Effects of defoliation and nitrogen on carbon dioxide (CO₂) emissions and microbial communities in soils of cherry tree orchards

Soil collection

Soil samples were obtained at three depths (0–10 cm [top soils], 10–30 cm [middle soils], and 30–60 cm [deep soils]) of the botanical garden test site at Tianshui Normal University (Tianshui, Gansu, China; 34°34′10″N and 105°41′47″E) in 2021. The test site was built in 2002, where the cherry rootstock was Gisela 5 and the cherry variety was Provence. The basic physical and chemical properties of top, middle, and deep soils were as follows: pH: 6.7, 7.8, and 8.1; total organic carbon: 14.74 g kg⁻¹, 12.54 g kg⁻¹, and 7.92 g kg⁻¹; total nitrogen: 0.74 g kg⁻¹, 0.69 g kg⁻¹, and 0.46 g kg⁻¹; available phosphorus: 4.93 mg kg⁻¹, 5.08 mg kg⁻¹, and 3.07 mg kg⁻¹; available potassium: 153.0 mg kg⁻¹, 148.0 mg kg⁻¹, and 100.6 mg kg⁻¹.

Experimental design

The incubation experiment was conducted as a complete factorial experiment of soils at three depths (top, middle, and deep soils) * two kinds of defoliation addition (no-defoliation [C₀] and defoliation [C_A, 1%]) * addition of three levels of nitrogen (0 mg kg⁻¹ [N₀], 90 mg kg⁻¹ [N_L], and 135 mg kg⁻¹ [N_H]) in three replicates with a fully randomized design. Total organic carbon was 451 g kg⁻¹ and total nitrogen was 12.47 g kg⁻¹ in defoliation.

Incubation tanks (1 L) were respectively added with 100 g soils (dry weight) at different depths for 7 days of pre-incubation at 25 °C (Mo et al., 2021). After that, soils were fully mixed with defoliation according to different treatments. Nitrogen and phosphate fertilizers were dissolved in distilled water and added to soils as a solution. Next, the soils were incubated at 25 °C with the moisture of 60% and the bulk density of 1.2 g cm⁻³ in the dark for 80 days.

Gas collection and soil sampling

For each treatment, three tanks were taken out at 1, 4, 13, and 80 days. The soils in the incubation tanks were stored at −20 °C for detecting microbial biomass carbon (MBC), soil enzyme activities, and microbial communities. Specifically, MBC was measured with the chloroform-fumigation extraction method (Vance, Brookes & Jenkinson, 1987). The activity of alkaline phosphatase was examined with the method of Tabatabai and Bremner (Tabatabai & Bremner, 1969) as described in a prior study (Miralles et al., 2012). The activity of cellulase was detected using the method of Xu and Zheng (Xu & Zheng, 1986). The activity of catalase was determined by titrating 0.1 mol L⁻¹ KMnO₄ (Guan, Zhang & Zhang, 1986).

DNA extraction and high-throughput sequencing

Total DNA was extracted from soils with the Fast DNA SPIN Kit for Soil and FastPrep-24 nucleic acid Extraction instrument (MP Biomedicals, Santa Ana, CA, USA) on the 4th day. The DNA was examined with 1% agarose gel electrophoresis, and its concentration was measured with NanoDrop 2000. The DNA was stored in a refrigerator at −20 °C for subsequent use. The stock solution of the DNA was diluted to about 5 mg/L as the template of polymerase chain reaction (PCR) amplification. Afterward, the 16S rRNA variable regions V3-V4 of bacteria were subjected to PCR amplification with the following primers: CCTAYGGGRBGCASCAG (341F) and GGACTACNNGGGTATCTAAT (806R). The PCR was conducted with a system of 50 µL, including 5 µL of 10 * buffer, 4 µL of dNTP, 0.5 µL of RTAQ (Takara), 1 µL of 10 µmol L⁻¹ each front and rear primers, 36.5 µL of ddH₂O, and 2 µL of template DNA, and the detection of each sample was repeated three times. The amplification conditions were as follows: pre-denaturation at 98 °C for 30 s, 30 cycles of chain disassembly at 98 °C for 10 s, annealing at 55 °C for 15 s, and extension at 72 °C for 1 min, and final extension at 72 °C for 10 min. The products obtained by three repeats of DNA amplification were mixed and tested with 1% agarose gel electrophoresis. The concentration of the obtained bacterial PCR products was measured with the PicoGreen kit. After the products were evenly mixed, DNA was purified and recovered with the DNA purification kit (TIANGEN Biotech, Beijing, China). Bacterial 16S rDNA was sequenced by Shenzhen Weicomeng Technology Group Co., Ltd. (Shenzhen, China) with the Novaseq 6000 PE 250 platform.

According to corresponding barcodes, the samples were subjected to paired reading, and then their barcode and primer sequence were removed. The reads at both ends were combined with the FLASH (V1.2.7) software. The QIIME data analysis package was utilized to remove low-quality raw sequences (length < 250 bp, ambiguous base “N”, and the average base quality score less than 20). The chimeric sequence was discarded with the MAARJAM database (https://maarjam.ut.ee/). The valid readings were assigned to operational taxon units or virtual taxa with an identity threshold of 97%, and the representative sequences were identified with SILVA (V128, http://www.arb-silva.de) and MAARJAM databases.

Calculation

CO₂ emissions (mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$) were calculated as per the titration results of hydrochloric acids. ${\displaystyle \begin{array}{c}C{O}_2=\frac{\left(V_0-V\right)\times C_{HCl}}{2}\times 12\times \frac{1}{m\left(1-a\right)}\times 1000.\hfill \end{array}}$

In this formula, V₀ was the volume of titration by standard hydrochloric acid during blank calibration (mL), while V represented the volume of titration by standard hydrochloric acid during samples (mL). C_HCl was the concentration of standard hydrochloric acid (1 mol L⁻¹), m was soil mass (g), and a was soil water content (%).

The CO₂ efflux rate (mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) was calculated with the following formula:

CO₂ efflux rate = CO₂ emission/t.

In this formula, t represented the day when the NaOH solution was placed in the incubation tanks.

The cumulative CO₂ emission (g CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$) was the cumulative CO₂ emission from each treatment over a given incubation time.

For a given incubation time (80 days), the priming index (PI) induced by exogenous substance addition was normalized to the proportion of added non-exogenous substances to cumulative CO₂ emissions based on the following formula:

Priming index (PI) = (CO_{2 add} − CO_{2 non-add})/ CO_{2 non-add}.

In this formula, CO_{2 add} represented cumulative CO₂ emissions after 80 days of defoliation and nitrogen treatment, and CO_{2 non-add} indicated cumulative CO₂ emissions after 80 days of treatment without defoliation and nitrogen.

The PI represented the intensity of the priming effect. The PI value of 1 suggested that the amount of organic carbon mineralization was not affected by exogenous substance addition. The PI value of >1 indicated that the addition of exogenous substances caused the priming effect of organic carbon mineralization, and the larger value was associated with a stronger priming effect. The PI value of <1 represented that the addition of exogenous substances reduced the mineralization of organic carbon and produced a negative priming effect, and the smaller value illustrated the stronger negative priming effect.

Statistical analysis

All statistical analyses were performed with the SPSS 18.0 statistical and R programming software. The effects of the CO₂ efflux rate, cumulative CO₂ emission, PI, MBC, catalase, alkaline phosphatase, cellulase, Chao1, Shannon, and Simpson and the relative abundance of Proteobacteria and Acidobacteria were analyzed with three-way analysis of variance combined with Duncan’s multiple range test. Partial least squares discrimination analysis was conducted to analyze the distribution of soil microbial communities in soils at different depths. The Spearman correlation analysis was used to clarify the correlations of defoliation, nitrogen, and soil depth with CO₂ emissions, soil microbial activity parameters, Proteobacteria and Acidobacteria. Significance was defined at P < 0.05. Additionally, Origin18.0 and R programming software were utilized for mapping.

CO₂ emissions

The CO₂ efflux rate and cumulative CO₂ emissions were increased under defoliation addition, which was extremely significantly affected by the soil depth, defoliation, and nitrogen, as well as interaction among these three factors (Figs. 1 and 2, and Table 1).

The CO₂ efflux rate peaked on the 1st day and then declined with the incubation time of each treatment (Fig. 1). Among all treatments, the CO₂ efflux rate was higher in C_A treatment (65.41 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) than in C₀ treatment (26.42 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) and lower in deep soils (36.1 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) than middle (51.56 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) and top (50.06 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) soils (Fig. 1). In addition, N₀ treatment (52.87 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) resulted in a higher CO₂ efflux rate than N_L treatment (51.56 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) and N_H (50.06 mg CO₂ ${\mathrm{kg}}_{\mathrm{soil}}^{-1}$ d⁻¹) on the 1st day.

After 80 days (Fig. 2), cumulative CO₂ emissions were averagely elevated by 2.62, 2.37, or 2.26 times under treatments of C_AN₀, C_AN_L, or C_AN_H when compared to under C₀N₀ treatment. Cumulative CO₂ emissions were 30.28% and 28.21% higher in top and middle soils than in deep soils, respectively. Meanwhile, cumulative CO₂ emissions were markedly lower under N_L and N_H treatment than under N₀ treatment by 7.87% and 15.75%, respectively.

PI

After 80 days of incubation, defoliation addition alone or in combination with nitrogen addition resulted in positive PI in soils at the three depths (Fig. 3). PI was extremely substantially affected by defoliation, nitrogen, and soil depth (Table 1). PI was the highest in deep soils (2.05 [C_AN₀], 1.83 [C_AN_L], and 1.76 [C_AN_H]) in defoliation and nitrogen additions. Moreover, nitrogen addition alone caused reverse PI in soils at three depths.

MBC

MBC was enhanced under defoliation and nitrogen addition, which was significantly or extremely significantly altered by defoliation, nitrogen and soil depth (Fig. 4 and Table 1). Moreover, defoliation addition alone or combined with nitrogen addition contributed to higher MBC than C₀N₀ treatment. On the 4th day, MBC in top soils was markedly higher under C_AN₀, C_AN_L, and C_AN_H treatment than under C₀N₀ treatment by 66.49%, 87.82%, and 97.81%. In contrast, nitrogen addition alone diminished MBC in soils at the three depths. MBC in top and middle soils was higher than that in deep soils by 127.07% and 104.31%, respectively.

Soil enzymes

The activity of soil catalase, alkaline phosphatase, and cellulase was enhanced, which was markedly changed by defoliation, nitrogen and soil depth, on the 4th day (Fig. 5 and Table 1). The activity of soil catalase was 32.95% and 36.94% higher, that of alkaline phosphatase was 84.64% and 83.03% higher, and that of cellulase was 174.67% and 251.12% higher under C_AN_L and C_AN_H treatment than under C₀N₀. Nitrogen addition alone had different influences on activity of the tested soil enzymes in soils at the three depths. The activity of catalase, alkaline phosphatase, and cellulase was higher in top soils than in deep soils.

Soil bacterial diversity and community structure

During the first 7 days, CO₂ emissions rapidly surged after exogenous carbon addition. Mounting studies (Tian et al., 2016a; Shen et al., 2021; Zhou et al., 2021) reported that bacterial communities were changed from day 3 to day 15. Moreover, our study demonstrated that MBC peaked on the 4th day under defoliation and nitrogen addition. Accordingly, in order to evaluate the effect of defoliation, nitrogen, and soil depth on the soil microbial community structure in dryland cherry orchards, soil DNA was extracted on the 4th day of incubation to analyze the bacterial diversity. The results manifested that defoliation and nitrogen addition exerted no significant effects on the Chao1, Shannon, and Simpson index of soil bacteria (Fig. 6). However, the relative abundance of Proteobacteria markedly increased in soils at the three depths under defoliation and nitrogen addition, and that of Acidobacteria in deep soils was substantially reduced under both defoliation and nitrogen addition.

In our research, 48 phyla, 133 classes, 209 orders, 260 families, and 480 genera of bacteria were obtained through high-throughput sequencing of bacterial 16S rRNA. At the level of bacteria phyla (Fig. 7), the relative abundance of Proteobacteria and Acidobacteria was greater than 10.0% in communities, which were 37.84%–60.62% and 10.70%–24.53%, respectively. The relative abundance of Proteobacteria was 15.91% higher under C_A treatment than under C₀ treatment and was 12.18% and 11.85% higher under N_H and N_L treatment than under N₀ treatment in deep soils. The relative abundance of Acidobacteria in deep soils was lower under C_AN_H treatment than under C₀N₀ treatment by 96.23%.

Effects of defoliation and nitrogen addition on CO₂ emissions in soils at different depths

As previously described (Fontaine, Mariotti & Abbadie, 2003), defoliation alone or both defoliation and nitrogen was added in our study to assess organsim matter decomposition and analyze the mechanism of “microbial mining”. Our data unveiled that cumulative CO₂ emissions under defoliation addition alone or combined with nitrogen addition were 1.94−2.78 times higher than those under C₀N₀ treatment, which may be attributable to the fact that defoliation and nitrogen addition increased the activity of soil microbes (Figs. 4 and 5), accelerated the conversion of microbial biomass (Blagodatskaya & Kuzyakov, 2008), and facilitated the secretion of extracellular enzymes to decompose soil organsim matter and then deciduous residues through microbes (Allison et al., 2010; Burns et al., 2013), thus resulting in CO₂ emissions in short-term incubation (Cotrufo et al., 2015; Liu et al., 2017; Zhou et al., 2021). These results are consistent with hypothesis (1). Additionally, our findings also elucidated that compared to defoliation addition alone, addition of both defoliation and nitrogen decreased CO₂ emissions and PI, which is ascribed to the “mining of nitrogen” (Fontaine, Mariotti & Abbadie, 2003). Moreover, nitrogen alone addition also reduced CO₂ emissions and caused negative PI, which is attributed to the “mining of nitrogen” (Fontaine, Mariotti & Abbadie, 2003). Our data also revealed that in soils with relatively low total nitrogen contents (0.47−0.74 g kg⁻¹), microbes were stimulated to decompose organsim matter to acquire the required nitrogen and induce CO₂ emissions under defoliation and nitrogen addition, similar to most research results (Chen et al., 2014; Li et al., 2017; Fang et al., 2018a; Mason-Jones, Schmücker & Kuzyakov, 2018; Hicks et al., 2019; Liao, Tian & Liu, 2021). However, our data showed no significant differences between low and high nitrogen, which indicated that the nitrogen demand of microbes could be saturated at low nitrogen levels. In addition, nitrogen addition alone reduced CO₂ emissions and then caused negative PI, which was concordant with the results of laboratory culture in the dark (Wang et al., 2014; Fang et al., 2018a) and outdoor testing (Ginting et al., 2003). Corresponding to the research with the DeNitrification-DeComposition (DNDC) model (Chi & Chen, 2001), Grant et al. (2004) also used the DNDC model to predict the impact of 50% CO₂ emissions on nitrogen. Therefore, nitrogen addition alone exerts little effect on soil respiration, whilst combination of nitrogen and organic carbon effectively promotes the decomposition of exogenous organic substances.

Soil depth is frequently associated with new increases in CO₂ emissions (Meyer et al., 2018; Liao et al., 2020). Throughout the incubation period, the CO₂ efflux rates and cumulative CO₂ emissions were higher in top and middle soils than in deep soils, which might be explained by the following factors: (i) high MBC and activities of soil catalase, alkaline phosphatase, and cellulase in top and middle soils (Figs. 4 and 5), (ii) relatively sufficient nutrients in top and middle soils, and (iii) suitable pH and favorable soil structure and properties in top and middle soils (Wang et al., 2014; Tian et al., 2016b), which provided advantageous conditions for the decomposition of deciduous residues by microbes. Conversely, PI was higher in deep soils than in top and middle soils in our study, concurrent with the result of a prior study (Liao et al., 2020). The proportion of residual organic carbon was calculated based on the input amount of organic carbon and emission amount of CO₂ in soils at different depths, which exhibited that the proportion of residual organic carbon was 88.88%, 88.05%, and 80.60% in top, middle, and deep soils, respectively. This result indicated that top and middle soils can retain a higher proportion of exogenous carbon than deep soils.

Effects of defoliation and nitrogen addition on microbial activities and communities in different soils

Microbes have been extensively recognized to play a key role in soil carbon mineralization (Herrmann & Bucksch, 2014; Chen et al., 2018; Li et al., 2018). MBC and enzyme activities are usually considered measurable proxies for microbial decomposition (Dorodnikov et al., 2009; Jiang et al., 2021) and can be optimized with the shortest incubation time to minimize microbial growth and enzyme production during the measurement (Shen et al., 2021). Defoliation addition alone or combined with nitrogen addition elevated MBC, increased the activity of alkaline phosphatase, and cellulase, and caused positive PI in soils (Fig. 3). Defoliation shared significantly positive correlations with MBC and the activity of catalase, alkaline phosphatase, and cellulase in soils (Fig. 8). These data illustrated that microbes obtained available carbon and nitrogen from defoliation and nitrogen through various enzymes to meet their stoichiometric requirements (Lashermes et al., 2016). In our study, nitrogen increased microbial biomass and the activity of the tested enzymes in defoliation and nitrogen additions (Figs. 4 and 5) as a regulator in organic carbon mineralization, which was supported by many previous studies (Frey et al., 2004; Cayuela, Sinicco & Mondini, 2009; Parajuli, Ye & Szogi, 2022). We also observed that at the initial stage of culture, addition of both defoliation and nitrogen (especially both defoliation and high nitrogen) markedly enhanced MBC and enzyme activities but diminished CO₂ emissions when compared with defoliation addition alone, further confirming the “microbial mining” mechanism (Fontaine, Mariotti & Abbadie, 2003). Meanwhile, our results also unraveled that only nitrogen addition significantly decreased soil microbial biomass and changed the activity of the tested enzymes at the three depths, accompanied by reduced CO₂ emissions and negative PI. These findings illustrated that the effective combination of nitrogen and organic carbon could accelerate organism decomposition, improve soil quality and fertility, and provide guidance for production practice.

In our study, Proteobacteria and Acidobacteria were dominant bacteria in soils regardless of soil depth or exogenous inputs (Fig. 7). Defoliation and nitrogen addition did not significantly change soil bacterial diversity and abundance (Fig. 6), indicating the ecological properties of Proteobacteria and Acidobacteria in soils. Defoliation addition alone or combined with nitrogen addition markedly elevated the relative abundance of Proteobacteria and reduced the relative abundance of Acidobacteria, thereby causing high CO₂ emissions. This result suggested the association of Proteobacteria and Acidobacteria with CO₂ emissions. As a copiotrophic group, Proteobacteria grows rapidly, and its abundance is increased by relying on more labile carbon sources under nutrient inputs (Ramirez, Craine & Fierer, 2012). With increasing stimulation, more Proteobacteria can participate in the synthesis of exoenzymes to mineralize substrates (Schimel, 2003; Fierer, Bradford & Jackson, 2007), corresponding to the result that defoliation addition alone or combined with nitrogen addition substantially augmented MBC and enzyme activities, which was consistent with hypothesis (2). Our data revealed that CO₂ emissions were significantly positively correlated with Proteobacteria (Fig. 8). In addition, Nottingham et al. (2009) observed that microbial biomass increased in the first step of plant residue decomposition, which was caused by labile C rather than macromolecular compounds. Altogether, these data suggest Proteobacteria as a main participant in organic carbon mineralization in cherry orchards. However, Acidobacteria (an oligotrophic group) could be colonized on mineral surfaces under rich nutrient conditions (Nemergut et al., 2010; Ramirez, Craine & Fierer, 2012).

Furthermore, our PLS-DA analysis (Fig. 9) demonstrated that bacterial communities were clearly separated in top and deep soils, which can be explained by the fact that nutrient composition determines the distribution of oligotrophic and hypertrophic bacteria (Fierer, Bradford & Jackson, 2007; Whitman et al., 2016), consistent with hypothesis (3). Compared to top soils, deep soils had the highest relative abundance of Proteobacteria under both defoliation and nitrogen addition, which can be attributed to the highest PI in deep soils that is related to different microbial products (Kögel-Knabner, 2017), soil pH (Madsen & Munk, 1987; Silveira et al., 2008; Takele, Chimdi & Abebaw, 2014), and soil nutrient (Madsen & Munk, 1987). Therefore, soil microbial communities may be partially determined by soil properties. Furthermore, fungi have also been reported to directly participate in organic carbon mineralization. Accordingly, further research is warranted to determine how fungal communities respond to defoliation and nitrogen added in soils of cherry orchards in northwest China.