Carbon dioxide emission and soil microbial respiration activity of Chernozems under anthropogenic transformation of terrestrial ecosystems

Nadezhda D. Ananyeva a,*, Sofia V. Rogovaya a, Kristina V. Ivashchenko a,b Vyacheslav I. Vasenev b, Dmitriy A. Sarzhanov d, Оleg V. Ryzhkov c Valeriy N. Kudeyarov a a Institute of Physicochemical and Biological Problems in Soil Science of the Russian Academy of Sciences, Pushchino, Moscow region, Russia b Peoples’ Friendship University of Russia, Agricultural Faculty, Moscow, Russia c The Central-Chernozemic State Biosphere Nature Reserve named by Prof. V. Alekhin (CCSBNR), Zapovednoe, Russia d Russian Timiryazev State Agrarian University, Moscow, Russia Abstract


Introduction
The circulation of carbon dioxide between soil and atmosphere provided by two main processes: plants photosynthesis and respiration of soil.Soil respiration (soil CO2 emission), in turn, is provided by respiration of soil microorganisms and plant roots.It is believed that about 70% of total soil CO2 emission was derived by soil microbial respiration (Zavarzin and Kudeyarov, 2006).According to many researchers the portion of respiration of soil microorganisms and plant roots in total soil CO2 emission depends on hydrothermal conditions (Ryan and Law, 2005;Martin and Bolstad, 2005), photosynthesis activity (Kuzyakov and Gavrichkova, 2010) and soil organic matter composition (Metting, 1993), but it mainly remains still unclear (Hanson et. al, 2000;Kuzyakov and Larionova, 2005).In addition, the separation of these soil CO2 fluxes in natural condition is difficult to defined (Kuzyakov and Larionova, 2005) and time-and labor-consuming (Yevdokimov et al., 2010).Moreover, the determination of these two portions in total soil CO2 emission might be very important for studying carbon cycle and modeling carbon change in terrestrial ecosystems (Wie et al., 2010).Chernozems is an important natural resource of Russia, the distribution of Chernozems area is about 6% of the country (National Soil Atlas..., 2011).Plowing Chernozems is currently reached up 50-60%, which made almost 2 / 3 of all agricultural production in Russia.For the last 150 years the carbon content of Chernozems was decreased by 20-30% (Mikhailova and Post, 2006).Besides, the urban area in Chernozems zone increases by 10% for 2010-2015 yrs (Statistical Pocketbook, 2015).Anthropogenic transformation of terrestrial ecosystems (agricultural use, urbanization) leads to the changes of soil microbial community functioning.It was shown that in arable Chernozems the soil microbial biomass content was dramatically decreased by almost 3-4 times (Senicovscaia, 2012;Ivashchenko et al., 2015).In Chernozems of Voronezh region (Russia) the soil microbial biomass carbon and its portion in total soil organic carbon were decreased by 2 times and by 40%, respectively, compared to natural analogue (Blagodatskii et al., 2008).Our study was focused on: i) the measurements of total CO2 emission from Chernozems and portion of soil microbial respiration in situ in natural and anthropogenically transformed ecosystems; ii) the parameters estimation of soil microbial community functioning (soil microbial biomass carbon content, basal respiration, specific respiration of microbial biomass and fungi-to-bacteria ratio); iii) the assessment of relationship between soil microbial respiration measured in situ and laboratory conditions.

Field measurement
The total soil CO2 emission was measured (closed-chamber, LI-820) in five spatially distant points on the plot (20 × 20 m) of each ecosystem (ground vegetation cut) and expressed as g CO2 m -2 d -1 (20 totally).In each point of the ecosystem the soil temperature and soil moisture were recorded at 10 cm depth.The measurements were carried out in early May, June and July, 2015 yr.Soil samples were taken from 10-cm layer for chemical and microbiological (60 totally) analyzes.Soil microbial respiration in situ was determined by substrate induced respiration method (Larionova et al., 2006;Yevdokimov et al., 2010).Into soil steppe, forest and urban the four "collar-base" were cut in 10-cm depth on the distance 1-2 m from points for total CO2 emission measurement (Figure 1).In two "collar-base" it was unsieved soil (with roots), in the other two it was sieved soil (mesh 3 mm, roots excluded).The measurement of soil CO2 from the surface of four "collar-base" started not earlier than in half an hour.The water or glucose solution was added (slow penetration) into unsieved and sieved soils of "collar-base" then.In preliminary experiments it was found that the volume (water or glucose solution) provided slow penetration of liquid through 10 cm soil layer was equaled 0.6, 0.9 and 1.0 L for forest, steppe and urban, respectively.The glucose concentration provided the highest initial soil substrate-induced respiration (Anderson and Domsch, 1978) was amounted 5 mg g -1 soil in our experiments.The time between addition of liquid to soil and soil CO2 measuring was 4 h (preliminary experiment).Soil microbial respiration (MR) of unsieved (with roots) and water-moistened soil was calculated according by following: where MR is soil microbial respiration, g CO2 m -2 d -1 ; GL is CO2 emission from enriched glucose soil, g CO2 m -2 d -1 ; W is CO2 emission from soil with water, g CO2 m -2 d -1 ; UNS is unsieved soil (with roots); S is sieved soil (without roots).The (GL/W)S ratio was characterized the excess of CO2 emission from sieved soil added glucose compared to soil added water.The (GL-W)UNS / (GL-W)S ratio was characterized the soil disturbance (sieving).The MR portion in total CO2 emission from unsieved (with roots) and water-moistened soil was expressed as MR / WUNS ratio (%).Then we calculated the MR portion in total soil CO2 emission from nearest point without any liquid addition.Soil MR of steppe, forest and urban was measured in two replicates.

Statistical analysis
The measurements were performed in three replicates for SIR, BR and fungi-to-bacteria ratio.The results were calculated for dry soil (105°C, 8 h) and expressed as mean ± standard deviation (Excel).Significance of the difference in total soil CO2 emission, hydrothermal and microbiological soil parameters between ecosystems was tested by one-factor analysis of variance (ANOVA) and Tukey's multiple comparison test.Statistic tests were chosen based on the preliminary analysis: normality distribution of experimental data was checked by Shapiro-Wilk test, variance homogeneity was checked by Levene's test.Relationship between MR and total soil CO2 emission, microbiological and hydrothermal soil parameters was analyzed by Pearson's correlation coefficient.Relationship between MR and BR was analyzed nonlinear regression.All experimental data was statistically analyzed and visualized (box-plot) using Statistica 10.0 software.A principal component analysis and ordination of experimental data was carried out by PCord 4.27.

Results and Discussion
The soil organic carbon content (Сorg) of natural ecosystems was by approximately 2 times higher than that of anthropogenically transformed, and the pH value of steppe and forest was by about one unit less than bare fallow and urban (Table 1).The CO2 emission from Chernozems in various months was ranged from 2.0 (fallow) to 23.2 (steppe) g CO2 m -2 d -1 , these values differ by almost an order of magnitude (Table 2).In May the highest average soil CO2 emission was found in steppe, forest and urban ecosystems, and the lowest value was in fallow, wherein the soil temperature in forest was significantly low and the soil moisture was high compared to other studied ecosystems.In June a significantly high soil CO2 emission was found in steppe, and the low one was in fallow, the soil of which was significantly high temperature and low moisture.In the warmest month (July) the soil CO2 emission from fallow was also the lowest (high soil temperature and low soil moisture).During the observed period (May-July) the high soil CO2 emission was found in steppe, and the low was in fallow (in average 20.3 and 3.6 g CO2 m -2 d -1 , respectively), the difference between these values was almost 6 times (Figure 2).The soil Cmic and BR of various ecosystems for May-June-July were ranged from 284 (urban) to 1710 (steppe) µg C g -1 soil and from 0.28 (fallow) to 1.64 (steppe) µg CO2-C g -1 soil h -1 , respectively (Table 3).The soil Cmic and BR values of steppe and forest were mainly significantly higher than fallow and urban for each studied month.The qCO2 value in urban soil was on the contrary significantly high in May and June, however in July, the difference of this index for studied ecosystems was not found.For the observed period, the soil Cmic content and BR rate of natural ecosystems (steppe, forest) were significantly higher than the anthropogenically transformed (fallow, urban), however the qCO2 of urban soil was significantly higher compared to other ecosystems (Figure 3).Therefore, there is a base to consider the "deterioration" of soil microbial community functioning in anthropogenically transformed ecosystems compared to natural analogues.In our experiments the highest SIR inhibition by antibiotics both was achieved 41-51% (Table 4).The fungi portion in urban and steppe soils was almost the same (82-85%), wherein the fungi / bacteria ratios were also approximately equal (3.4 and 3.8, respectively).However, the Cmic / Corg (as an indicator of soil organic matter "quality") and Cfungi / Corg ratios for urban soil were 2.6 and 2.4 times less than those for steppe.It might be indicated the essential "deterioration" of soil microbial community functioning under anthropogenic impact.The soil MR of steppe, forest and urban ecosystems for three months of observation was varied from 4.8 (urban) to 17.5 (forest) g СО2 m -2 d -1 and amounted in average 6.9, 9.1 and 10.8 g СО2 m -2 d -1 for urban, steppe and forest, respectively, however these values were not significantly differ (data not shown).The MR portion in total soil CO2 emission in May was varied from 27% in urban to 91% in forest (Figure 4).The highest portion of MR was found in forest, it was 76 и 91% for two replicates (points).The highest difference of MR between replicates was in forest and urban.In the first replicate of forest it was a rich undergrowth of shrubs (more roots), and in the second replicate it was rare (less roots).The first replicate of urban industrial zone was covered by rich grasses and had sod cover (less MR), the second replicate was almost without grass cover (more MR).The MR portion in total soil CO2 emission for the three studied months was the highest in forest and amounted in average 83% (Figure 5).The MR portion in total soil CO2 emission of steppe and urban was less and amounted in average 51 and 60%, respectively.
Between MR and soil total CO2 emission (or BR, Cmic and soil water content) the positive correlation was found (r = 0.48, 0.51, 0.34, respectively) for studied months.Between MR and soil temperature the correlation was weak (r = 0.06).Between MR (in situ) and BR (lab test) was revealed the regression relationship with high satisfactory R 2 (Figure 6).So, the relationship might be allowed to predict soil MR (time-and labor-consuming procedure) by soil BR measurement.A principal component analysis showed that the first and second axes (components) produce 47 and 22% of the experimental data variation, respectively (Figure 7).The highest correlation of the first axis was found with the content of Cmic, BR, CO2 emission, soil temperature and moisture.The highest correlation of the second axis was found with qCO2 value.The first axis can be considered as the gradient of ecosystem changes.The soils of undisturbed (natural) ecosystems are mainly collected in the right part of figure, and the soils of disturbed (anthropogenically transformed) ecosystems are located in the left part.This might be indicated a more "optimum" functioning of the soil microbial community in Chernozems of natural ecosystems.

Conclusion
Along a gradient of Chernozems ecosystems (steppe, forest, fallow, urban) the significant decrease of soil Cmic, BR and Cfungi / Corg ratio was found (by 2-4 times less), while the qCO2 value increased.It might be illustrated an "deterioration" of soil microbial community functioning under anthropogenic transformation of terrestrial ecosystems.Soil basal respiration (lab test) might be characterized soil microbial respiration in situ.

Figure 1 .
Figure 1.Design of soil microbial respiration measurement in field conditions

Figure 4 .
Figure 4.The portion of soil microbial respiration in total soil CO2 emission of Chernozems in different ecosystems (1 and 2 are numbers of measurement point, May, 2015)

Table 4 .
Soil organic carbon content (Corg), soil microbial biomass carbon (Cmic), Cmic portion in Corg, the highest inhibition of substrate-induced respiration (SIR) by streptomycin and cycloheximide both, fungi / bacteria (F / B) and Cfungi / Corg ratios in different ecosystems of typical Chernozems