Soil aggregate size influences the impact of inorganic nitrogen deposition on soil nitrification in an alpine meadow of the Qinghai–Tibet Plateau

Soil sampling and sieving of aggregates

The soil was collected in an alpine meadow located at the Haibei Demonstration Zone of the Plateau Modern Ecological Animal Husbandry Science and Technology (36°55′N, 100°57′E) in Qinghai Province, China, at an altitude of 3,040 m. The mean annual temperature and precipitation of this site are −0.45 °C and 400 mm, respectively. The minimum monthly mean air temperature is −29 °C in January, with the maximum of 27 °C occurring in July. The dominant plant species belong to the Gramineae family and include Elymus dahuricus and Stipa capillata. The soil underlying the site is a clay–loam classified as Mat–Gryic Cambisol (Ma et al., 2017b).

In June of 2017, soil cores (0–15 cm) were collected at random from a natural alpine meadow with an area of about 50 × 50 m that was not subject to any management or use practices. We set up three sampling points along the diagonal of the selected area as three replicates of field soil sampling. At each sampling point, we randomly collected soil cores (5 cm in diameter) and took approximately 50 kg of bulk soil. We then transported samples to the laboratory where plant roots and fine stones were carefully removed by hand and soils were sieved to the different soil aggregate sizes required for the experiment. The three field sampling points corresponded to the three replicates in the laboratory. Four aggregate-size classes were obtained by dry sieving 100 g of fresh soil through a series of four sieves (4, 2, 1, and 0.25 mm) as follows: large macro-aggregates (2–4 mm), macro-aggregates (1–2 mm), meso-aggregates (0.25–1 mm), and micro-aggregates (<0.25 mm). Soil was placed on a four mm sieve, then manually moved up and down by 10 cm 60 times during a period of 2 min. The material passing through the four mm sieve was then transferred to the next smaller-sized sieve (two mm) for further fractionation, ultimately generating four aggregate fractions (Jiang et al., 2015; Yang, Liu & Zhang, 2017). This process was repeated until the amount of each soil aggregate size fraction required was obtained.

Determination of physicochemical properties for soil aggregates

Soil organic carbon (SOC) was determined using an auto-analyzer (TOC, Elementar, Germany). Ten gram of fresh soil was extracted with 50 ml of 2 M KCl to measure soil NH₄⁺-N and NO₃⁻-N using a flow-solution analyzer (Flowsys, Ecotech, Germany). Soil pH was measured using a pH meter after shaking a 1:2.5 air-dried soil/water suspension for 30 min. SP was calculated from the bulk density (BD) and the particle density (2.65 g cm⁻³) using the following equation (Munkholm et al., 2016): $$\mathrm{S}\mathrm{o}\mathrm{i}\mathrm{l}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}=\left(1-\mathrm{S}\mathrm{o}\mathrm{i}\mathrm{l}\mathrm{b}\mathrm{u}\mathrm{l}\mathrm{k}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}/2.65\right)\times 100\mathrm{\%}$$

Soil BD was determined using oven-dried soils (Regelink et al., 2015). Briefly, we placed three replicates of 500 g aggregates in 1,000 ml jars. We then adjusted the moisture content to 30%, which was the maximum field water capacity of the soil. After allowing the samples to settle for 1 day, a foil sampler with a volume of 100 cm³ was used to obtain the samples. This was followed by drying at 105 °C for 1 day.

N addition and soil aggregate incubation

Soil aggregate samples (200 g) of four classes were placed at the bottom of 1,000 ml plastic bottles. Polyethylene film punctured with needle holes was then placed on all bottles to maintain aerobic conditions. The soil moisture content was adjusted to 60% field moisture capacity, and the bottles were pre-incubated at 25 °C for 7 days. Our experiment employed a two-factor design that consisted of four levels of aggregate sizes (2–4, 1–2, 0.25–1, and <0.25 mm) and two forms of nitrogen (N) addition (i.e., NH₄⁺-N and NO₃⁻-N). Two forms of N were applied as NH₄Cl and Ca (NO₃)₂, which were added to the four levels of soil aggregates to give a gradient of 0, 5, and 10 mg N kg⁻¹ soil. Each treatment had three replicates. After incubation for 0, 7, 14, 21 and 28 days, 10 g of wet soil was collected from three replicate bottles of each treatment. The NO₃⁻-N was then extracted with 50 ml of 2 M KCl, after which the filtrate was used to determine the NO₃⁻-N concentrations.

Data calculations and analysis

The net nitrification rate N_r (mg NO₃⁻-N kg⁻¹ aggregate day⁻¹) was calculated from the equation below (Xin et al., 2014): $$N_{\mathrm{r}}=\left[{\left({\mathrm{N}\mathrm{O}}_3^{-}\text{-}\mathrm{N}\right)}_a-{\left({\mathrm{N}\mathrm{O}}_3^{-}\text{-}\mathrm{N}\right)}_b\right]/T_{\mathrm{d}}$$ where a and b are the NO₃⁻-N concentrations measured after and before each incubation period, respectively, and T_d indicates the incubation time in days.

The contribution of each type of soil aggregates to the net nitrification rate of bulk soil was determined from the following equation (Yang, Liu & Zhang, 2017): $${\mathrm{C}}_{\mathrm{r}}={\mathrm{N}}_{\mathrm{r}}\times {\mathrm{A}}_{\mathrm{r}}$$ where C_r is the contribution rate (mg NO₃⁻-N kg⁻¹ soil day⁻¹), N_r is the nitrification rate observed during the fourth week and A_r is the aggregates proportion (%).

The homogeneity and normality of variances were verified for all data using the Levene and Kolmogorov–Smirnov tests, respectively. Repeated-measures analysis of variance (ANOVA) was employed to test the effects of the incubation time, soil aggregates, and N addition on N_r. One-way ANOVA was used to test the physicochemical properties among soil aggregate sizes and the differences of NO₃⁻-N concentrations, N_r and C_r among soil aggregate sizes or the differences of NO₃⁻-N concentrations, N_r and C_r under nitrogen addition treatment with different concentrations and forms. Following ANOVA, post hoc comparisons of the means were calculated using Tukey multiple comparison (P < 0.05). Pearson’s correlation was used to determine the correlation between soil physicochemical properties and N_r for all nitrogen addition treatments. All statistical analyses were performed using the SPSS statistical package (version 19.0; IBM, Armonk, NY, USA) and figures were obtained using SigmaPlot 12.5.

Physicochemical properties of soil aggregates

The SOC, NH₄⁺-N, and NO₃⁻-N concentrations were significantly different among the aggregate sizes (Table 1). The level of SOC associated with the <0.25 mm aggregates was significantly higher than 2–4 and 0.25–1 mm soil aggregates. The NH₄⁺-N concentration in the 2–4 mm aggregates was significantly higher than in the other soil aggregates sizes, which did not significantly differ in their concentration. The NH₄⁺-N concentration in the 2–4 mm aggregates was 14.1% higher than that in the 0.25–1 mm aggregates. The NO₃⁻-N concentrations in the 2–4 mm and 1–2 mm aggregates were significantly higher than in the 0.25–1 mm and <0.25 mm aggregates. The soil pH was alkaline for all the soil aggregates sizes, varying between 8.67 and 8.93. Soil pH was significantly lower in 1–2 mm and <0.25 mm aggregates than in aggregates with other sizes. The BD of <0.25 and 0.25–1 mm aggregates were significantly higher than for the other soil aggregates sizes, and the BD of the 2–4 mm aggregate fraction was the lowest (0.78 g cm⁻³; which was 28.4% lower than the BD of <0.25 mm aggregates). The 2–4 mm aggregates had the highest porosity (70.48%), which was significantly higher than in the other soil aggregate sizes.

Response of NO₃⁻-N concentration to soil aggregate sizes and N addition

Under no N addition and the addition of 5 mg NH₄⁺ kg^-1 soil, as the incubation time increased, the NO₃⁻-N concentration in the 2–4 mm, 1–2 mm and <0.25 mm aggregates increased, while it first decreased and then increased in the 0.25–1 mm aggregates (Figs. 1A and 1B). Following the addition of 10 mg NH₄⁺ kg⁻¹ soil, the NO₃⁻-N concentration in the four soil aggregate size fractions increased as the incubation time increased (Fig. 1C). Under the addition of 5 mg and 10 mg NO₃⁻ kg⁻¹ soil, as the incubation time increased, the NO₃⁻-N concentration in the 2–4 mm, 1–2 mm and <0.25 mm aggregates increased (Figs. 1D and 1E). For the 0.25–1 mm aggregates, the NO₃⁻-N concentration first decreased, then increased under the addition of 5 mg NO₃⁻ kg⁻¹ soil. However, it decreased under the addition of 10 mg NO₃⁻ kg⁻¹ soil as the incubation time increased (Figs. 1D and 1E). The largest NO₃⁻-N concentration was observed for the 2–4 mm aggregates, while the lowest was found in the 0.25–1 mm aggregates under both forms of N addition and rates (Figs. S1 and S2). Under the addition of 10 mg NH₄⁺ kg⁻¹ soil and 10 mg NO₃⁻ kg⁻¹ soil, the NO₃⁻-N concentration in the 2–4 mm aggregates was significantly higher than that of the other aggregates on day 7, 14, 21, and 28 (Figs. S1B–S1E and S2B–S2E, P < 0.05). For the 2–4 mm aggregates, the NO₃⁻-N concentration was significantly higher under the addition of 10 mg NO₃⁻ kg⁻¹ soil than the addition of 0 and 5 NO₃⁻ kg⁻¹ soil on day 7, 21, and 28 (Figs. S2B, S2D and S2E, P < 0.05).

Response of N_r to soil aggregate sizes and N addition

The rate of nitrification fluctuated greatly during the first and second weeks of incubation, especially in 0.25–1 mm and 1–2 mm aggregates (Figs. 2A–2D). After the 3rd week of incubation, the change of nitrification between aggregate sizes tended to be stable (Figs. 2E and 2F). During the 4 week incubation, nitrification rate in the 0.25–1 mm aggregates showed a consistent change with the addition of 10 mg NO₃⁻ kg⁻¹ soil, all of which were negative, but increased with the incubation time. Especially for the change in nitrification rate after 4 weeks of incubation, nitrification patterns associated with different sizes of aggregates were similar under the two N addition treatments, except for the 2–4 mm aggregates where NH₄⁺ addition showed the highest N_r (Figs. 2G and 2H). The N_r value during the 4 week incubation period differed between aggregate sizes, with the 2–4 mm aggregates having significantly higher N_r than other aggregate sizes at the addition rate of 10 mg NH₄⁺ kg⁻¹ soil and 10 mg NO₃⁻ kg⁻¹ soil (Fig. 2H, P < 0.05). The largest N_r value (0.75 and 0.46 mg kg⁻¹ day⁻¹ for 10 mg NH₄⁺ kg⁻¹ soil NH₄⁺ and 5 mg NO₃⁻ kg⁻¹ soil NO₃⁻ addition, respectively) was observed for large macro-aggregates (2–4 mm), and the lowest for meso-aggregates (0.25–1 mm) (0.26 and −0.085 mg kg⁻¹ day⁻¹ for 10 mg NH₄⁺ kg⁻¹ soil NH₄⁺ and 10 mg NO₃⁻ kg⁻¹ soil NO₃⁻ addition, respectively). The N_r was significantly higher under the addition of 10 mg NH₄⁺ kg⁻¹ soil than in untreated soil, except for the 1–2 mm aggregates (Fig. 2H, P < 0.05). N_r was negative in the 0.25–1 mm aggregates under the addition of 10 mg NO₃⁻ kg⁻¹ soil.

The incubation period, soil aggregate size, nitrogen addition, and interactions of these variables had significant impacts on the N_r (Table S1, P = 0.005 and P < 0.0001). NH₄⁺-N in soil aggregates was positively correlated with the soil N_r, and the porosity of soil aggregates was positively correlated with the soil N_r under all N addition treatments (Table 2).

Contribution of different types of aggregates to the N_r in bulk soil

The proportion of the aggregate sizes in the bulk soil was 10.48, 12.21, 28.35 and 48.73% for the 2–4 mm, 1–2 mm, 0.25–1 mm, and <0.25 mm aggregates, respectively (Table 1). The <0.25 mm aggregates had a significantly higher proportion than the other aggregate sizes (P < 0.05). The 2–4 mm aggregates had higher soil N_r, but the soil aggregate proportion for the 2–4 mm aggregates was low. Conversely, the 0.25–1 mm and <0.25 mm aggregates had lower soil N_r, but their soil aggregates proportions were high. For the 2–4 mm aggregates, C_r was significantly higher with the addition of 10 mg NH₄⁺ kg⁻¹ soil than with the addition of 0 and 5 mg NH₄⁺ kg⁻¹ soil (Fig. 3A, P < 0.05). However, for the 0.25–1 mm aggregates, C_r was significantly lower with the addition of 10 mg NO₃⁻ kg⁻¹ soil than with the addition of 0 and 5 mg NO₃⁻ kg⁻¹ soil (Fig. 3B, P < 0.05).

Variability of nitrification associated with soil aggregate sizes

Aggregate sizes had a significant effect on the nitrification rate (Jiang et al., 2011). In the present study, higher nitrification rates were found in 2–4 mm aggregates than in the other three aggregates under both N addition treatments. Muruganandam, Israel & Robarge (2010) found that a higher nitrification rate associated with 0.5–1 mm aggregates than 2–4 and <0.25 mm aggregates. In addition, Jiang et al. (2011) reported that nitrification rates were higher for the 2–0.25 mm fraction than the 0.25–0.053 mm fraction. We further separated the 0.25–2 mm aggregates into 0.25–1 and 1–2 mm aggregates and found higher nitrification rates in the 1–2 mm aggregates. However, a study of N turnover associated with aggregates in wetland soils revealed that the >2 mm aggregate fraction played a vital role in the release of inorganic N (Song et al., 2017), which was consistent with our results. Across different aggregate fractions, the heterogeneous distribution of microclimatic conditions and substrates of different qualities influence microbial community structure and nitrification rates (Bach & Hofmockel, 2014; Blaud et al., 2017; Davinic et al., 2012; Muruganandam, Israel & Robarge, 2010). Nitrification rates had no correlation with SOC. This may be due to the fact that the nitrification process was dominated by autotrophic microorganisms in our experiments. However, the NH₄⁺-N in soil aggregates was positively correlated with the soil nitrification rate, which means that high concentration of NH₄⁺ might improve the soil nitrification capacity (Zhao, Cai & Xu, 2007). In addition, there was a significant positive correlation between the porosity of soil aggregates and the nitrification rate, which may be a factor affecting the rate of nitrification. Specifically, a higher SP results in a higher specific surface area and therefore higher aerobic microbial biomass and activity. Oxygen availability is one of the most important factors controlling the rate of nitrification in soil (Princic et al., 1998). Aerobic microorganisms associated with nitrification are dependent on the presence of oxygen (Ke, Lu & Conrad, 2015). Because the rate of O₂ diffusion is related to the aggregate radius, O₂ concentrations are different in various aggregate sizes (Sexstone et al., 1985), and thus differentially affect the nitrification process. The 2–4 mm aggregates was the most active, which might be due to the high levels of SP allowing sufficient oxygen supply (Sexstone et al., 1985), and this in turn influences activities and compositions of soil microbial communities (Blackwood et al., 2006). In addition, our previous study has found that the abundance of bacteria in 2–4 mm and 1–2 mm aggregates were much higher than that in 0.25–1 mm and <0.25 mm aggregates (Yang, Liu & Zhang, 2019). We found that there was a significant positive correlation between nitrification rate and bacterial abundance of soil aggregates with different sizes under the treatment of 5 mg NH₄⁺ addition kg⁻¹ aggregate (Fig. S3). Our results suggest that the 2–4 mm and 1–2 mm aggregates could have a dominant influence on the response of soil N transformation processes to future nitrogen deposition in the studied alpine meadow.

Relationship of nitrification rate with nitrogen addition

The nitrification rate of different aggregate fractions responded differently to two forms of N addition. Specifically, nitrification rate under the 10 mg NH₄⁺ treatment was greater than that under the 10 mg NO₃⁻ treatment across all soil aggregate fractions, which is in agreement with the results of previous studies (Jiang, Jin & Sun, 2014; Zhang et al., 2017). NH₄⁺ concentration is important to the determination of ammonia oxidizer growth and whether oxidization will be conducted by bacteria or archaea (Verhamme, Prosser & Nicol, 2011). However, NO₃⁻ is another primary inorganic N form in soils, and the impact of NO₃⁻ on N_r is likely to be based on the product inhibition mechanism (Ying et al., 2017), which means that excessive concentration of the final product (NO₃⁻) could inhibit the activity of nitrosating bacteria and nitrifying bacteria, thereby affecting soil nitrification capacity (Painter, 1977). In the present study, N_r was negative in the 0.25–1 mm aggregates when treated with 10 mg NO₃⁻ kg⁻¹ soil. This may have been caused by a product inhibition mechanism. On the one hand, the SP of the 0.25–1 mm aggregates was small and oxygen diffusion might be limited; therefore, the microorganisms in the aggregates may have been in an anoxic state (Ebrahimi & Or, 2016; Schlueter et al., 2018). On the other hand, a large amount of product was accumulated, which may have enabled denitrification (Su et al., 2019). As a result, the net nitrification rate of the 0.25–1 mm aggregates was negative under high NO₃⁻-N addition. Deposition of NO₃⁻-N has become increasingly important to the total N deposition (Zhao et al., 2009). Neumann et al. (2013) investigated the effects of fertilization on microbial communities in soil aggregates, and found that fertilizations increased the abundance of microbial communities in the larger-sized fractions than in fine silt. In our study, we found that, the soil nitrification rate in 2–4 mm aggregates was larger after N addition. Although we have not characterized the microbial community, our study also showed that the large aggregate size responds more strongly to fertilization. Blaud et al. (2017) found that different sieving methods affect the observed bacterial diversity and abundance. The dry sieving method used in our study can lead to different results as compared to the wet sieving method. Blaud et al. (2018) compared the abundance of N cycling genes in different land use and soil aggregates sizes, and found land use patterns had a significant impact on the abundances for all genes, while the effect of soil aggregates was relatively small. Our study found that there are differences in nitrification rates between different soil aggregates, and we suspect that their microbial distribution may also differ. The results of the present study showed that the soil nitrification in samples amended with 10 mg N kg⁻¹ soil was greater than that of those treated with 5 mg N kg⁻¹ soil. For the 0.25–1 mm aggregates, the NO₃⁻-N concentrations in the 1st week were relatively stable or reduced in response to treatment with both forms of N, especially when 10 mg N kg⁻¹ soil was added. These findings indicated that the nitrification ability differed among the soil aggregates of different sizes, which could be due to the SP and oxygen content. In macro-aggregates, the NO₃⁻-N concentration increased linearly with the incubation time, indicating a stronger nitrification ability. In micro-aggregates, the NO₃⁻-N concentration was exponentially related to the incubation time, presenting a stable or decreased nitrification ability.

Contribution of aggregates to the nitrification rate in bulk soil

In this study, dry sieving revealed that the <0.25 mm aggregates was significantly higher than that of the other aggregates, representing 48.73% of the aggregate distribution. A previous study showed that soil fractions <0.25 mm represented only 2–20% of the aggregate distribution (Blaud et al., 2017). These findings indicated that different soils have different aggregate distributions, even when the same sieving method is used, so their contribution to nitrification will also differ. Fernández et al. (2010) compared the effects of no-till and conventional tillage on carbon contents and respiration rates for different aggregate size fractions, and they found that intermediate size aggregates showed the highest difference of C contents and the highest amount of respired C. This would indicate that C losses from soil through mineralization are mostly associated with intermediate aggregate size. Our previous research found that 0.25–1 and <0.25 mm aggregates had higher contribution rates to bulk soil SOC mineralization than 2–4 mm and 1–2 mm aggregates (Yang, Liu & Zhang, 2017). In our study, the 2–4 mm aggregates was the least represented, but had the highest nitrification rates. These findings imply that soil nitrification was mostly associated with the 2–4 mm aggregates. Almost half of the soil was composed of <0.25 mm aggregates, which resulted in higher contribution rates to the bulk soil nitrification rate than the other three aggregate sizes, despite the lower nitrification rate of the <0.25 mm aggregates.