Introduction
Soil respiration (SR), or the emission of CO2 from soil, is one of the most important fluxes in the exchange of carbon between ecosystems and the atmosphere. It consists mainly of autotrophic root respiration and heterotrophic microbial respiration and is also a key indicator of soil health and fertility (Ward et al., Reference Ward, Kirkman, Hagenah and Tsvuura2017). Even after two decades of research on this topic and undoubted successes, many uncertainties remain in this field (Bond-Lamberty et al., Reference Bond-Lamberty, Ballantyne, Berryman, Fluet-Chouinard, Jian, Morris, Rey and Vargas2024).
In agroecosystems, the respiration of arable soils is influenced by both natural and anthropogenic factors. Important factors affecting SR include soil and air temperature, soil moisture, the quantity and quality of organic compounds, the population dynamics of different microbial groups, the soil’s physicochemical properties, including oxygen supply, acidity and redox potential (Rochette and Hutchinson, Reference Rochette, Hutchinson, Hatfield and Baker2005), ammonium and ammonia nitrogen content (Sosulski et al., Reference Sosulski, Szymańska, Szara and Sulewski2021), the depth of soil horizons, slope steepness and the silt fraction (Adhikari et al., Reference Adhikari, Anderson, Smith, Owens, Moore and Libohova2023). Among the anthropogenic factors, the application of fertilizer is particularly notable, as it increases bacterial community diversity in the soil and consequently increases SR (Wang et al., Reference Wang, Xie, Li, Effah, Xie, Luo, Zhou and Jiang2022). However, this temporary effect may last no more than two weeks, after which the respiration level decreases to initial values (Kulachkova et al., Reference Kulachkova, Derevenets, Korolev and Pronina2023). Intra-seasonal CO2 emission dynamics from arable soils also depend on the plant growth stage, row width and tillage. This is because ploughing, which involves mixing of soil layers, consequently contributes to the increased decomposition of plant residues and emissions (Zapata et al., Reference Zapata, Rajan, Mowrer, Casey, Schnell and Hons2021).
Among the key factors influencing SR, researchers cite the content of organic carbon in the upper soil layer, which in some cases accounts for more than 90% of the variance in respiration (Li et al., Reference Li, Wang, Li, Wang, Liu, Shi and Cao2019). Soil organic matter can be divided into two fractions. The first is particulate organic matter (POM), which is distributed in the sand fraction (2–0.05 mm); it is considered relatively more labile and enriched in organic carbon than the soil as a whole (Semenov et al., Reference Semenov, Lebedeva, Sokolov, Zinyakova, Lopes de Gerenu and Semenov2023). The second is mineral-associated organic matter (MAOM), contained in silt and clay (< 0.05 mm), which is relatively stable and has a lower carbon concentration than the soil (Semenov et al., Reference Semenov, Lebedeva, Sokolov, Zinyakova, Lopes de Gerenu and Semenov2023). In arable soils, the impoverishment of the active carbon pool further increases the proportion of MAOM relative to POM, which is typically 60:40. However, as POM contains 2.5–6.4 times more potential mineralizable carbon, it dominates the formation of soil CO2 emissions (Semenov et al., Reference Semenov, Lebedeva, Sokolov, Zinyakova, Lopes de Gerenu and Semenov2023).
Considering the multitude of listed controls and the complexity of their interactions, it is clear that studying their influence on SR under agricultural production conditions is only possible in long-term field experiments. Based on this, our study hypothesized that within the framework of a field experiment, it would be possible to separate the influence of the two key anthropogenic factors – applied fertilizers and cultivated crops – on the respiration rate of arable soils.
Initially, long-term field experiments were designed to evaluate the impact of natural factors (e.g., weather conditions, soil characteristics, water regime) and anthropogenic factors (e.g., fertilizers, melioration, crop rotation, variety selection, mechanization) on crop yields and the conservation of soil fertility within a confined area (Ward et al., Reference Ward, Kirkman, Hagenah and Tsvuura2017). More recently, experiments have studied the influence of factors such as crop rotation and permanent crops (Zavyalova et al., Reference Zavyalova, Vasbieva and Fomin2020), the ploughing and no-till systems (Gelybó et al., Reference Gelybó, Barcza, Dencső, Potyó, Kása, Horel, Pokovai, Birkás, Kern, Hollós and Tóth2022). To date, several dozen long-term field experiments lasting over 20 years have been conducted, and their data have been systematized by organizations such as the American Society of Agronomy (https://www.agronomy.org/membership/sections-communities), the Global Long-Term Agricultural Experiment Network (https://glten.org/) and the BonaRes project (https://www.bonares.de/service-portal/data-repository#data%20reuse).
Field experiments provide an opportunity not only to assess the impact of farming practices on crop yields and investigate soil organic carbon content and composition but also allow the development of measures to adapt crop production to mitigate climate change and obtain data to validate simulation models (Grosse et al., Reference Grosse, Hierold, Ahlborn, Piepho and Helming2020). Long-term experiments enable field samples to be collected for the assessment of plant and soil characteristics, particularly properties that change slowly and affect fertility. Archival materials are an invaluable source of information for future research and for developing models that describe soil–plant system processes (Johnston and Poulton, Reference Johnston and Poulton2018; Grosse et al., Reference Grosse, Hierold, Ahlborn, Piepho and Helming2020). In our opinion, although long-term experiments are expensive and require years of logistical support, they are the most cost-effective research method. The scientific knowledge gained leads to increases in yield and product quality, reduces the negative environmental impact of agriculture and maintains soil quality and natural resource conservation (Körschens, Reference Körschens2006).
Based on this, the objective of the present study was to determine the influence of the species of cultivated crop and the type of fertilizer applied on arable SR and to assess the dependence of SR on soil characteristics. Similar studies investigating the impact of these two factors on SR have been conducted in the world’s oldest field experiment at Rothamsted Research in England (Storkey et al., Reference Storkey, Macdonald, Bell, Clark, Gregory, Hawkins, Hirsch, Todman and Whitmore2016), as well as in long-term field experiments in the Czech Republic (Cerhanová et al., Reference Cerhanová, Kubát and Nováková2006) and Poland (Sosulski et al., Reference Sosulski, Szymańska, Szara and Sulewski2021). Similar measurements were also taken in the long-term field experiment at the Timiryazev Academy (Moscow, Russia), but these were only taken in the summer using a different method (manometric) to estimate SR (Savoskina and Polin, Reference Savoskina and Polin2015).
Materials and methods
Site description
The object of the study was the long-term field experiment of the Timiryazev Agricultural Academy located in the north of the Moscow area (55°50′25″ N, 37°33′25″ E, 157 m a.s.l.). Established in 1912 by Professor A.G. Doyarenko, the experiment aimed to evaluate the impact of different fertilizers, liming practices and crop rotation on crop yields (Mazirov and Safonov, Reference Mazirov and Safonov2010). The experimental area is 1.5 ha, and the accounting plot area is 50 m2. The experiment is divided into two parts: monocropping and crop rotation. The crop rotation scheme is as follows: fallow–winter rye–potato–barley–clover–flax.
Fertilizer variants (nine variants in crop rotation and 11 variants in monocropping) were applied across the fields: N, P, K, without fertilizer, NP, NK, PK, NPK + manure, NPK and two additional variants on monocropping – manure and without fertilizer. Since 1973, full mineral fertilizer NPK has been applied to all even-numbered crop rotation plots (taking into account the aftereffects of previously applied fertilizer variants), and only odd-numbered plots have retained variants of differential fertilization. The fertilizer doses have also changed several times over the last 100 + years and are currently 100 kg N/ha in the form of ammonium nitrate, 150 kg P/ha in the form of double superphosphate, 120 kg K/ha in the form of potassium chloride and 20 t/ha of farm yard manure. Soil liming was performed once per rotation on the longitudinal half of each plot according to the value of hydrolytic acidity (4.5 t/ha), with dolomite limestone being applied (Mazirov and Safonov, Reference Mazirov and Safonov2010). The last liming was carried out in autumn 2020.
The soils at this site have been cultivated for more than 200 years. Despite the small size of the experiment, the soil cover is quite diverse and comprises three soil types (Khitrov, Reference Khitrov2012): Stagnic Cutanic Albeluvisol (Siltic, Eutric, Ruptic), Stagnic Albic Podzol (Siltic, Eutric, Ruptic) and Haplic Regosol (Siltic, Eutric). The first type is predominantly found in monocropping, while the second and third types are found in crop rotation areas. The thickness of the anthropogenically transformed surface horizon, which has been subjected to mechanical treatment, varies from 25 to 55 cm (Khitrov, Reference Khitrov2012).
Field measurements
For the measurements, the rotation area closest to real farm conditions was chosen, as were the most significant crops for the central part of European Russia: rye, barley, potatoes and fallow. Three contrasting fertilizer variants were selected: full mineral fertilizer (variant N100P150K120, hereafter referred to as NPK); a combination of mineral and organic fertilizer (variant N100P150K120 + 20 t/ha manure, referred to as NPK + manure); and a variant without fertilizer. Thus, 12 plots were analysed each year, all with lime application (Figure 1).
Scheme of the long-term field experiment of the Timiryazev Academy with the location of the plots under study. Fertilizer options are shown on the right, and crop rotation is shown at the top. N, nitrogen; P, phosphorus; K, potassium.
In 2023, the potato and fallow plots were located on the even-numbered fields fertilized with NPK, whereas winter rye and barley were cultivated on the odd-numbered fields, with fertilizer applied according to nine different options. However, in 2024, the scheme was reversed: different fertilizer variants were applied to the plots with potatoes and fallow, whereas rye and barley were grown under the same application of NPK. It is important to note that the spring rye was reseeded in the winter rye plots due to poor overwintering.
Measurements were taken at bi-weekly intervals during the growing season and once a month during the bare soil period, in a five-fold repetition in each plot, from May 2023 till November 2024. Ploughing (autumn transition) and seeding (spring transition) were considered as seasonal boundaries. During the growing seasons (May–September), the averaging of SR values was used for crops and fertilizer options. During the bare soil period (October–April), averaging was performed for fertilizer options only as the soil was mixed between plots with different crops after ploughing. The total sample size was 108 averages for the 2023 growing season, 90 averages for the 2024 growing season and 30 averages for the bare soil period.
Carbon dioxide emission measurements from the soil were taken using the closed chamber technique and portable infrared CO2 analysers based on an AZ 77535 sensor (AZ Instruments, Taiwan), which were modified for fieldwork (patent 174321 U1). Five opaque, cylindrical, PVC chambers, each with an area of 90 cm2 and a height of 20 cm, were installed in each plot: between the rows for rye and barley, on top of the ridges for potatoes, and at randomly selected points on fallow land. Air and soil temperatures at the depths of 5 and 10 cm (HI 98509, Hanna Instruments, USA) and soil volumetric moisture (SM 150 Kit, Delta-T, UK) were estimated simultaneously.
In July 2023, mixed soil samples were collected from the 0–20 cm layer in each plot. After drying, sieving and treatment with hydrochloric acid, the organic carbon (SOC) and total nitrogen (Ntotal) contents were determined. The POM and MAOM fractions were then isolated from the same samples according to the method described by Semenov et al. (Reference Semenov, Lebedeva and Pautova2019). The carbon (C-POM and C-MAOM) and nitrogen (N-POM and N-MAOM) content of these fractions was determined using a Vario EL Cube CHNS analyser (Elementar, Germany) in the Laboratory of Radiocarbon Dating and Electronic Microscopy (Institute of Geography, Russian Academy of Sciences). Additionally, the pH of the soil samples was measured (HI 98121, Hanna Instruments, USA).
Statistical analysis
Statistical analysis was performed using the PAST software (Hammer et al., Reference Hammer, Harper and Ryan2001). Due to the small size of the samples and the differing group variances, nonparametric methods were used:
-
1. Two-way permutation multivariate analysis of variance (two-way PERMANOVA) was used to separate and assess the significance of the effect of two factors on SR: cultivated crop and fertilizer application for two growing seasons (P = 0.05). This method does not work with the values themselves, but with a distance matrix based on dissimilarity criteria (Anderson, Reference Anderson2001). In the current study, the Bray–Curtis criterion was chosen.
-
2. The Mann–Whitney test (P = 0.05) was used to compare the SR rates under different crops and fertilizer options.
-
3. Spearman’s correlation (P = 0.05) was used to assess the dependence of the SR rate on hydrothermal indicators (air and soil temperatures, volumetric soil moisture) and various pools of carbon and nitrogen (SOC and Ntotal, C-POM and C-MAOM, N-POM and N-MAOM).
Results
Soil respiration rate
A graphical analysis of SR in the 12 target plots (Figures 2 and 3) shows that they are all characterized by significant variation. Nevertheless, the ranges overlap and the median levels are relatively close to each other (0.063–0.178 g C/(m2·h) in 2023 and 0.099–0.276 g C/(m2·h) in 2024), regardless of the fertilizer used; the mean values are most often greater than the medians (0.074–0.253 g C/(m2·h) in 2023 and 0.118–0.288 g C/(m2·h) in 2024). The only exception is the NPK + manure variant, where CO2 emission from the soil is significantly greater than that in the other plots, averaging 0.430 g C/(m2·h) under winter rye in 2023 and 0.371 g C/(m2·h) on fallow in 2024.
Soil respiration during the vegetation period of 2023. NPK, nitrogen, phosphorus, potassium.
Soil respiration during the vegetation period of 2024. NPK, nitrogen, phosphorus, potassium.
During both growing seasons, the low SR under potatoes is notable, which is caused by the ridging method used in its cultivation. As a result, the soil temperature increases significantly in the summer period, and its moisture drops sharply. Active SR under cereal crops (rye and barley) is due to great plant biomass, which preserves moisture and prevents a significant increase of soil temperature in shaded areas. SR on fallow is also quite intensive, which can be attributed to constant overgrowth of weed vegetation and subsequent cultivation, thereby activating soil microbiota.
The bare soil period can be divided into subperiods: winter, with negative air temperatures and snow cover, and autumn and spring periods, which are characterized by positive temperatures and no snow cover. SR values vary considerably in autumn and spring (Figure 4), whereas under snow crust, SR drops sharply to 0.006–0.018 g C/(m2·h), which is almost 10–20 times lower than during the growing season.
Soil respiration from autumn 2023 to spring 2024. NPK, nitrogen, phosphorus, potassium.
Multivariate analysis of variance
The results of the two-way PERMANOVA analysis (Table 1) show that, in plots receiving continuous NPK application, SR is independent of the cultivated crop species or type of fertilizer. However, for plots divided by fertilizer variants, both crop and fertilizer significantly affect soil CO2 efflux, although their interaction is insignificant. Using the sum-of-squares values, the relative contribution of the factors to SR variation can be calculated. Fertilizer application explains 17.2–24.0% of the total variance, whereas the crop species explains only 6.5–7.1%. The majority of the variance in SR is attributed to unaccounted factors (65.0–75.9%).
Results of two-way PERMANOVA
NPK, nitrogen, phosphorus, potassium.
Significant factors (P < 0.05) are written in bold.
Pairwise comparison using the Mann–Whitney test shows that SR during the growing season on most plots with different crops and fertilizers does not differ significantly: in 2023, out of 66 compared pairs, a significant difference was observed in only 22 pairs, that is, one-third (Table 2), and in 2024, only six pairs differed significantly, that is, one-tenth (Table 3). As expected, the highest number of differences was obtained for SR in the NPK + manure variant (winter rye in 2023 and fallow in 2024). Conversely, CO2 emission from the soil was almost completely coincident with that from all other plots for all crops under continuous NPK application. During the bare soil period, nonparametric comparison showed that SR did not differ significantly between fertilizer variants (P > 0.05).
Results of pairwise comparison of soil respiration in the study plots during the 2023 growing season
NPK, nitrogen, phosphorus, potassium.
The values of the significance level P for the Mann–Whitney criterion are presented. Significant differences (P < 0.05) are written in bold.
In brackets are fertilizer variants that were applied to these plots before 1973, the effects of which are taken into account.
Results of pairwise comparison of soil respiration in the study plots during the 2024 growing season
NPK, nitrogen, phosphorus, potassium.
The values of the significance level P for the Mann–Whitney criterion are presented. Significant differences (P < 0.05) are written in bold.
In brackets are fertilizer variants that were applied to these plots before 1973, the effects of which are taken into account.
Correlation analysis
Strong correlations between soil CO2 emission and the hydrothermal characteristics of the environment were obtained (Table 4). However, it is impossible to identify any general tendency in the influence of temperature and moisture on SR, as different crops and different fertilizer variants show significant differences. For plots under potatoes and fallow, where the proportion of bare soil is high, the temperature dependence of CO2 emissions from the soil is clearly traced. At the same time, however, very few significant correlations were obtained for spring cereal crops such as barley and rye.
Correlations of soil respiration with hydrothermal parameters during the growing seasons
NPK, nitrogen, phosphorus, potassium.
Spearman’s correlation coefficients R in the numerator and significance levels P in the denominator are presented.
‘−’ significant correlations were not obtained.
In brackets are fertilizer variants that were applied to these plots before 1973, the aftereffects of which are taken into account.
When applying mineral and especially organic fertilizers, the soil was supplied with carbon (up to 1.84%) and nitrogen (up to 0.14%) (Figure 5). These elements accumulated especially intensively (four times higher) in the active POM fraction (0.62% C and 0.04% N) compared to the variant without fertilizers (0.14% C and 0.01% N). Interestingly, organic fertilizers did not increase MAOM, and the accumulation of carbon and nitrogen under the NPK variant (0.88% C and 0.08% N) was significantly greater than that in NPK + manure, which, in terms of carbon and nitrogen content, was equal to that of the variant without fertilizers (0.54–0.58% C and 0.05% N).
Carbon (a) and nitrogen (b) content in different pools of soil organic matter (mean ± standard error). C, carbon; POM, particulate organic matter; MAOM, mineral-associated organic matter; SOC, soil organic carbon; N, nitrogen.
The average values of CO2 emissions during the vegetation period generally showed significant moderate correlations with the total content of carbon and nitrogen and the content of these elements in POM (the most active fraction of soil organic matter), as well (Table 5). However, MAOM, which is strongly associated with clay particles, apparently does not play a role in the formation of CO2 efflux from the soil, despite its higher carbon and nitrogen content (Semenov et al., Reference Semenov, Lebedeva, Sokolov, Zinyakova, Lopes de Gerenu and Semenov2023).
Correlation coefficients of growing season averages of soil respiration with the content of different pools of carbon and nitrogen
SOC, soil organic carbon; N, nitrogen; C, carbon; POM, particulate organic matter; MAOM, mineral-associated organic matter.
‘−’ significant correlations were not obtained.
Spearman’s correlation coefficients R in the numerator and significance levels P in the denominator are presented.
No relationship was found between SR and pH. This is because this indicator varied slightly during the experiment, remaining within the range of 6.01–6.68 (the soil solution was slightly acidic and close to neutral). Most probably, it is a consequence of liming (Sosulski et al., Reference Sosulski, Szymańska, Szara and Sulewski2021). During periods without crops in the fields, there is a strong positive correlation with air temperature for the NPK + manure variant (R = 0.762; P = 0.037). However, such a relationship was not detected in the NPK plots or in plots without fertilizer.
Discussion
The results of the current study have revealed a significant contribution of crops and fertilizers to soil CO2 emission, confirming results observed in prior studies by other researchers. Similar observations were made in a long-term field experiment in Poland, where the respiration of light soils was found to depend more on the species of cultivated crop and the type of fertilizer applied than on the crop rotation system (Sosulski et al., Reference Sosulski, Szymańska, Szara and Sulewski2021). For example, CO2 emission from the soil under winter rye in the NPK + manure variant with lime application in no-till crops did not differ from that in the crop rotation system (Sosulski et al., Reference Sosulski, Szymańska, Szara and Sulewski2021). However, according to the results of an experiment conducted in the Czech Republic, the crop exerts a stronger effect on SR than the fertilizer (Cerhanová et al., Reference Cerhanová, Kubát and Nováková2006).
Similar fertilizer variants were applied to experimental fields at the All-Russia Research Institute of Organic Fertilizers in the Vladimir region, which has the same type of sod-podzolic soil as in our experiment. SR naturally increased in the following order: without fertilizer, NPK, NPK + manure: during the growing season, this amounted to 1554, 2461 and 2792 kg C/ha for barley and 1902, 2471 and 2658 kg C/ha for potatoes, respectively (Shilova, Reference Shilova2014). In the Vladimir experiment, during the growing season, CO2 emission averaged over crops in the variant without fertilizer was 1296 kg C/ha on fallow, 1753 kg C/ha under potatoes, and 1959 kg C/ha under winter wheat (Shilova, Reference Shilova2014; Lukin, Reference Lukin2015). Analogous values calculated for the long-term experiment at the Timiryazev Academy amount to 2900, 2100 and 4800 kg C/ha, respectively; that is, they are 1.2–2.4 times higher. The most probable reason for this difference is the method used to measure emissions, since the Vladimir experiment used a less accurate method of CO2 absorption by alkali.
It is important to emphasize that SR depends on the type of fertilizer used. Indeed, several authors have noted that the use of organic fertilizers can significantly increase CO2 emissions from soil (Shilova, Reference Shilova2014; Lukin, Reference Lukin2015; Körschens, Reference Körschens, Mueller, Sychev, Dronin and Eulenstein2021; Wang et al., Reference Wang, Xie, Li, Effah, Xie, Luo, Zhou and Jiang2022). For example, Shilova (Reference Shilova2014) estimates that mineral fertilizers increase emissions by 20%, whereas the application of organic-mineral fertilizer increases them by 1.3–1.5 times. The absence of mineral fertilizers significantly suppresses SR (Sosulski et al., Reference Sosulski, Szymańska, Szara and Sulewski2021). In our study, respiration enhancement was 10–60% for the NPK variant and 2–3 times higher for the NPK + manure variant than for plots without fertilizers. In addition to enhancing emissions, organic fertilizers increase net carbon uptake by soil (Yang et al., Reference Yang, Xiao and Xu2018), thereby significantly rising its carbon and nitrogen content (Wang et al., Reference Wang, Li and Li2023). In accordance with the results obtained in our work, the content of SOC and Ntotal in the NPK + manure variants showed a two-fold increase relative to the control and a four-fold increase in C-POM and N-POM.
Numerous long-term European experiments have shown that the SOC content varies from 0.15% in sandy soils to 2.29% in chernozems (usually less than 1.0%) (Körschens, Reference Körschens, Mueller, Sychev, Dronin and Eulenstein2021). Mineral fertilizers increase SOC content by 0.06–0.08% annually, while manure application increases it by 0.24% (Körschens, Reference Körschens, Mueller, Sychev, Dronin and Eulenstein2021). In an experiment conducted at the Timiryazev Academy, 60 years after its establishment, the average values were found to be 1.03% for humus carbon and 0.079% for total nitrogen, with a C/N ratio of 13 (Mazirov and Safonov, Reference Mazirov and Safonov2010). In samples taken in 2023, that is, 110 years after establishment and 50 years after the last measurement, SOC averaged 1.36%, total nitrogen equaled 0.108% and the C/N ratio was 12.6. Therefore, we can conclude that the use of fertilizers contributes to soil fertility and the increase in biogenic element content, albeit more slowly than in European experiments. However, such discrepancies in values may be due to differences in chemical analysis methods, highlighting the need for further studies.
The dependence of SR on temperature and moisture has been extensively discussed in previous studies. The stimulating effect of temperature on CO2 emission intensity from soil has been observed in both farm conditions (Brito et al., Reference Brito, Azenha, Janusckiewicz, Cardoso, Morgado, Malheiros, La Scala, Reis and Ruggieri2015; Anokye et al., Reference Anokye, Logah and Opoku2021; Lei et al., Reference Lei, Wang, Zhang and Chen2022; Wang et al., Reference Wang, Li and Li2023) and field experiments (Kong et al., Reference Kong, Liu, Wang, Akhtar, Li, Ren, Feng and Yang2019; Kulachkova et al., Reference Kulachkova, Derevenets, Korolev and Pronina2023). Furthermore, the impact of increased temperature is more noticeable in wetter conditions, as well as in plots with lower pH values and high SOC content (Yang et al., Reference Yang, Pan, Wang, Tian, Zhang, Zhao, Hu, Yang, Yan, Ma, Chen, Quan, Wang and Niu2023). The relationship with moisture is generally nonlinear: SR decreases with both low precipitation (Apostolakis et al., Reference Apostolakis, Schöning, Michalzik, Klaus, Boeddinghaus, Kandeler, Marhan, Bolliger, Fischer, Prati, Hänsel, Nauss, Hölzel, Kleinebecker and Schrumpf2022) and excessive moisture (Yilmaz, Reference Yilmaz2019). However, in arid ecosystems, where moisture is a limiting factor, increased precipitation always leads to a significant rise in SR (Francioni et al., Reference Francioni, Trozzo, Toderi, Baldoni, Allegrezza, Tesei, Kishimoto-Mo, Foresi, Santilocchi and D’Ottavio2020; Morris et al., Reference Morris, Hornum, Crystal-Ornelas, Pennington and Bond-Lamberty2022).
The direct dependence of SR rate on soil organic matter content is undisputed (Feng et al., Reference Feng, Wang, Song and Zhu2018; Byanjankar et al., Reference Byanjankar, Dhamala, Maharjan and Kayastha2020). However, recent studies have clarified this dependence, revealing that soil CO2 emission is more dependent on the POM pool, which is the most active fraction of organic matter (Koutika et al., Reference Koutika, Dassonville, Vanderhoeven, Chapuis-Lardy and Meerts2008; Schlüter et al., Reference Schlüter, Roussety, Rohe, Guliyev, Blagodatskaya and Reitz2022).
Thus, the similarity of SR under different crops and its predominant dependence on the carbon and nitrogen content of POM, as observed in the current study, allows us to recognize that the level of CO2 emission from soil at the farm level is relatively stable. A sharp increase in SR following the application of organic-mineral fertilizer could inform the development of future compensation measures for carbon sequestration in response to specific fertilizer doses.
Conclusion
Applied fertilizers and cultivated crops significantly affect SR, while the interaction between them and the aftereffect of fertilizers are insignificant. Applied fertilizers account for 17.2–24.0% of the total variance in SR, whereas crops account for only 6.5–7.1%.
During the growing season, SR exhibits extremely high variance and does not differ significantly between plots with various crops and fertilizers. During the bare soil period, the SR rates are similar for different fertilizer variants, and under snow cover, SR is 10–20 times lower than within the growing season.
A direct dependence of SR on SOC and POM content was observed. However, it is impossible to identify a general trend in the influence of temperature and moisture on SR, as various crops and fertilizers show significant differences.
Author contributions
OS: conceptualization, investigation, visualization, writing the original draft; AR: field measurements; AP and TL: analysis of soil samples; DK: scientific consulting; IZ: supervision of the field experiment. All authors participated in the review and editing of the manuscript.
Funding statement
The research was conducted as a part of the Institute of Geography RAS State Assignment, Project No. FMWS-2024-0007.
Competing interests
All authors declare that they have no conflicts of interest.
Ethical standards
Not applicable.
