Changes in soil phosphorus fractions after 9 years of continuous nitrogen addition in a Larix gmelinii plantation

Abstract

• The key message

N addition decreased soil inorganic P availability, microbial biomass P, and acid phosphatase activity in the larch plantation. Soil inorganic P availability decreased after N addition due to the changes in both microbial properties and plant uptake.

• Context

Soil phosphorus (P) availability is considered an important factor in influencing the biomass production of plants. Sustained inputs of nitrogen (N) through atmospheric deposition or N fertilizers, particularly in temperate forests, may change the composition and availability of P and thus affect long-term forest productivity.

• Aims

The objective of this study was to assess soil P availability, P fractions, and microbial properties including microbial biomass P and acid phosphatase activity after 9 consecutive years of N addition in a larch (Larix gmelinii) plantation, northeastern China.

• Methods

From 2003 to 2011, NH₄NO₃ was added to replicate plots (three 20 m × 30 m plots) in the larch plantation each year at a rate of 100 kg N ha⁻¹ year⁻¹. Soil samples from 0–10-cm and 10–20-cm depths were collected in N addition plots and control (no N addition) plots.

• Results

N addition significantly decreased soil NaHCO₃-Pi (Pi is inorganic P), microbial biomass P, and acid phosphatase activity but increased the NaOH-Pi concentration. N addition appeared to induce a decrease in soil inorganic P availability by changing pH and P uptake by trees. In addition, N addition significantly decreased the NaOH-Po (Po is organic P) concentration, possibly because of increased P mineralization. However, the total P and other P fractions were unaffected by N fertilization.

• Conclusion

Our results suggested that N addition enhanced P uptake by trees, whereas it reduced soil inorganic P availability as well as microbial biomass and activity related to soil P cycling in the larch plantation.

1 Introduction

Nitrogen (N) is generally believed to be the one nutrient limiting primary productivity in a wide variety of terrestrial ecosystems (LeBauer and Treseder 2008), particularly in temperate forest ecosystems with relatively young soils (Vitousek and Howarth 1991). Thus, fertilization with N often stimulates tree growth, resulting in increased plant production in such ecosystems (Thomas et al. 2010). Recently, increased plant production after N enrichment has led to a concern that forest ecosystems may be limited by other important nutrients such as phosphorus (P) (Braun et al. 2010). Although studies have focused on the impact of increased N on carbon (C) cycling (Sinsabaugh et al. 2005; Nave et al. 2009; Cusack et al. 2010), the question of how increased N affects the availability and cycling of P in forest soils receives little attention. Given that P is an important structural component of nucleotides, phospholipids, and nucleic acids in plants (Schachtman et al. 1998), it is important to understand the effect of N addition on the forms and availability of P in soils.

The P in terrestrial ecosystems exists in many complex chemical forms, including inorganic mineral forms and organic forms, which differ markedly in their behavior, mobility, and bioavailability in the soils. Studies have reported that soil P availability, especially inorganic P, is directly linked to net primary productivity (Herbert and Fownes 1995; Paoli and Curran 2007; Baribault et al. 2012). In forest soils, N addition may stimulate primary production and increase biotic P demand, therefore decreasing soil inorganic P availability (Vitousek et al. 2010). Mycorrhizae may play an important role in the plants because they may affect uptake of relatively immobile nutrients, such as P, then changing soil P availability (Clark and Zeto 2000). It has been suggested that N inputs may significantly enhance nutrient cycling in the terrestrial ecosystems via stimulation of root growth and mycorrhizal functioning (Tu et al. 2006). In addition, N addition decreases soil pH, which then mobilizes soil aluminum and iron, therefore reducing available P through increased P sorption and decreased mineralization of organic matter (Carreira et al. 2000). The content and forms of soil organic P may also be influenced by N addition through changes in organic matter input (Khan et al. 2008). Taken together, if continued N addition changes the soil P cycle, then understanding how the soil P status changes in response to N addition will be important for predicting forest ecosystem function.

Nitrogen addition could influence soil P availability through microbial changes, especially phosphatase enzyme activity (Marklein and Houlton 2012). For example, phosphatase enzyme is rich in N, meaning that soil microorganisms must sacrifice N to get P via this path (Olander and Vitousek 2000). In addition, from a functional perspective, the phosphatase enzyme response to N addition has received considerable attention because this enzyme contributes to the processes controlling organic P mineralization (Criquet et al. 2004). Studies that examined phosphatase enzyme activities after N addition in different forests have produced contradictory results. One study reported that soils underlying a Eucalyptus dunnii plantation after N addition increased in both phosphatase activity and P availability (Wang et al. 2008), while another study showed no significant change in phosphatase activity after N addition in young spruce stands in Sweden (Clarholm 1993). Importantly, the response of phosphatase enzyme activity to added N has direct effect (increase in N availability) or indirect effects (changes in pH or concentrations of cations/anions other than N availability). To better illustrate soil P status after N addition in temperate forests, we need to understand how soil phosphatase enzyme activity was affected by added N.

During the last few decades, rapid economic development in China has resulted in accelerated N deposition (Lü and Tian 2007). Northern China is considered one of the main N deposition areas in Asia. The N deposition levels in northern China are in the range of 28.5–100.4 kg N ha⁻¹ year⁻¹ (Pan et al. 2012). Larch species are widely distributed throughout northern Asia, and many are important forestry species (Wang et al. 2006; Ohsawa and Shimokawa 2011). In previous studies, the effects of N addition on soil respiration, microbial properties, and methane and nitrous oxide fluxes have been investigated (Hu et al. 2010; Jia et al. 2010; Kim et al. 2012). However, little is known about how N addition may alter soil P status in larch plantations.

In this study, we investigated soil P forms and the microbial properties affecting P cycling and transformation (i.e., microbial biomass P and acid phosphatase activity) in response to 9 consecutive years of N addition in a larch (Larix gmelinii) plantation in northeastern China. At this site, previous study has shown that N addition significantly decreased soil pH and microbial biomass (Hu et al. 2010) but increased P concentration in plant tissues (Zhao et al. 2010). We hypothesized that soil P availability and P fractions would change after N addition due to the changes in both biochemical (pH and microbial biomass and activity) properties and plant uptake.

2 Material and methods

2.1 Study site and experimental design

This study was conducted at the Maoershan Experimental Station of Northeast Forestry University, located in a mountainous area, Heilongjiang Province, China (45° 21′–45° 25′ N, 127° 30′–127° 34′ E). The region has a continental monsoon climate with mean annual temperature of 2.8 °C, mean minimum temperature of −19.6 °C in January, and mean maximum temperature of 20.1 °C in July. The growing season is from early May to late September. Mean annual precipitation is approximately 700 mm, with >60 % falling between June and August. The frost-free period is between 120 and 140 days (Zhou 1994). The soil is Hap-Boric Luvisol (Gong et al. 1999). Soil depth of the study site exceeds 50 cm and with high organic matter content. The soil is a loam soil in the 0–10-cm depth and a sandy loam soil in the 10–20-cm depth. More details about the soil were given by Wang et al. (2006).

Larch is the key commercial tree species and represents 65 % of conifer plantations in northeast China (Wang et al. 2006). In this study, a larch (L. gmelinii) plantation was selected on a southwest facing slope of approximately 13°, with an elevation ranging from 450 to 500 m above sea level. The plantation was established in 1986 by planting 2-year-old seedlings using a 1.5 m × 2.0 m planting grid. At the beginning of the experiment, the tree density, mean tree diameter at breast height, and tree height were 2267 trees ha⁻¹, 10.9 cm, and 11.3 m in the larch plantation; see Mei et al. (2010) for more details on stand characteristics.

Nitrogen addition plots were established in the plantation in May 2002. Three 20 m × 30 m ammonium nitrate (10 g N m⁻¹ year⁻¹) fertilized plots were paired with control plots of the same size, with each plot surrounded by ≥10-m-wide buffer strip. Therefore, a total of six plots were established for soil sample collection. N addition experiments were initiated in 2003. Ammonium nitrate was applied in pellet form each year between 2003 and 2011. All of the 10 g N m⁻¹ year⁻¹ was applied each year during the growing season, with 15.25 % of the fertilizer added in May and in September, 21 % in June and in August, and 27.5 % in July. The amount of fertilizer applied each month was designed to track the natural temporal patterns of N mineralization that occurs at the site in each particular month (Chen et al. 1999).

2.2 Soil sampling

Soils were sampled from each plot in August (approximately 3 weeks after the last addition of N in July) 2011. Fifteen soil cores (5 cm in diameter) were taken randomly from each plot after removal of the forest floor litter and sectioned into 0–10-cm and 10–20-cm depths. The cores were then mixed to form a composite soil sample for each soil depth. Soil was sieved (<4 mm) and stored at 4 °C for later analysis after visible roots and plant residues were removed. Soil samples were further divided into three sets of sub-samples. One set of the sub-samples was processed to pass through a 2-mm sieve and stored at 4 °C for measurement of microbial biomass P and acid phosphatase activity, one set was air-dried and passed through a 0.5-mm sieve for analysis of soil pH and P fractionation, and the remaining set of sub-samples was oven-dried and homogenized and passed through a 0.25-mm sieve for analyses of soil organic carbon (C) and total nitrogen (N). The selected soil chemical properties are listed in Table 1.

Table 1 Changes in selected soil chemical properties after 9 years of N addition in the larch plantation

2.3 Soil analyses

Soil organic C and N were analyzed by dry combustion using a Vario EL III elemental analyzer (Elementar Analysensysteme GmbH, Hanau, Germany). Ammonium (NH₄ ⁺-N) and nitrate (NO₃ ⁻-N) were extracted with 2 M KCl and analyzed on an auto analyzer (Auto Analyzer III, Bran + Luebbe GmbH, Germany). Extractable Al and Fe were extracted with acid ammonium oxalate, and extracts were analyzed by inductively coupled plasma mass spectrometry (Agilent Technologies Co. Ltd, USA). Soil pH was estimated from a 1:2.5 soil-water slurry.

Soil P was sequentially fractionated following a modified version of the Hedley et al. (1982) fractionation scheme. Briefly, 0.5 g of soil was successively extracted with 30 mL deionized water, 0.5 mol L⁻¹ NaHCO₃ (pH 8.5), 0.1 mol L⁻¹ NaOH, and 1 mol L⁻¹ HCl for 17 h each. The inorganic P (Pi) concentration of each extract was determined and referred to as either water-soluble (H₂O-soluble) or NaHCO₃-, NaOH-, or HCl-extractable Pi fractions. The residue was digested using H₂SO₄ + H₂O₂ at 360 °C for the measurement of residual P (Tiessen and Moir 1993). The total P concentration of each extract was also determined after acid persulfate digestion (autoclaving for 30 min at 121 °C and 1.05 kg cm⁻¹ pressure). The organic P (Po) concentration of these extracts was calculated as the difference between total and inorganic P, referred to as the H₂O-, NaHCO₃-, NaOH-, or HCl-extractable Po fractions. The soil total P was the sum of all P fractions.

Soil microbial biomass P was determined using the chloroform fumigation extraction method. The microbial biomass P was estimated from the relationship E _P/K _EP, where E _P is the difference between inorganic P extracted from fumigated and unfumigated soils and K _EP = 0.40 (Brookes et al. 1982). The correction for chloroform-released P that was absorbed by soil colloids during extraction was made by adding 25 mg P kg⁻¹ soil during extraction and then correcting for its recovery (Brookes et al. 1982).

Acid phosphatase activity was determined according to the method of Tabatabai (1994). Briefly, soil samples were incubated with p-nitrophenyl phosphate (p-NPP) as the substrate for 1 h, the reaction was terminated with 0.5 M NaOH and 0.5 M CaCl₂, and the absorbance was determined spectrophotometrically at 400 nm. Controls without enzymes were processed in parallel to determine non-enzymic hydrolysis of the substrate and to correct for background coloration. Enzyme activities are expressed as mg p-NPP kg⁻¹ soil h⁻¹. Controls without soil were processed in parallel to correct for background coloration.

2.4 Statistical analyses

Analysis of variance (ANOVA) of a split-plot design with the treatments as the whole plot and the sampling depths as the split plot was performed to determine whether N addition significantly affected soil P fractions, microbial biomass P, and acid phosphatase activity. These statistical analyses were conducted using SAS version 9.2 (SAS Institute Inc., Cary, NC, USA).

3 Results

3.1 Soil properties

Of the selected soil chemical properties, soil organic C and N, NH₄ ⁺-N, and extractable Fe were not affected by N addition (Table 1). By comparison, the soil-extractable Al and C-to-N ratio increased significantly after N addition at a depth of 0–10 cm. Soil NO₃ ⁻-N increased significantly after N addition in both 0–10-cm and 10–20-cm depths. The soil pH value at the 0–10-cm depth in the N addition plots was 0.5 units lower than those in the control plots. Soil chemical properties changed with soil depth (Table 1). Soil organic C and N and the soil C-to-N ratio at the 0–10-cm depth were greater than those at the 10–20-cm depth for both the control and the N addition plots. However, differences in soil pH between the 0–10-cm and 10–20-cm depths were only observed in the N addition plots.

3.2 Soil total P and P fractions

Soil total P concentrations, i.e., the sum of the concentrations of all P fractions, at both sampling depths were not affected by N addition (Fig. 1). Soil total P at the 0–10-cm depth was greater than that at the 10–20-cm depth for both the control and N addition plots. The relative size of the P pools increased in the order H₂O-Pi ≈ H₂O-Po < HCl-Po < NaHCO₃-Pi < NaHCO₃-Po < HCl-Pi < NaOH-Pi < NaOH-Po < Residual-P (Figs. 2 and 3) for both the control and N addition plots.

Fig. 1

Changes in soil total P concentration (the sum of the concentrations of all fractions) after 9 years of N addition in the larch plantation. Bars represent means with standard errors for n = 3. Different letters indicate significant differences at P < 0.05

Fig. 2

Changes in soil inorganic phosphorus (Pi) fractions (extracted by H₂O, NaHCO₃, NaOH, and HCl) after 9 years of N addition in the larch plantation. Bars represent means with standard errors for n = 3. Different letters indicate significant differences at P < 0.05

Fig. 3

Changes in soil organic phosphorus (Po) fractions (extracted by H₂O, NaHCO₃, NaOH, and HCl) and residual P fraction after 9 years of N addition in the larch plantation. Bars represent means with standard errors for n = 3. Different letters indicate significant differences at P < 0.05

N addition significantly reduced the NaHCO₃-Pi concentration that is commonly considered as labile P pool (Fig. 2). In contrast, the NaOH-Pi concentration at the 0–10-cm depth in the N addition plots was significantly higher than that in the control plots. There were no significant changes in the H₂O-Pi and HCl-Pi concentrations in the N addition plots. The Pi fractions extracted by H₂O, NaHCO₃, and NaOH at the 0–10-cm depth were markedly higher than those at the 10–20-cm depth for both treatments but not for the HCl-Pi fraction.

The NaOH-Po concentration at the 0–10-cm depth in the N addition plots was significantly lower than that in the control plots. However, the Po fractions extracted by H₂O, NaHCO₃, and HCl were similar between the control and N addition plots for a given depth (Fig. 3). The H₂O-Po concentration was also significantly higher at the 0–10-cm depth than at the 10–20-cm soil depth, but the other three Po fractions were similar between the two sampling depths. The residue P concentration did not change with N addition in both 0–10-cm and 10–20-cm soil depths.

3.3 Soil microbial biomass P and acid phosphatase activity

Soil microbial biomass P decreased approximately 40 % after 9 years of N addition at the 0–10-cm soil depth (Fig. 4). There was no difference in microbial biomass P concentration between the control and N addition plots at the 10–20-cm soil depth. The microbial biomass P concentration at the 0–10-cm depth was greater than that at the 10–20-cm depth in same treatment.

Fig. 4

Changes in soil microbial biomass P after 9 years of N addition in the larch plantation. Bars represent means with standard errors for n = 3. Different letters indicate significant differences at P < 0.05

N addition also significantly reduced acid phosphatase activity at the 0–10-cm depth. However, soil acid phosphatase activity in the N addition plots was not different from that in the control plots at the 10–20-cm depth (Fig. 5). The acid phosphatase activity was approximately 1.5-fold greater at the 0–10-cm depth than that at the 10–20-cm depth in both the control and N addition plots.

Fig. 5

Changes in soil acid phosphatase activity after 9 years of N addition in the larch plantation. Bars represent means with standard errors for n = 3. Different letters indicate significant differences at P < 0.05

4 Discussion

Our results show that the soil P fraction, microbial biomass P, and acid phosphatase activity are highly responsive to changes in N applied to the larch plantation. This study has yielded two main findings. First, added N decreased the concentration of labile inorganic P in the soil, especially the NaHCO₃-Pi concentration. Second, soil acid phosphatase activity decreased after N addition, and this result contradicts most of our current understanding of the interaction between nutrient availability and microbial function. Based on allocation theory and prior studies conducted in forest ecosystems (Treseder and Vitousek 2001; Saiya-Cork et al. 2002; Marklein and Houlton 2012), N addition should increase acid phosphatase activity. However, we found that N addition resulted in decreased acid phosphatase activity in the larch plantation.

4.1 N addition changes soil P fractions

After 9 years of N addition, soil total P did not change significantly in the N addition plots. Total P content represents the net result between the input of litter and the output of plant uptake. In our study, no significant change was observed in soil total P between the control and N treatment, suggesting that the input and output of P via aboveground and belowground may be in balance. However, the application of N fertilizer to the larch plantation resulted in a significant decrease in the NaHCO₃-Pi concentration in the topsoil compared with the control. This finding disagreed with the result from northern hardwood forests in the eastern USA, where soil P availability and biotic P sufficiency differed by tree species but were unaffected by N fertilization (Weand et al. 2010b). Weand et al. (2010b) reported that P concentrations in foliar, litter, and fine roots were not affected by N fertilization. In this study, N addition stimulates potential net N mineralization and potential net nitrification processes (Hu et al. 2010) and may increase biotic P demand. A recent study in our study area showed that N addition could increase the N and P concentrations in the stem, branch, and roots of the larch plantation (Zhao et al. 2010; Table 2). N addition enhanced P uptake by trees and might account for the decease of soil NaHCO₃-Pi concentration. In this study, the aboveground and belowground biomass, including stem, branch, and coarse root (>2 mm in diameter) biomass, increased approximately 20 % after N addition in our study area (Gu, Northeast Forestry University, data not published), which may also be attributed to the decrease of soil NaHCO₃-Pi concentration. Therefore, N addition increased plant growth and root P uptake and then assigned to the different parts of the trees could be responsible for the decrease in the soil NaHCO₃-Pi concentration (Zhao et al. 2010). However, it should be noted that the decrease in the NaHCO₃-Pi concentration observed in this study could also be influenced by soil microbial biomass and activity. A recent study also reported that the contents of microbial biomass and Olsen-P (i.e., NaHCO₃-Pi) showed a highly significant linear relationship under different management conditions (Khan and Joergensen 2012). This finding disagreed with the result from an evergreen Mediterranean forest that the decrease in soil available P would reduce the P uptake (Sardans and Peñuelas 2004). Our result was the opposite and such discrepancies may be explained by differences in soil labile inorganic P concentration. In our study, the labile inorganic P concentration was high, and N addition might not result in P limitation or saturation of tree demand for N. Furthermore, mycorrhizae play an important role in the P uptake by trees. In our study area, we know little regarding mycorrhizal associations important for P acquisition, or mechanisms of P uptake. Future investigations comparing mycorrhizal communities between the control and the fertilized plots, and their abilities to access P, could further elucidate the decline of labile inorganic P.

Table 2 Effects of N addition on nutrient contents in leaf, branch, stem and roots of Larix gmelinii

The NaOH-Pi fraction is generally considered the less readily available P associated with Fe and Al oxides and may represent P adsorption sites (Cross and Schlesinger 1995). In the present study, the NaOH-Pi fraction was significantly increased following long-term N application. NH₄NO₃ addition likely enhances soil nitrification, a microbially mediated process of ammonia oxidation with H⁺ as the by-product resulting in soil acidification. The low pH values in the N addition plots might enhance the solubility of Al³⁺ and Fe³⁺ and then increase the P binding by Al and Fe in the soils (Chen 2003; Dossa et al. 2010). Therefore, the overall combination of changes in the inorganic P fractions observed in our study demonstrates that N input may induce a decrease in soil P availability through changes in plant P uptake, microbial biomass, and inorganic P solubilization.

N fertilization did not increase soil organic P fractions, although biomass inputs, including stem, branch, and coarse root biomass, increased in the fertilized plots (Gu, Northeast Forestry University, data not published). In the present study, the NaOH-Po content was the dominant organic P fraction. The high content of NaOH-Po in the larch plantation was consistent with previous research in Pinus radiata forest soil (Liu et al. 2004). Interestingly, the larch plantation in this study depleted the NaOH-Po fraction, but other organic P fractions were unaffected by N addition, suggesting that the NaOH-Po pool is important in providing plant-available P through mineralization. Therefore, the increased growth of L. gmelinii after N addition, and hence higher nutrient demand, is reflected in the NaOH-Po fraction in the soil. It has been suggested that the NaOH-Po pool can be the major sink of available P in the soil and may maintain the levels of plant-available P through P mineralization (Beck and Sanchez 1994).

4.2 N addition effects on microbial biomass P and acid phosphatase activity

Soil microbial biomass is an important component influencing plant nutrient availability. In this study, N addition caused microbial biomass to decrease in fertilized larch plots, supporting the previous finding of Hu et al. (2010) that soil microbial biomass was lower in fertilized plots in the same study site. There are two potential mechanisms to explain the decrease in soil microbial biomass with N addition. On the one hand, N addition could retard mineralization of old and humified soil organic matter (Neff et al. 2002; Hagedorn et al. 2003) then reduce soil labile organic C, which is the energy source for soil microorganisms (Sjöberg et al. 2003). On the other hand, soil pH is an important factor regulating microbial properties. Generally, soil acidification after N addition could reduce soil microbial biomass and activity (Liu et al. 2011). A recent meta-analysis using 82 field studies after N addition concluded that microbial biomass declined 15 % on average under N fertilization, and these declines in microbial biomass were more evident in studies of longer duration and with higher total amounts of N addition (Treseder 2008). Our study site receives a high N dose compared with other N addition studies in forest ecosystems (Will et al. 2006; Kritzler and Johnson 2010). Therefore, it is not surprising that microbial biomass was reduced by approximately 40 % after 9 years of N addition in our study site.

Phosphatase plays a critical role in catalyzing the hydrolysis of organic P compounds in soil. In general, N addition is thought to influence P availability through the N-rich phosphatase enzymes produced by plant roots and microbes. Biota can invest N into the production of hydrolytic enzymes such as phosphatase for acquiring P when N availability is sufficient (Houlton et al. 2008). In addition, the decrease in soil P availability after N addition would induce higher phosphatase activity. However, we observed that N addition depressed acid phosphatase activity. This finding disagreed with the results that increasing N availability or increasing N deposition increased phosphatase activity (Clarholm and Rosengren-Brinck 1995; Turner et al. 2002). Enzyme activity is known to decline with increasing acidity (Weand et al. 2010a), and N application may have acidified the fertilized plots to the point where enzyme activity is suppressed. Therefore, the decrease in acid phosphatase activity observed in the N addition plots than those in control plots was mainly indirect effect caused by the added N, such as the decrease in pH, rather than a direct result of the increase in soil N availability.

4.3 Implications

This study extends our knowledge of soil P status in the larch plantation with long-term N addition. Unlike other studies (Finzi 2009; Weand et al. 2010b), we found that N addition altered P distribution and availability in the larch plantation. The soil labile P pool significantly decreased following long-term N application. High N input in some forest ecosystems can increase the growth and nutrient uptake of tree species, thereby leading to a decline in soil P availability (Will et al. 2006). Soil pH appears to be another important factor affecting the sorption of Al- and Fe-associated P and accordingly P availability in terrestrial ecosystems (Waldrip et al. 2011). In the larch plantation, N addition inhibited soil acid phosphatase activity and microbial biomass P. The results suggest that N addition reduced soil P availability as well as microbial process related to soil P cycling in the L. gmelinii plantation. Although soil types and seasonal effects of N addition were not investigated, our study confirms soil P status about responses of larch plantation to N addition and provides new insights into the causes of these responses. Further research that examines the impact of N addition on P status in larch plantation ecosystem with different soil characteristics and sampling time would help assess the generality of our results.