatland, dominated by Sph. balticum, C. limosa, and Scheuchzeria palustris, with the presence of E. vaginatum on the periphery. The plant composition inside the chambers was not determined, but the most homogeneous and similar areas were selected for installation (Figure 2).

Carbon dioxide flux measurements were carried out using the automated chamber method, using a portable soil respiration analysis system LI-8100A (LI-COR, USA). The flues were measured by four automated chambers installed in the raised bog area of Mukhrino (Figure 3). The first group of chambers – NEE (2 LI-COR 8100-104s chambers), measured net ecosystem exchange (NEE); the second group – R_eco (2 LI-COR 8100-104 cameras), measured ecosystem respiration (R_eco). Measurements were taken for 2 minutes every 30 minutes for all cameras. Wooden walkways were installed in the peatland area to minimize potential negative impacts on the study surface.

The fluxes were calculated using a linear model of specialized software LI-8100 File Viewer 3.0.0 (LI-COR). R programming language packages dplyr [Wickham, 2016], ggplot2 [Wickham, 2016], lubridate [Grolemund, Wickham, 2011] were used for data processing and visualization. To analyze the dynamics of NEE and R_eco fluxes, the obtained values were averaged between LI-COR 8100-104s chambers (for NEE) and LI-COR 8100-104 chambers (for R_eco). Gross primary production (GPP) was calculated using the equation GPP=NEE-R_eco [Connolly et al., 2009]. For further analysis, measurements with a coefficient of determination (R²) of linear regression above 0.5 were selected to minimize significant noise in the data. Spearman's rank correlation method was chosen to identify dependencies of flux on hydrometeorological properties. The dependence was determined based on the data of the flux and hydrometeorological properties averaged over 30 minutes.

RESULTS AND DISCUSSION

The average daily variation of CO₂ flows for July, September, October 2021 is shown in Fig. 5. The simultaneous use of dark and light chambers allowed to assess the flows that are released in the ecosystem as a result of the respiration of plants, animals and microorganisms (R_eco), the intensity of CO₂ absorption in the process photosynthesis (GPP), and net ecosystem exchange (NEE), which is the difference between the specific absorption rate (GPP) of carbon dioxide excretion (R_eco). The average daily variation of R_eco (Fig. 5) in July was featured by the highest values during daylight hours; the CO₂ flux reaches its maximum value at 11:00 (1.44 µmol m^‑2s^‑1). For September and October, the daily dynamics of R_eco were weakly expressed. The highest CO₂ emissions were typical for evening and night time. The maximum R_eco in the daily cycle was observed at 19:00 (0.47 µmol m^‑2s^‑1) for September, and at 00:00 (0.17 µmol m^‑2s^‑1) for October. The average daily cycle of GPP (Fig. 5) had a pronounced absorption maximum during daylight hours with maximum radiation, for July – at 11:00 (-3.47 µmol m^‑2s^‑1), for September – at 12:00 (-1.53 µmol m^‑2s^‑1), for October – at 11:00 (-0.45 µmol m^‑2s^‑1). The absorption of carbon dioxide from the atmosphere (GPP) had different daily durations depending on the month (Fig. 5), which is associated with a decrease in daylight hours by autumn. In July, carbon dioxide absorption was observed from 4:00 to 20:00 (16 hours), in September – from 5:00 to 18:00 (13 hours), in October – from 7:00 to 17:00 (10 hours). For the diurnal cycle of NEE (Fig. 5), the CO₂ absorption process (GPP>R_eco) predominated in the daytime, while the carbon dioxide emission process (GPPeco) dominated at night. The maximum NEE value in the daily cycle in July was estimated at 1.01 µmol m^‑2s^‑1 at 22:00, in September 0.49 µmol m^‑2s^‑1 at 20:00, in October 0.17 µmol m^‑2s^‑1 at 21:00. The minimum NEE value in July was -2.03 µmol m^‑2s^‑1 at 11:00, in September: -1.01 µmol m^‑2s^‑1 at 12:00, in October 0.39 µmol m^‑2s^‑1 at 11:00.

A total of 1711, 2625 and 1597 R_eco measurements were taken in July, September and October, respectively. The highest average daily rate of ecosystem respiration R_eco occurred in the third ten days of July (July 19); by the last days of October, ecosystem respiration reached its minimum in the annual course (Fig. 7). The average R_eco in July was 1.05±0.25 µmol m^‑2s^‑1, and in October 0.13±0.01 µmol m^‑2s^‑1. These estimates were obtained on a sufficient array of data and therefore can be considered reliable. The peak intensity of photosynthesis was recorded on July 22, when vegetation absorbed the largest amount of CO₂. After July 22, there was a gradual decline in GPP; the rate of carbon dioxide absorption in the last days of October decreased significantly, but did not drop to zero. The presence of photosynthesis in the hollow of an oligotrophic bog even in late autumn and at low air temperatures is probably due to the activity of sphagnum mosses. Net ecosystem exchange (Fig. 7) was negative every day in July, thereby the absorption of carbon dioxide from the atmosphere daily dominated its release. In September, ecosystem absorption of carbon dioxide prevailed until September 10, after which both negative and positive NEE values were observed. During this period, intense precipitation occurred, a decrease in air temperature and the amount of incoming radiation, which led to the ecosystem switching from a sink to a temporary source of CO₂. In October, the number of days on which the ecosystem acted as a carbon sink decreased; on most days, carbon dioxide emissions predominated. According to average monthly values, carbon dioxide absorption prevailed in July (-0.53±0.13 µmol m^‑2s^‑1) and September (-0.11±0.18 µmol m^‑2s^‑1), in October (0.02±0.04 µmol m^‑2s^‑1) CO₂ evolution predominated. The number of measurements according to NEE (Table 2) is greatest in September (2584) and least in July (1709).

R_eco was most influenced in July (Table 3) by air and soil temperature; in September – soil temperature and marsh water level. In October, when daily temperature variability decreased, the most significant factor for R_eco was PAR (-0.59). The degree of correlation of R_eco with T_avg and T_soil in July qualifies as high; these factors are directly related to R_eco – the higher the temperature, the greater the release of carbon dioxide into the atmosphere by the ecosystem. This is caused by an increase in the activity of microorganisms under the influence of increased temperature [Nikonova et al., 2019]. In September, the influence of T_soil (0.81) and water level (-0.78) increased, while the influence of T_avg (0.54) decreased. The degree of correlation of these parameters with R_eco in September was classified as high. It is assumed that the strong influence of water level (-0.78) on the R_eco flux in September may be associated with a sharp rise in water level (Fig. 6F), which could lead to a disruption of the optimum life activity of microorganisms. Similar flow behavior was found for North American peatlands [Miao et al., 2013]. In October, the greatest influence on R_eco was exerted by PAR (-0.59), the degree of correlation is weak negative; At the same time, the correlation of the indicator with PAR in July was weakly positive. The highest correlation for GPP (Table 3) was obtained with photosynthetically active radiation for all months of the study. The PAR correlation level for all months was classified as high. The inverse correlation is due to the fact that as PAR increases, CO₂ absorption increases (negative GPP flux). PAR is a key factor influencing plant photosynthesis, which in turn affects their ability to assimilate CO₂ and produce GPP. As PAR intensity increases, plants increase the rate of photosynthesis and absorb carbon dioxide from the atmosphere faster, which increases GPP. The greatest influence on NEE was caused by PAR (Table 3) in July, in September and October (-0.91, -0.74 and -0.71, respectively). The level of PAR correlation in July and September was high, in October it was moderate. When PAR levels increase, plants use carbon dioxide more actively to produce organic matter and increase the level of GPP in the ecosystem, which leads to an increase in NEE flux. On the other hand, when PAR levels decrease, plants become less active in photosynthesis, which leads to the prevalence of R_eco and a decrease in NEE flux. Analysis of correlation coefficients calculated from data for the entire field season, the best relationship for R_eco was found with soil temperature (0.88), air temperature (0.71) and water level (-0.73). PAR has the greatest influence on GPP (-0.89) and NEE (-0.73).

CONCLUSIONS

Automated high temporal resolution chamber measurements of carbon dioxide flux provided a data for analyzing CO₂ fluxes in the peatland area. The results provided detailed information that was used to analyze the impact of environmental hydrometeorological parameters on the flux. The highest ecosystem respiration (R_eco) value during a 24-hour period was recorded in July at 11:00 (1.44 µmol m^‑2s^‑1), in September at 19:00 (0.47 µmol m^‑2s^‑1), and in October at 00:00 (0.17 µmol m^‑2s^‑1). The maximum gross primary production (GPP) for all months occurred between 11-12 hours: in July at 11:00 (-3.47 µmol m^‑2s^‑1), in September at 12:00 (-1.53 µmol m^‑2s^‑1), and in October at 11:00 (-0.45 µmol m^‑2s^‑1). By autumn, the duration of GPP throughout a day decreased, as well as the amplitude of diurnal variation for all flux indicators. The highest average daily CO₂ flux for all indicators was recorded in July, while the lowest was in October. In net ecosystem exchange (NEE), absorption predominated from July 14 to September 9, with days dominated by ecosystem respiration from September 10 onwards. The amplitude of the average daily flux for all indicators decreased by October.

Based on the Spearman correlation data, the highest seasonal correlation for ecosystem respiration (R_eco) was with soil temperature (0.88), air temperature (0.71), and water level (-0.73). In July, the best correlation is with air temperature (0.70) and soil temperature (0.68), in September with soil temperature (0.81) and water level (-0.78), and in October with photosynthetically active radiation (PAR) (-0.59). Gross primary production (GPP) correlates best with PAR. In July, the correlation coefficient is -0.95, in September -0.86, in October -0.79, and for the entire field season -0.89. Net ecosystem exchange (NEE), similar to GPP, is most dependent on PAR. In July, the correlation coefficient is -0.91, in September -0.74, in October -0.71, and for the entire field season -0.73.

In general, the article calculates carbon dioxide fluxes from the surface of a hollow in an oligotrophic peatland. The seasonal and average daily dynamics of hydrometeorological properties are described, and their influence on CO₂ flows is assessed. It is worth noting that throughout the entire growing season, the influence of external factors on fluxes decreases, reaching a minimum mutual correlation in the coldest month (October).