The Effect of Mineral Fertilizers on Soil Respiration in Urban Lawns

Application of mineral fertilizers to regulate microbial respiration and carbon dioxide emissions from urban lawn soils was evaluated due to the need to develop technologies for reducing CO2 emissions and for increasing the C-absorption capacity of natural and anthropogenic ecosystems. The studies were performed in the Botanical Garden of Moscow State University on Leninskie Hills in a small-plot experiment with the fractional application of four types of complex fertilizers (NPKS 27 : 6 : 6 : 2, NPKS 21 : 10 : 10 : 2, NPK 15 : 15 : 15 and NPK 18:18:18 + 3 MgO + trace elements (TE)) at the doses of 60 and 120 kg N/ha during the growing season. We studied the basal respiration (BR) of soils, carbon content of microbial biomass (Cmic) by substrate-induced respiration, and the CO2 emission from soils by the method of closed static chambers. Cmic in soil of the control plot in the summer period was 1300–1450 µg/g. Application of NPKS 21 : 10 : 10 : 2 and NPK 18 : 18 : 18 + 3 MgO + TE at a low dose increased Cmic by 12–35% within the first two weeks, and then it dropped. All types of fertilizers applied for a short period of time increased BR of soils and CO2 emission maximum on the sixth day. After two weeks and onwards, their growth decreased or their intensity dropped to the control values (500 mg CO2 m2/h–1 and 1.5 μg C–CO2 g/h, respectively) and lower. The lowest intensity of CO2 emission, a rise in basal respiration, and an increase in microbial biomass were recorded after the application of fertilizer NPKS 21 : 10 : 10 : 2. The change in the functioning of the soil microbial community detected by the maximal qCO2 was the greatest in case of NPKS 27 : 6 : 6 : 2 application. The dynamics of CO2 emission from the soils of the small-plot experiment from April to October correlated with the soil temperature (rS = 0.66, p < 0.05, n = 135). Emissions of CO2 were minimal for the plot with NPKS 21 : 10 : 10 : 2 in all periods of the study.


INTRODUCTION
The increase in the carbon dioxide content in the atmosphere of our planet remains an urgent problem due to its effect on climate change.Internationally accepted agreements and strategies of Russian economic development are aimed at reducing greenhouse gas emissions and increasing their absorption by ecosystem components.About a third of atmospheric CO 2 is of soil origin (Kudeyarov, 2015), and the global value of soil respiration according to recent estimates is 94.3 ± 17.9 Gt C/year (Xu et al., 2016).Soils in cities may also be a source of CO 2 entering into the atmosphere.Their contribution has not yet been sufficiently assessed, but may be significant, as it is shown, for example, for soils under lawns (Decina et al., 2016;Shchepeleva et al., 2019).
Fertilizers are one of the many factors affecting greenhouse gas emissions from soils.Their application may play an important role in regulating CO 2 fluxes in the city (Livesley et al., 2010;Decina et al., 2016).According to Decree of the Government of Moscow no.743-PP of 2002, fertilizers are intensively used during creation and maintenance of urban lawns to increase their productivity and sustainability.Fertilizers affect vegetation and soil biota and thus may participate in the formation of fluxes of climatically active gases in urban ecosystems.
There is no unequivocal opinion in published works, concerning the effect of mineral fertilizers on greenhouse gas emissions from soils.In addition, works on urban areas devoted to this issue are not numerous, managed forests and agroecosystems are mainly studied.There are data on a decrease in the intensity of soil respiration as a result of application of nitrogen fertilizers in some studies (Ding et al., 2007;Huth et al., 2010;Wang et al., 2017) and on an increase in others (Comeau et al., 2016;Chen et al., 2018;Huang et al., 2021;Zhang et al., 2021).The contradictory results are first related to differences in soils in the experiments, in particular, in their fertility, content and types of organic matter, texture, temperature, and humidity.Second, the plant species used in the experiments and their needs for fertilizers due to the lack of nutrients in soils differed.Third, the design of the experiments was not uniform: various types and rates of fertilizers and methods of their application were used.This determines the relevance of such studies under specific urban conditions.
The purpose of this study was to assess the effect of mineral fertilizers on soil respiration of urban lawns under long-term functioning in one of the botanical gardens of Moscow.
The tasks included establishing the short-term effect of different types and doses of applied complex mineral fertilizers on microbial respiration and CO 2 emission from soils in the summer period and identifying the seasonal dynamics of these processes.
We have suggested that different types and doses of fertilizers may exert a multidirectional effect on CO 2 formation and emission from soils due to stimulation or suppression of their microbiota, which can be used during creation and maintenance of lawns in the city to control greenhouse gas emissions.

MATERIALS AND METHODS
The study object was represented by gray-humus urban-stratified medium loamy soils on a technogenic substrate of a small-plot experiment with the application of complex mineral fertilizers.Soil names were given according to the classification from the work (Prokofeva et al., 2018).The small-plot experiment was laid in the Botanical Garden of Moscow State University at the Leninskie Hills on a restored lawn in the sector Exposition of Summer Annuals in July 2020 and described in detail in (Korolev et al., 2022).Lawn grasses included ryegrass (Lolium perenne L.), fescue (Festuca pratensis Huds.), bluegrass (Poa pratensis L.), and timo-thy (Phleum pratense L.).
Prior to the experiment, the content of mineral forms of nitrogen in soils was low and soil provision with available forms of phosphorus and potassium was high (Korolev et al., 2022).Four types of complex mineral fertilizers were used in the experiment.They differed in NPK content, nitrogen forms (ammonium, nitrate, and amide), and the presence of additional elements (Table 1).These were three fertilizers manufactured by Uralchem under the commercial names Azofoska NPKS 27 : 6 : 6 : 2 (hereinafter, Azofoska), Universal NPK 18 : 18 : 18 + 3 MgO + trace elements (hereinafter, Universal), and Nitroammofoska NPKS 21 : 10 : 10 : 2 (hereinafter, Nitroammofoska-1), as well as one fertilizer produced by the company NOV-AGRO under the commercial name Nitroammofoska NPK 15 : 15 : 15 (hereinafter, Nitroammofoska-2).Azofoska, Nitroammofoska-1, and Nitroammofoska-2 were granular, and Universal was powder.The fertilizers were applied by portions during the growing season after lawn mowing: they were evenly spread over the surface without disturbing the grass cover, which was then irrigated.The experiment was performed in three replications, and the area of one experimental plot was 4 m 2 .The fertilizers were applied in 2021 on May 28, July 2, August 3, and September 1 by 15 and 30 kg N/ha on each date and in 2022 on May 26, June 30, and July 29 at the doses of 20 and 40 kg N/ha.The control plot was located at a distance of 10 m from the plots with fertilizers.Air and soils were sampled in 2021 on October 12 and in 2022 on April 19 and June 30 (prior to the application of the second portion of fertilizers) and on July 6, July 12, and September 13.
Since a long-existing lawn was used for the experiment and the heterogeneity of soil properties could make it difficult to identify the effect of fertilizers, one plot in each experiment variant with properties closest to the control ones was chosen for monitoring soil respiration.Measurements and sampling of the surface horizons of soils (0-5 cm) at each plot were carried out three times.The soil profile was studied by drilling wells (0-80 cm) on the control plot and in the center of the plots with fertilizers on June 29, 2022.
The soil temperature was measured by an HI98501 Checktemp electronic thermometer (Hanna Instruments) with a resolution of 0.1°C and an accuracy of ± 0.2°C.We determined soil moisture by the thermostatic-weight method, total water capacity by the laboratory method of tubes (Vadyunina, Korchagina, 1983), soil density by the method of cutting rings (GOST (State Standard) 5180-2015), the content of carbon of organic matter by the Tyurin method modi- Basal respiration (BR) and carbon content of microbial biomass (C mic ) in soils were determined in fresh samples at natural humidity, which was 38-66% of the total water capacity (TWC) except for April 2022, when it increased to 70-90% of TWC, and September 2022, when it decreased to 22-30% of TWC.Prior to the analysis, the soil was sifted through a 2-mm screen and plant residues were removed.Two weighted portions (2 g) of each soil sample were placed in a penicillin bottle and pre-incubated in a desiccator over distilled water for two days (ISO 18400-206) at a temperature of +22°C.BR was evaluated by CO 2 release by native soil.The vials with soil were hermetically sealed, air samples were taken at the beginning of the experiment and after 23-24 h (ISO 16072-2002).The intensity of BR was calculated per absolutely dry soil and given in μg of C-CO 2 g/h.The content of C mic in soils was determined by the method of substrateinduced respiration (SIR) by adding a glucose solution (0.2 mL) at the concentration of 10 mg/g of absolutely dry soil as a substrate, which was easily available for microorganisms (Anan'eva et al., 2009).The incubation period was 2-3 h.C mic (μg/g) was calculated according to the conventional formula (Andersen, Domsh, 1978): (1) Microbial metabolic coefficient qСО 2 (specific microbial respiration, μg C-CO 2 mg C mic /h) was calculated by the BR to C mic ratio.
The potential microbial production of CO 2 was estimated by the formula where MP is the microbial production of CO 2 , mg m 2 /h, Vbas is the basal respiration of soils, μg С-СО 2 g/h; m s is the soil mass in the 5-cm-thick layer per an area of 1 m 2 , g; 3.66 is the conversion factor of C to CO 2 ; and 1000 is the coefficient to take into account the conversion of micrograms to milligrams.
The carbon dioxide emission from soils was determined by the closed static chamber method.We used small chambers of regular cylindrical shape (h = 11.5 cm, d of the base = 10 cm), which were set after removing the aboveground part of plants on plastic bases embedded into the soil to a depth of 3-4 cm.The exposure period was 3 min (Kulachkova, Kovalenko, Vbas 2021).Air was sampled by insulin syringes during one day from all plots from 10.00 a.m. to 1.00 p.m., and the analysis at a laboratory was carried out at the same day.The CO 2 emission from soil was calculated by the formula: (3) where q is CO 2 emission, mg m 2 /h, ΔC is the change in gas concentration in the chamber (g/m 3 ) during the exposure period, and h is the height of the chamber, m (Smagin, 2005).
Quantitative analysis of the carbon dioxide content was carried out on a Cristalluks 4000M gas chromatograph (Meta-khrom, Yoshkar-Ola, Russia) with a thermal-conductivity detector.The length of the columns was 3.2 m; the filler was represented by Polisorb-1; helium was used as the carrier gas, with its flow rate being 30 mL/min; the temperature was 50°C for the columns, 60°C for the detector, and 60°C for the evaporator; the current of the measuring elements was 100 mA; and the volume of the analyzed sample was 0.5 cm 3 .The stability of the chromatograph calibration was monitored prior to the analysis.The error of the measurement results determined during the chromatograph verification was 2.0%.
Statistical data processing was performed in the StatSoft Statistica 12.0 program.The correspondence of the data to the normal distribution was determined by the Shapiro-Wilk criterion.Nonparametric methods of analysis were used for the parameters, whose normal distribution was not confirmed, as well as for small samplings (n = 3).Data were compared by the Mann-Whitney U test for soils of different plots and by the Wilcoxon matched pairs test for the same plots in different periods.The diagrams show the medians and the interquartile range.The relationships between soil parameters were evaluated by Spearman rank correlation coefficient r S .The significance level was 0.05.

Soil Characteristics
The properties of gray-humus urban-stratified soils of the control plots and plots with fertilizers are similar.The upper AY horizon is dark gray with a powdery lumpy structure and medium loamy, which as single inclusions of brick, crushed stone, and glass (the total mass of artifacts is <5%).The thickness of the grayhumus horizon is 17-18 cm, the density is 1.0-1.1 g/cm 3 , and the C org content is 7.2-7.3%.The medium reaction and the electrical conductivity of the water extract at the control plot and the plot with fertilizers are slightly different: pH is 7.0 and 6.4, and EC is 69 and 104 μS/cm, respectively.It is underlain by three technogenic ТСН horizons: to a depth of 52 cm on the control plot and to a depth of 65 cm on the plot with fertilizers.Technogenic horizons are of inhomogeneous grayish-brown color with whitish and rusty = Δ × × Δ 1000 / , q C h t spots and black iron-manganese specks and are distinguished by the change in the texture from mediumloamy to heavy-loamy and by the abundance of technogenic inclusions.The technogenic horizons on the control plot are characterized by more alkaline medium reaction, greater electrical conductivity, and a higher content of C org as compared to the fertilized plots.The рН parameter of technogenic horizons increases with depth from 7.9 to 8.2 on the control plot and from 7.0 to 7.6 on the plots with fertilizers; EC decreases with depth from 125 to 81 μS/cm and from 62 to 58 μS/cm, respectively; and the C org content decreases from 5.4 to 3.0% and from 3.9 to 1.9%, respectively.Natural eluvial and subeluvial horizons of texturally differentiated soils are discovered under the technogenic layer; they are characterized by a neutral medium reaction (рН = 7.5), low electrical conductivity (26-40 μS/cm), and low content of carbon of organic matter (0.7-0.8%).
The C mic content and BR decrease significantly down the soil profile.The maximal values are typical for the upper 5-cm layer of the AY gray-humus horizon, so we have studied their dynamics and will discuss them below.On the plots with the studied profile characteristics, the C mic content varies from the surface within 1000-1300 μg/g, and the BR varies within 0.70-1.50μg C-CO 2 g/h.The C mic content is 650-760 μg/g, and the BR is 0.60-0.79μg C-CO 2 g/h in the middle part of the AY horizon and decreases with depth from 333-500 to 122-223 μg/g and from 0.33-0.50 to 0.13-0.25 μg C-CO 2 g/h, respectively, in technogenic horizons.The C mic and BR values are minimal in the lower eluvial and subeluvial horizons: 60-62 μg/g and 0.08-0.09μg C-CO 2 g/h.
The characteristics of the surface soil horizons of the monitoring plots are given in Table 2.The soil surface reaction was neutral on the control plot and slightly acid on plots with fertilizers, and pH values varied slightly (the coefficient of variation (CV) was ≤3%).The variation of EC and C org was greater, and their distribution did not correspond to normal.The electrical conductivity of soils with fertilizers applied was 1.2-2.8times higher as compared to the control.Greater doses of Nitroammofoska 1 and 2 caused a rise in this parameter to higher levels than in other variants.When the doses were similar, the electrical conductivity of the soils on plots with various fertilizers did not significantly differ.The C org content in soils on fertilized plots differed by no more than 1% from the control, except for the sites with high doses of Nitroammofoska 1 and 2. Soils were contaminated with heavy metals on all the plots: the content of mobile forms exceeded the MPC 4.5-6.5 times for copper and 2.1-2.9 times for zinc (SanPiN 1.2.3685-21) (the distribution was normal, and the CV was 12% for the copper content in soils and 8% for zinc).
The temperature of the surface horizon of soils on the monitoring plots varied slightly: the CV did not exceed 10%.The temperature difference between neighbor plots during sampling did not exceed 1.2-2.2°C(Fig. 1a) and was not statistically significant.The soil was the least heated in April (4.2-5.9°C) and the most heated in early July (21.9-24.0°C).In summer, the soil temperature increased by about 3°C from June 30 to July 6 and decreased by 1.5-2°C by July 12. Temperature differences over the measurement periods were statistically significant (n = 27, p < 0.00001).
The nonuniformity of soil moisture (W) was more pronounced than that of temperature (Fig. 1b): the CV ranged from 13 to 18% in various periods, but statistically significant differences between the plots were not found.The driest conditions were observed in September, when W was 15-22% (22-30% of TWC), and the wettest in April after snow melting (W = 40-66% (57-97% of TWC)).During the summer studies, soil

Dynamics of Microbial Respiration and CO 2 Emission from Soil during Fertilization in Summer
The short-term effect of fertilizers on soil respiration parameters was studied in summer: a week, two weeks, and a month after the application.The C mic content in soils of the control plot varied slightly and increased by only 1.1 times during the studied summer periods.On all fertilized plots, the C mic content was lower than on the control a month after the application of the first portion (June 30, 2022), and its decrease was the greatest under the effect of Nitroammofoska-2 and Azofoska (to 30-40%) and the smallest in the case of Nitroammofoska-1 application (≤16%) (Fig. 2, points).
It should be pointed out that, in soils with high doses of fertilizers, the content of microbial biomass was comparable to or even smaller than in soils with lower doses.The introduction of the second portion of all types of fertilizers stimulated the growth of biomass of microorganisms, which was usually more intensive during the first week (the differences were statistically significant, n = 27, p < 0.00005).During the second week, the biomass growth was slowed down or even decreased in soils with the application of both rates of Azofoska and Nitroammofoska-2 or high doses of other fertilizers (the differences were not significant).Two weeks later, the C mic content was greater in soils with the application of low fertilizer doses and was maximal under the effect of Universal (1962 μg/g) and Nitroammofoska-1 (1633 μg/g).The C mic content exceeded the median of the contol values (1454 μg/g) only on plots with low doses of these two fertilizers: the increase was 35 and 12%, respectively.The stimulating effect of high doses of these fertilizers and of both doses of Azofoska and Nitroammofoska-2 for the formation of microbial biomass lasted only one week.
The basal respiration of soils on the control plot was less intensive than in soils with fertilizers in both the short-term and the seasonal studies, which are discussed below.Excess BR in soils of control over BR in soils with fertilizers was noted in no more than 20% of cases.In soils of the control plot, BR increased by 1.8 times from June 30 to July 6, which was related to a rise in soil temperature and moisture.On July 12, it remained at the same level (Fig. 2, columns).On fertilized plots, BR of soil also increased one week after the application of the second fertilizer portion, but usually decreased by the end of the second week (differences between the data are statistically significant, n = 27, p < 0.008).High rates of fertilizers caused a stronger increase in BR of soil, as compared to low doses: by 1.5-4.7 and 1.2-2.0times, respectively.The increase in the BR rate of soils on plots with both doses of Azofoska and high doses of Nitroammofoska-1 and Universal exceeded the increase on the control plot.
The change in BR relative to the control was the smallest in soils under the effect of low doses of Nitroammofoska-1 and the highest among low doses under the effect of Azofoska.The increase in BR of soils was maximal in case of the application of high doses of Azofoska and Universal.
Metabolic coefficient qCO 2 reflects the efficiency of the use of the substrate and at the same time is a measure of stress of microorganisms (Anan'eva et al., 2009).On the control plot, qCO 2 was often lower than on fertilized plots (Fig. 3).The differences from the control were the strongest under arid conditions at the end of June, a month after the application of the first fertilizer portions (54-215%), and the lowest a week after the application of the second fertilizer portion (2-16%) except for Azofoska (33-91%).
The control qCO 2 values increased by 1.8 times from June to July (0.65 and 1.18 μg С-СО 2 mg С mic /h, respectively).In soils with fertilizers applied, the increase in qCO 2 during the same period was characterized by a smaller amplitude, with the exception of sites with a high rate of Nitroammofoska-1 and Universal a week after the application of the second fertilizer portion (the difference between 30.06 and 06.07 was statistically significant, n = 27, p = 0.003).According to absolute values, the specific microbial respiration of microorganisms was maximal on the plot with Azofoska.
Prior to the application of the second portion of fertilizers, CO 2 emission from soils of all plots was comparable to the control or insignificantly lower (Fig. 4).On the control plot, the CO 2 emission from soils remained constant on June 30 and July 6 (about 500 mg СО 2 m 2 /h), but decreased 1.5 times to July 12. Contrary to the control, the fertilized plots were characterized by a 1.1-to 2.3-fold increase in the intensity of soil respiration a week after fertilizer application (the differences were statistically significant, n = 27, p = 0.0001).The magnitude of increase in soil respiration did not depend on the applied fertilizer dose.Low doses of Azofoska and Universal increased soil CO 2 emission to a greater rate than their high doses: by 24 and 36% as compared to the control, respectively, contrary to 1 and 29%.High rates of Nitroammofoska-1 and 2 exerted a greater effect: the emission increased by 80% and 50% as compared to the control contrary to 8 and 25%, respectively.The impact on the increase in CO 2 emission was minimal for the low dose of Nitroammofoska-1 and a high dose of Azofoska.

Pg/g
Two weeks after fertilizers application, the rise in CO 2 emission continued under the effect of high doses of Azofoska and Universal: it was two times greater than the control.On other fertilized plots, the intensity of soil respiration significantly decreased.It was similar to the control values in areas with low fertilizer doses and exceeded them in the case of high doses.
We evaluated the correlation between the intensity of microbial CO 2 production and CO 2 emission from soils.We calculated the emission rate of microbial CO 2 by the surface 5-cm-thick soil horizon, taking into account the soil density and BR (formula (2)).Since BR was determined in resently taken soil samples, the values of CO 2 production by microbes in summer may be considered the closest to reality.A month after the first portion of fertilizers was applied, the microbial CO 2 production in areas with a low fertilizer dose amounted to 40-50% of soil emission into the atmo-

Azofoska
Nitroammofoska-1 Nitroammofoska-2 Universal sphere, except for the variant with Nitroammofoska-2, which was 72%, as opposed to 18-36% at its high dose.A week after the application of the second portion of fertilizers, the percentage of microbial respiration increased to 50-70% of the total CO 2 emission from soils in areas with a low dose and to 34-99% in the variant with high fertilizer dose.A week later, these shares amounted 50-94 and 28-81%, respectively.

Seasonal Dynamics of Microbial Respiration and CO 2 Emission from Soils under the Effect of fertilizers
Seasonal changes in the hydrothermal soil parameters affected the biomass of microorganisms and BR and CO 2 emission from soils: they increased from spring to summer and decreased in autumn, following the temperature.The absolute values of microbial biomass and soil BR were the lowest on all plots in dry September 2022, and soil CO 2 emission was the smallest in the wettest and cold October 2021 and April 2022 (the differences were statistically significant, n = 27, p < 0.00005).
The impact of fertilizers on microbial biomass did not depend on the survey season: the difference from control plots could be greater or smaller at any time.An increase in biomass in comparison with the control was seen under the effect of all fertilizers applied in the middle of summer and in all seasons under the effect of low doses of Universal (Fig. 2, points).
Changes in the basal respiration of soils on fertilized plots varied from season to season.If the first week after fertilization is not taken into account, the largest increase of BR relative to the control was typical for soils with a residual effect of high fertilizer doses in April (64-184% for Azofoska and both Nitroammofoskas).This correlated with the microbial biomass and could be explained by greater availability of nutrients during biota activation after winter.The exceeding of the control levels became from two to three times smaller on the same plots by mid-July (Fig. 2, columns).In autumn, the effect of fertilizers was minimal, and BR of soils on almost all plots was similar to the control.Low doses of Nitroammofoska-1 changed the BR of soils in all the studied periods to a smaller rate than the other fertilizers.The dynamics of specific respiration of microorganisms was similar, and the differences from the control were smaller in all seasons for the variant with Universal and low rate of Nitroammofoska-1.
The CO 2 emission from soil at the plots with fertilizer significantly increased from April to July (n = 27, p = 0.000006) and decreased or remained almost stable in September (Fig. 4).In spring, soil respiration on plots with fertilizers was significantly higher than on the control, which corresponded to regularities of microbial CO 2 production.In mid-July, it was higher than the control only in areas with high fertilizer doses; and in autumn it was lower or similar for all doses.The CO 2 emission from soils was most often minimal for Nitroammofoska-1 in all seasons.

DISCUSSION
In summer, the content of carbon of microbial biomass in lawn soils ranged from about 150 to 800 μg/g in recreational areas of the cities of Moscow oblast and from 200 to 1000 μg/g in residential zones (Ivashchenko et al., 2014).In our studies, C mic was usually higher than 1000 μg/g even in soils of the control plot and could exceed this level by more than two times in fertilized soils during the first week.This great microbial biomass was obviously related to the high content of organic matter in soils of the Botanical Garden of Moscow State University.According to our data, the BR of soil in the first two weeks after fertilization, by high doses in particular, was almost two times greater than in recreational areas of the cities of Moscow region (0.23-1.37 μg C-CO 2 g/h) and 1.4 times higher than the parameters of residential areas (0.22-2.04 μg C-CO 2 g/h) (Ivashchenko et al., 2014).The CO 2 emission from soils observed for the lawn of a city park with similar soil types studied by us varied from about 400 mg of CO 2 m 2 /h in April and September-October to 1875 mg of CO 2 m 2 /h in July (Schepeleva et al., 2019).It was comparable to our data at the beginning and the end of the growing season, but higher in July.The differences could be related to vegetation types.In (Schepeleva et al., 2019), it was represented by the native meadow with minimal maintenance, whereas, in our study, this was a restored regularly mowed lawn with a probably smaller contribution of plant roots to soil respiration.
It is known that the parameters of the status and of the functioning of microorganisms and the CO 2 flow are influenced by hydrothermal soil properties (Kurganova et al., 2020;Goncharova et al., 2022).Temperature and soil moisture varied from season to season and over the two weeks when we tried to assess the effect of the applied fertilizers.A moderate positive correlation between microbial biomass and basal respiration of soil with temperature (r S = 0.47 and 0.52, respectively, p < 0.05, n = 81) and soil moisture (r S = 0.59 and 0.35, respectively, p < 0.05, n = 81) was determined for the period from June 30 to July 12 (the entire dataset).The relationship of CO 2 emission from soil with its temperature during this period was weak (r S = 0.30, p < 0.05, n = 81), and there was no relation with the humidity.This testifies that the variation in temperature during the summer period within 4-5°C and in humidity within 15-20% (provided that it was within 30-70% of the TWC) is not a factor for the studied soils, which determines the CO 2 flux from it.We revealed a slight positive correlation of C mic and BR with soil temperature (r S = 0.19 and 0.23, respectively, p < 0.05, n = 135) and moderate correlation with soil humidity (r S = 0.50 and 0.39, respectively, p < 0.05, n = 135) for the entire array of data from April to October, except for the period of the maximal soil respiration after fertilizer application during the first week of July (July 6, 2022).The seasonal dynamics of CO 2 emission from soil more strongly depended on hydrothermal parameters: there was a noticeable positive relationship with temperature (r S = 0.66, p < 0.05, n = 135) and negative correlation with soil moisture (r S = -0.51,p < 0.05, n = 135).Low temperatures and humidity above 70% of the TWC restricted CO 2 flows from soils.The negative correlation of CO 2 emission from soils with soil moisture was observed only in cold periods, when the soil temperature was low (r S = -0.66,p < 0.05, n = 81).Positive correlations of CO 2 fluxes with soil temperature in seasonal cycles confirm the previously identified regularities (Ding et al., 2007;Schepeleva et al., 2019).
Basal respiration of soils significantly correlated with microbial biomass (r S = 0.63, p < 0.05, n = 162) during all periods of the research, and CO 2 emission into the atmosphere, in turn, was moderately related with soil BR (r S = 0.32, p < 0.05, n = 162).These regularities confirm the results obtained by other authors (Ivashchenko et al., 2014;Kurganova et al., 2020;Goncharova et al., 2022).
The increase in the biomass and in the specific respiration of microorganisms, in soil BR, and in CO 2 emission to the atmosphere during the first week after the application of fertilizers is related to their total dose.This was confirmed by the significant Spearman correlation coefficients between these parameters individually for the plots with low and high doses of fertilizers.The correlation with the added portion of fertilizers was the slightest for the increase in microbial biomass (r S = 0.36 and 0.30 for plots with low and high doses, respectively, p < 0.05, n = 30), stronger for BR (r S = 0.45 and 0.61 for plots with low and high doses, respectively, p < 0.05, n = 30), and maximal for CO 2 emission (r S = 0.53 and 0.65 for low and high doses, respectively, p < 0.05, n = 30).Soils of the small-plot experiment were initially characterized by a lack of nitrogen, and so its input with fertilizers stimulated microbial activity and growth of small roots, which was reflected in the CO 2 emission.With the depletion of the additional nitrogen source over a relatively short time period (about two weeks in our experiment), the microbial activity and soil CO 2 emission became smaller.The increase that we observed in soil respiration immediately after fertilizer application was related to the phenomenon of priming effect described in published works (Kuzyakov et al., 2000).
A strong increase in BR in the studied soils of the Botanical Garden of Moscow State University in areas with higher doses of nitrogen fertilizers was not accompanied by a proportional rise in biomass.This may be related to partial immobilization of nitrogen, which increases, when its large quantities are applied to the soil.As a result, it is not involved in biosynthetic reactions (Evdokimov et al., 2005).
The effect of long-term application of various forms of nitrogen fertilizers on microbial respiration and carbon-nitrogen status of gray forest soil was previously investigated (Susyan et al., 2008).The authors showed an increase in the BR of soils and in the rate of specific respiration of soil microorganisms by 32-142% and a decrease in the portion of microbial carbon in soil organic matter contrary to an increase in crop yields as compared to plots without fertilizers.Increased parameters of specific respiration of microorganisms and BR of soils were also seen in our experiment on plots with fertilizers.The absolute specific microbial respiration was maximum on the plot with Azofoska, which testifies to the most stressful status of soil biota under the effect of this fertilizer.
The increased CO 2 emission from soil on the studied plots after fertilizer application was short-term.Fluxes of CO 2 from soils became smaller than those from the control plot or equal to them two weeks after the application of low doses of Nitroammofoska-1 and a month after the application of the rest portion fertilizers.
Smaller CO 2 emission by soils under the effect of fertilizers was recorded in a number of works for agroecosystems and managed forests.Experiments on the treatment of corn plantations by urea at doses of 150 and 200 kg N/ha were performed on well-drained slightly alkaline sandy loam alluvial plowed soils in the Chinese province of Henan (North China Plain).As a result, CO 2 fluxes decreased by 10.5% as compared to untreated soils (Ding et al., 2007).The authors suggested that the decomposition of soil organic matter was suppressed under the impact of nitrogen fertilizers.Several possible mechanisms were proposed.First, a lower pH related to nitrification could inhibit microbial activity.Second, the increase in the concentration of dissolved substances could exert a negative impact on microbial populations of the soil.Third, high nitrogen levels in nitrogen-rich soils inhibited the synthesis and activity of some enzymes (Ding et al., 2007).
Emissions of CO 2 and N 2 O under long-term effect of nitrogen fertilizers (urea) on wheat and sorghum with the introduction of leguminous crops in the rotation were modeled for the Northern grain region of Australia.Nitrogen addition with fertilizers increased the yield and thus slowed down the depletion of soil carbon and decreased twice CO 2 emission, but significantly increased the amount of produced N 2 O.The total decrease in greenhouse gas emissions from soils under the effect of fertilizers recalculated per CO 2 equivalent amounted to 10% (Huth et al., 2010).
Significant in comparison with the published data, but short-term (priming effect) impact of applied fertilizers (urea at the dose of 70 kg N ha/year) on CO 2 flows was revealed for plantations of oil palm on peat deposits of a deep-peat coastal plain in Indonesia (Comeau et al., 2016).During the second fertilizer application at the dose of 1 kg of urea per one palm tree, the studied parameters of heterotrophic soil respiration were 3.4 times lower than during the application of the first dose (0.5 kg of urea per palm tree).
Thus, the application of mineral nitrogen fertilizers affects soil respiration, and the study of mechanisms of this effect can be important for the development of relevant carbon sequestration technologies.

CONCLUSIONS
The work confirms our hypothesis that different types of fertilizers may exert various impacts on CO 2 formation and emission from soils due to their effect on the functioning of soil microbial communities.Various impacts of mineral fertilizers on short-term and seasonal dynamics of CO 2 emission from soils of urban lawn are shown.
The application of complex mineral fertilizers for a short period increases soil CO 2 emission and BR under the lawn in the city, becoming maximal on the sixth day.Within two weeks and one month, their rise is slowed down or becomes equal to the control values or lower.
The analysis of the short-term dynamics of CO 2 emission from soil after applying the second portion of fertilizers shows that the intensity of soil respiration and the surge in BR along with an increase in microbial biomass are the after Nitroammofoska-1 application (NPKS 21 : 10 : 10 : 2).Changes in the functioning of the microbial community of soils detected by the maximal qCO 2 are the greatest in the variant with Azofoska (NPKS 21 : 10 : 10 : 2).
A significant role of basal respiration in CO 2 emission from soils is revealed: it sometimes comprises to 81%, when higher doses of fertilizers are applied, and 94% in the case of their lower doses.In this regard, it is recommended to develop ways to control the functioning of soil microbial community to reduce carbon dioxide emissions.
Seasonal dynamics of CO 2 emission and microbial respiration of soils in the small-plot experiment are related to changes in temperature: the intensity of the processes increases from spring to summer and decreases in autumn.In all seasons, the CO 2 emission from soil is minimal on the plot with applied Nitroammofoska-1 (NPKS 21 : 10 : 10 : 2).
Our results show that the use of mineral fertilizers is possible not only to create and maintain an decoratively attractive lawn, but also to regulate the CO 2 fluxes from soil.The studies should not be limited by the effect of fertilizers on the microbial component of soil respiration, it is necessary to assess the impact on vegetation and to evaluate its contribution, which it is planned to be done in future published works.

Fig. 1 .
Fig. 1.(a) Temperature and (b) humidity of the surface (0-5 cm) soil horizons of monitoring plots: L, with a low dose (60 N kg/ha), and H, with a high dose (120 N kg/ha) of fertilizers; dates of short-term effect of fertilizers in summer are highlighted in gray.

Fig. 2 .
Fig.2.Content of carbon of microbial biomass (C mic , points) and basal respiration (BR, columns) of soils ( -low dose,high dose of fertilizers, -control; dates highlighted in gray reflect the period of short-term effect of fertilizers). 3

FUNDING
This work was performed within the framework of a state order of the Ministry of Science and Higher Education of the Russian Federation (theme no.121040800147-0 "Soil Information Systems and Optimization of the Use of Soil Resources") and of the program for the development of the Interdisciplinary Scientific-Educational School of Moscow State University "The Future of the Planet and Global Environmental Changes."

Table 1 .
Composition of complex fertilizers (%) according to manufacturers

Table 2 .
Properties of surface (5-cm-thick) soil horizons of monitoring plots L is a low dose of fertilizer (60 N kg/ha), H is a high dose of fertilizer (120 N kg/ha), рН is the mean ± standard deviation, EC and C org are the median (min-max), Cu and Zn are the result of a single determination in a mixed sample (July 2020), and n = 3 for other parameters (July 2022).