Responses of ecosystem nitrogen cycle to nitrogen addition: a meta‐analysis

New PhytologistVolume 189, Issue 4 p. 1040-1050 Full paperFree Access Responses of ecosystem nitrogen cycle to nitrogen addition: a meta-analysis Meng Lu, Meng Lu Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, China Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USA Department of Environmental Science and Engineering, Fudan University, 220 Handan Road, Shanghai 200433, China These authors contributed equally to this work.Search for more papers by this authorYuanhe Yang, Yuanhe Yang Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USA These authors contributed equally to this work.Search for more papers by this authorYiqi Luo, Yiqi Luo Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, China Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USASearch for more papers by this authorChangming Fang, Changming Fang Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this authorXuhui Zhou, Xuhui Zhou Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USASearch for more papers by this authorJiakuan Chen, Jiakuan Chen Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this authorXin Yang, Xin Yang Department of Environmental Science and Engineering, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this authorBo Li, Bo Li Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this author Meng Lu, Meng Lu Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, China Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USA Department of Environmental Science and Engineering, Fudan University, 220 Handan Road, Shanghai 200433, China These authors contributed equally to this work.Search for more papers by this authorYuanhe Yang, Yuanhe Yang Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USA These authors contributed equally to this work.Search for more papers by this authorYiqi Luo, Yiqi Luo Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, China Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USASearch for more papers by this authorChangming Fang, Changming Fang Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this authorXuhui Zhou, Xuhui Zhou Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USASearch for more papers by this authorJiakuan Chen, Jiakuan Chen Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this authorXin Yang, Xin Yang Department of Environmental Science and Engineering, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this authorBo Li, Bo Li Coastal Ecosystems Research Station of the Yangtze River Estuary, Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering, The Institute of Biodiversity Science, Fudan University, 220 Handan Road, Shanghai 200433, ChinaSearch for more papers by this author First published: 07 December 2010 https://doi.org/10.1111/j.1469-8137.2010.03563.xCitations: 290 Authors for correspondence:Yiqi LuoTel: +1 405 325 8578Email: yluo@ou.eduBo LiTel: +86 21 65642178Email: bool@fudan.edu.cn AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Summary • Anthropogenic nitrogen (N) addition may substantially alter the terrestrial N cycle. However, a comprehensive understanding of how the ecosystem N cycle responds to external N input remains elusive. • Here, we evaluated the central tendencies of the responses of 15 variables associated with the ecosystem N cycle to N addition, using data extracted from 206 peer-reviewed papers. • Our results showed that the largest changes in the ecosystem N cycle caused by N addition were increases in soil inorganic N leaching (461%), soil NO3− concentration (429%), nitrification (154%), nitrous oxide emission (134%), and denitrification (84%). N addition also substantially increased soil NH4+ concentration (47%), and the N content in belowground (53%) and aboveground (44%) plant pools, leaves (24%), litter (24%) and dissolved organic N (21%). Total N content in the organic horizon (6.1%) and mineral soil (6.2%) slightly increased in response to N addition. However, N addition induced a decrease in microbial biomass N by 5.8%. • The increases in N effluxes caused by N addition were much greater than those in plant and soil pools except soil NO3−, suggesting a leaky terrestrial N system. Introduction Humans have approximately doubled the input of reactive nitrogen (N) to the Earth’s land surface (Galloway et al., 2008; Gruber & Galloway, 2008; Schlesinger, 2009). The increase of anthropogenic reactive N emissions via agricultural fertilization and combustion of fossil fuel has induced significant atmospheric N deposition, with an average rate of 105 Tg N yr−1 (Vitousek et al., 1997; Galloway et al., 2008). The enhanced N input may exert strong effects on both the structure (Clark & Tilman, 2008; Bobbink et al., 2010) and the functioning (Reay et al., 2008; Janssens et al., 2010) of terrestrial ecosystems. As the fundamental components of ecosystem functioning, terrestrial carbon (C) and N cycles may be sensitive to enhanced N deposition (Gruber & Galloway, 2008). It is well known that plant growth is usually constrained by soil N availability in most terrestrial ecosystems (Vitousek & Howarth, 1991; LeBauer & Treseder, 2008). However, it is still uncertain whether this N-induced stimulation of plant growth results in ecosystem C and N accumulation (Neff et al., 2002; Reay et al., 2008). Therefore, improved understanding of the responses of ecosystem C and N cycles to N addition is much needed to enable prediction of the effects of N fertilization and deposition on terrestrial ecosystems. A number of meta-analyses have examined the effects of additional N input on both C pools (e.g. Treseder, 2008; Xia & Wan, 2008; Liu & Greaver, 2010) and fluxes (e.g. Knorr et al., 2005; LeBauer & Treseder, 2008; Liu & Greaver, 2009; Janssens et al., 2010) in terrestrial ecosystems. However, little is known about how the ecosystem N pools and fluxes respond to atmospheric N deposition. The lack of a comprehensive understanding of the effects of N addition on the ecosystem N cycle greatly limits our ability to explore the responses of the ecosystem C cycle to N fertilization and deposition, as C and N cycles are coupled in terrestrial ecosystems (Gruber & Galloway, 2008). Therefore, to gain insights into the responses of ecosystem C and N cycles to additional N input, it is imperative to examine how the ecosystem N cycle responds to N fertilization and deposition. Numerous individual studies have been conducted to examine how the ecosystem N cycle responds to N fertilization or deposition. Previous studies have demonstrated that ecosystem N pools (Mack et al., 2004), microbial biomass N and enzyme activities (Ajwa et al., 1999), and nitric oxide (NO) and nitrous oxide (N2O) emissions (Butterbach-Bahl et al., 1997) can all be significantly influenced by the external N input. However, experimental results from various individual studies are highly variable, particularly for the soil N pool. For instance, the total N pool in mineral soil has been reported to exhibit an increase (Fisk & Schmidt, 1996), a decrease (Mack et al., 2004) or an insignificant change (Johnson et al., 2000) in response to external N input. Similarly, the responses of microbial biomass N and associated fluxes (i.e. N mineralization, nitrification, and denitrification) to N addition are also highly variable. For example, net N mineralization may increase (Brenner et al., 2005), decrease (Kowaljow & Mazzarino, 2007) or show minor changes (Riley, 1998) in response to N addition. Thus, a general pattern of the responses of N pools and fluxes to N fertilization and deposition is still unavailable. The highly diverse results from individual experiments are unlikely to reveal a general pattern that can be applied to various ecosystems. However, the results across individual studies can be synthesized to reveal a central tendency of changes in ecosystem N cycle induced by the additional N input (Hedges & Olkin, 1985). By compiling data from 206 individual studies, we conducted a meta-analysis to identify the central tendency of the effects of N addition on ecosystem N cycle. More specifically, this study aimed to investigate the responses of N pool sizes (including plant, litter and microbial biomass, organic horizon and mineral soil pools) to N addition; to explore the responses of N fluxes (i.e. net N mineralization, nitrification, denitrification and leaching) to the external N input; and to examine whether ecosystem types and other factors affect the responses of the ecosystem N cycle to N addition. Materials and Methods Data compilation We selected 206 papers from 2000 peer-reviewed publications that reported N dynamics in response to N fertilization (Supporting Information Notes S1). The compiled database included the responses to N addition of 15 variables related to N pools, fluxes, and other associated parameters (Table S1). The following five criteria were applied to select appropriate studies. (1) N fertilizers were directly added to terrestrial ecosystems and at least one of our selected variables was measured. (2) The treatment and control plots were started with the same plant species and soil types. For crop rotation experiments, the selected data were obtained using the same tillage management system, crop species and rotation sequences. (3) The N application rate, experimental duration and soil depth were clearly recorded and the measurements of treatment and control groups were conducted at the same temporal and spatial scales. (4) To investigate the long-term effect of N addition on the soil N pool, experiments shorter than 1 yr were excluded to avoid short-term noise. (5) The means, standard deviations or standard errors and sample sizes of the chosen variables were directly reported or could be indirectly calculated from the chosen papers. It should be noted that measurements for different N application rates were considered as independent observations if more than one level of N addition was applied in the same experiment (Curtis & Wang, 1998; Liu & Greaver, 2009). The latest sampling was used if more than one measurement at different temporal scales was available for the same experiment (Treseder, 2008; Liu & Greaver, 2009). The aboveground plant N pool was obtained from direct measurements of aboveground plant N content or indirectly calculated from aboveground plant biomass and N concentration. The belowground plant N pool was quantified using the reported root N content. The litter N pool was determined from the litter N content or the N content of returned residues in agricultural ecosystems, and direct measurements of litter N stock in nonagricultural ecosystems. The soil N pool was calculated for the organic horizon and mineral soil, respectively. The soil N concentration was also used to represent the soil N pool size because of the insignificant effects of N addition on soil bulk density (Fig. S2). To reveal the effects of N addition on ecosystem N fluxes, we extracted data from the studies that directly reported the average or cumulative net N mineralization, N immobilization, nitrification, denitrification, inorganic N and/or NO3− leaching, and N2O emission in response to external N input. In addition, data on mean annual temperature (MAT) and mean annual precipitation (MAP) at each study site were either extracted from the published papers or, if they were not reported in the paper, from the global database at http://www.worldclim.org/ using latitude and longitude coordinates. Considering that agricultural and nonagricultural ecosystems may respond differently to N addition, we examined the effects of N addition on the ecosystem N cycle for these two ecosystems separately. Given that only a few studies of N-fixing plants were found for the N addition experiment in nonagricultural ecosystems, plant species were grouped into N-fixing and non-N-fixing plants within agricultural ecosystems to examine their responses to N addition. We also grouped our data according to N application rate (0–5, 5–10 and > 10 g N m−2 yr−1) and experimental duration (0–5, 5–10 and > 10 yr) to explore their effects on the responses of the ecosystem N cycle to N addition. In agricultural ecosystems, fertilization and control groups received the same irrigation treatment. In nonagricultural ecosystems, experiments involving irrigation treatments were excluded as only experiments comparing fertilization treatments (with or without N fertilization) were selected. Thus, irrigation treatments in agricultural ecosystems should not greatly affect the general patterns observed in this meta-analysis. Statistical analyses The response ratio (RR) was used to reflect the effects of N addition on terrestrial ecosystem N pools and fluxes (Hedges et al., 1999). The RR, the ratio of the mean value of the chosen variable in the N addition group () to that in the control group (), is an index of the effect of N addition on the corresponding variable (Eqn 1). More specifically, the mean, standard deviation (S) or standard error, and sample size for each treatment were extracted to calculate the logarithm of RR, the variance (v), the weighting factor (wij), the weighted response ratio (RR++) and the 95% confidence interval (CI) of RR++ for the purpose of statistical tests (Eqns 2–6) (Curtis & Wang, 1998; Gurevitch & Hedges, 1999; Luo et al., 2006). The frequency distribution of logeRR was assumed to follow normal distribution and fitted by a Gaussian function (Eqn 7, Luo et al., 2006). If the 95% CI values of RR++ for a variable did not cover zero, the effects of N addition on the variable were considered to differ significantly between two treatments. Otherwise, they were not considered to differ significantly. We also used a t-test to examine whether the RR++ of a variable differed significantly between agricultural and nonagricultural ecosystems, between N-fixing and non-N-fixing plants, among different N application rates, and among various experimental durations. The per cent change in a variable was estimated by. In addition, the relationships between logeRR and environmental and/or forcing factors were examined using a single-factor regression approach. (Eqn 1) (Eqn 2) (Eqn 3) (Eqn 4) (Eqn 5) (Eqn 6) (Eqn 7) (x, the mean of logeRR in an individual interval; y, the frequency in an interval; a, the expected number of logeRR values at x = μ; μ and σ2, the mean and variance of the normal distribution of logeRR, respectively.) Results Responses of N pools and fluxes to N addition N-induced changes in N pools exhibited great variability across the studies, ranging from a minimum logeRR of −0.084 to a maximum of 1.31 in the leaf, from −0.083 to 1.69 in the aboveground plant, from −0.40 to 1.73 in the belowground plant, and from −0.16 to 1.25 in litter (Fig. 1). On average, the overall effects of N addition on plant N pools were positive, with an increase of 23.9% in leaf N pool (P < 0.05; Fig. 1a), 44.2% in the aboveground plant N pool (P < 0.05; Fig. 1b), 53.2% in the belowground plant N pool (P < 0.05; Fig. 1c), and 24.2% in the litter N pool (P < 0.05; Fig. 1d). Figure 1Open in figure viewerPowerPoint The frequency distributions of the natural logarithm of the response ratio (logeRR) for leaf (a), aboveground plant (b), belowground plant (c), and litter (d) nitrogen (N) pool responses to N addition. The solid curve is a Gaussian distribution fitted to the frequency data. The x-axis is logeRR and the y-axis is frequency. The vertical dashed line is at logeRR = 0. N addition significantly decreased microbial biomass N by 5.8% (P < 0.05; Fig. 2a), while total N pools in both the organic horizon and mineral soil increased by 6.1 and 6.2% under N enrichment, respectively (P < 0.05; Fig. 2b,c). Also, averaged dissolved organic N (DON) increased by 21.1% in the N addition group in comparison with the control group (P < 0.05; Fig. 2d). In addition, N addition significantly increased soil inorganic N (SIN) by 114.8% (P < 0.05; Fig. 2e), with a 47.2% rise in the NH4+ pool and a 428.6% rise in the NO3− pool (Fig. S2). Among all N pool variables, the N-induced increases in organic horizon and mineral soil N were among the smallest (Fig. 2f). Figure 2Open in figure viewerPowerPoint The frequency distributions of the natural logarithm of the response ratio (logeRR) for microbial biomass nitrogen (N) (a), organic horizon N pool (b), soil N pool (c), dissolved organic N (d) and soil inorganic N (e) responses to N addition, and the weighted response ratio (RR++) for the responses to N addition of nine variables related to the ecosystem N pool (f). The solid curve is a Gaussian distribution fitted to the frequency data. The x-axis is logeRR and the y-axis is frequency. The vertical dashed line is at logeRR = 0. Both N influx and efflux were stimulated under N addition (Fig. 3). Compared with those in control groups, soil net N mineralization, nitrification, denitrification and inorganic N leaching increased in the N fertilization groups by 24.9, 153.9, 84.3 and 460.9%, respectively (P < 0.05; Fig. 3). Moreover, N2O emissions increased by 133.6% in response to N addition (P < 0.05; Fig. 3c). However, N immobilization showed only a minor change with external N input (Fig. 3d). Figure 3Open in figure viewerPowerPoint The frequency distributions of the natural logarithm of the response ratio (logeRR) for net nitrogen (N) mineralization (a), nitrification (b) and N2O emission (c) responses to N addition, and the weighted response ratio (RR++) for N immobilization, mineralization, denitrification, N2O flux, nitrification and soil inorganic leaching (d) responses to N addition. The solid curve is a Gaussian distribution fitted to the frequency data. The x-axis is logeRR and the y-axis is frequency. The vertical dashed line is at logeRR = 0. Differential responses in agricultural and nonagricultural ecosystems N-induced changes in N pools and fluxes in agricultural ecosystems were different from those in nonagricultural ecosystems (Fig. 4). The increments of leaf, aboveground plant and litter N pools in agricultural ecosystems were larger than those in nonagricultural ecosystems (P < 0.05), while the increment of SIN in agricultural ecosystems was significantly smaller than that in nonagricultural ecosystems (P < 0.05). Moreover, the increments of net N mineralization, nitrification, and inorganic N leaching in nonagricultural ecosystems were greater than those in agricultural ecosystems (P < 0.05). However, the increases in belowground plant N, soil N pool and DON showed no significant differences between agricultural and nonagricultural ecosystems (P > 0.1). In addition, N addition decreased microbial biomass N in both agricultural and nonagricultural ecosystems (P < 0.05), and the changes in microbial biomass N did not reveal significant differences between agricultural and nonagricultural ecosystems (P = 0.32). Figure 4Open in figure viewerPowerPoint The weighted response ratio (RR++) for the responses to nitrogen (N) addition of 15 variables related to the ecosystem N cycle in agricultural (open bars) and nonagricultural (closed bars) ecosystems. Bars represent RR++ ± SE. The vertical line is drawn at logeRR = 0. The sample size for each variable is shown next to the bar. DON, dissolved organic N. Factors affecting the responses of N pools and fluxes to N addition Both aboveground plant and litter N pools in non-N-fixing crops exhibited larger responses than those in N-fixing crops under N addition (P < 0.05; Fig. 5). Moreover, an increase in the N application rate from 0–5 to 5–10 g N m−2 yr−1 led to greater increases in leaf, aboveground plant and litter N pools (P < 0.05; Fig. 6). In addition, litter N accumulation tended to increase, while changes in microbial biomass N tended to decrease with experimental duration (P < 0.05; Fig. 7, Tables S2, S3). Figure 5Open in figure viewerPowerPoint The weighted response ratio (RR++) for the responses to nitrogen (N) addition of six variables related to the ecosystem N cycle, with two functional groups of N fixation (open bars, N-fixing; closed bars, non-N-fixing). Bars represent RR++ ± SE. The vertical line is drawn at logeRR = 0. The sample size for each variable is shown next to the bar. Figure 6Open in figure viewerPowerPoint The weighted response ratio (RR++) for the responses to nitrogen (N) addition of 11 variables related to N pools and fluxes, with three N application rates (0–5 N m−2 yr−1, dark grey bars; 5–10 N m−2 yr−1, light grey bars; > 10 g N m−2 yr−1, black bars). Bars represent RR++ ± SE. The vertical line is drawn at logeRR = 0. The sample size for each variable is shown next to the bar. Figure 7Open in figure viewerPowerPoint The weighted response ratio (RR++) for the responses to nitrogen (N) addition of 11 variables related to nitrogen (N) pools and fluxes, with three experimental durations (0–5 yr, dark grey bars; 5–10 yr, light grey bars; > 10 yr, black bars). Bars represent RR++ ± SE. The vertical line is drawn at logeRR = 0. The sample size for each variable is shown next to the bar. N-induced changes in the aboveground plant N pool slightly increased with latitude (r2 = 0.07, P < 0.05), but were negatively correlated with MAT (r2 = 0.21, P < 0.001) and MAP (r2 = 0.14, P < 0.05) (Fig. 8). The changes in the aboveground plant N pool were positively related to N application rate (r2 = 0.07, P < 0.05). However, N-induced changes in the aboveground plant N pool did not show any significant correlations with experimental duration or the cumulative amount of N. Moreover, the relationships between the logeRR of the aboveground plant N pool and environmental factors did not differ significantly between agricultural and nonagricultural systems, except that the changes in the aboveground plant N pool were not significantly correlated with N application rate in nonagricultural ecosystems (P = 0.13; Fig. 8d). Both environmental and forcing factors also regulated the responses of the litter N pool and soil N pool to N addition. Specifically, the logeRR of the litter N pool significantly increased with the cumulative amount of N and experimental duration, but significantly decreased with MAT (P < 0.05) (Table S2). The logeRR of the organic horizon N pool significantly increased with the logeRR of the litter N pool (r2 = 0.46, P < 0.05) (Table S3). However, the logeRR of the soil N pool only increased with the logeRR of the belowground plant N pool (r2 = 0.36, P < 0.05) (Fig. S1), and slightly decreased with soil depth (r2 = 0.02, P < 0.05) (Table S3). Figure 8Open in figure viewerPowerPoint Relationships between the natural logarithm of the response ratio (logeRR) for the response to nitrogen (N) addition of the aboveground plant nitrogen (N) pool (APNP) and latitude (a), mean annual temperature (MAT) (b), mean annual precipitation (MAP) (c), N application rate (d), experimental duration (e), and cumulative N amount (f). Open circles, agriculture; closed circles, nonagriculture. Discussion Response of ecosystem N pools to N addition N addition stimulated N sequestration in both plant and litter pools. The net N accumulation in plants under N addition could be attributed to the increased plant biomass and plant N concentration. As a growth-limiting factor, the external input of N usually leads to increases in both plant N concentration (Xia & Wan, 2008) and net primary production (LeBauer & Treseder, 2008) in terrestrial ecosystems, and thus results in increased plant N accumulation. The increased litter N pool is logically consistent with N accumulation in both above- and belowground plant pools. As the major input to litter, the N-induced increase in plant N content could ultimately lead to net N accumulation in litter (Vanotti et al., 1995; Mack et al., 2004). Our results also showed that the changes in litter N pool were positively correlated with those in the aboveground plant N pool (r2 = 0.54, P < 0.01) (Fig. S1), indicating that the accumulation of the litter N pool could be driven by the increases in the aboveground plant N pool under N addition. By contrast, N addition decreased microbial biomass N. In ecosystems there may be a number of processes that lead to a decline in the microbial N pool. First, the increased amount of soil inorganic N caused by N addition can react with soil organic matter and result in the accumulation of recalcitrant compounds (Soderstrom et al., 1983; Fog, 1988), which may be unavailable for microbial growth in the N addition scenario (Treseder, 2008; Janssens et al., 2010). Secondly, N addition significantly decreased soil pH by 3.5% across various ecosystems (Fig. S2). N-induced soil acidification may result in calcium and magnesium leaching and other corresponding changes in soil physical–chemical properties, which may limit microbial biomass growth (Vitousek et al., 1997; Treseder, 2008). In addition, it has also been reported that the potential N saturation in the N addition scenario may constrain the activities of β-glucosidase in mineral soil, causing a decline in microbial C acquisition and a decrease in microbial biomass (DeForest et al., 2004). N addition led to N sequestration in both the organic horizon and mineral soil, but to a much lower degree than in plant and litter pools. The smaller response of the soil N pool to experimental N addition compared with the plant N pool may reflect the difference in the sizes of these N pools (Batjes, 1996). It usually takes longer to increase N content in a large pool than in a small one. The smaller response of the soil N pool may also be partly attributable to minor increases in soil organic mass (Liu & Greaver, 2010) and low stoichiometrical flexibility in the narrow-range C : N ratio in soil organic matter. Nevertheless, our analysis showed that the changes in the soil N pool were significantly correlated with those in the belowground plant N pool (r2 = 0.36, P < 0.05) (Fig. S1), indicating that belowground plant N dynamics may contribute to changes in the mineral soil N pool. However, the N added to the soil could be lost via stimulated N fluxes, such as nitrification, denitrification, N2O emission and inorganic N leaching. Thus, N addition resulted in low increases in total N pools in both the organic horizon and mineral soil. Enhanced N fluxes in response to N addition N addition stimulated net N mineralization in terrestrial ecosystems. An increase in N mineralization may be induced by increases in DON and the soil N pool (Chapin et al., 2002; Booth et al., 2005), and decreases in the C: N ratio in mineral soil (Barrios et al., 1996). We found that DON and the total N pool in mineral soil increased by 21.1 and 6.2%, respectively (Fig. 2), while the C : N ratio in mineral soil significantly decreased by 1.9% in response to N addition. As a consequence, increased substrate quantity and quality under N enrichment may exert positive effects on N mineralization. External N input increased nitrification in terrestrial ecosystems. An increase in nitrification may be driven by changes in soil NH