Denitrification characteristics of pond sediments in a Chinese agricultural watershed

Authors

  • Feiyue LI,

    1. State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008
    2. College of Urban Construction and Environment Science, Anhui Science and Technology University, Fengyang 233100, China
    Search for more papers by this author
  • Rong YANG,

    1. State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008
    Search for more papers by this author
  • Chaopu TI,

    1. State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008
    Search for more papers by this author
  • Man LANG,

    1. State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008
    Search for more papers by this author
  • Sonoko D. KIMURA,

    1. Department of International Environmental and Agricultural Science, Tokyo University of Agriculture and Technology, Fuchu 183-8509, Japan
    Search for more papers by this author
  • Xiaoyuan YAN

    1. State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008
    Search for more papers by this author

X. YAN, State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, No. 71 Beijing East Rd, Nanjing, Jiangsu 210008, China. Email: yanxy@issas.ac.cn

Abstract

Ponds are widely distributed in rice-based agricultural watersheds, particularly in southern China, and may play an important role in nitrate (inline image) removal. However, the denitrification rate of pond sediment, measured using the acetylene (C2H2) inhibition technique, indicated that the amount of nitrogen removed by denitrification accounted for <1% of the total nitrogen applied. The present study was undertaken to determine the effects of sediment depth and temperature on denitrification of pond sediment using the C2H2 inhibition technique. The highest denitrification potential was found in the upper 5 cm of sediment, but this only accounted for approximately 34% of the total denitrification of the upper 0–30 cm of sediment, suggesting that sediment denitrification potential would be underestimated if only the upper 5 cm of sediment was measured. The denitrification potential was low and showed a weak response over a temperature range of 6–18°C, whereas denitrification increased significantly from 18 to 30°C, indicating that denitrification may play an important role in the removal of inline image in warm seasons. A comparison of the inline image disappearance and C2H2 inhibition methods showed that they were significantly (P < 0.01) and positively correlated. However, the C2H2 inhibition method gave only approximately 25% of the values determined by the inline image disappearance method.

Introduction

The massive use of nitrogen (N)-based fertilizer in agricultural systems has aggravated the N loads of water bodies in China. Annually, a considerable amount of N enters the Yangtze and Yellow Rivers, approximately 92 and 88%, respectively, of this N originates from agriculture and, in particular, approximately 50% originates from N fertilizer (Zhu et al. 2005). This can lead to eutrophication of water bodies, which is an increasing environmental concern in China. An investigation has shown that 56% of the area of the Yangtze River estuary and its adjacent sea were heavily polluted in terms of N concentration (Quan et al. 2005).

Denitrification plays a major role in the N cycle of aquatic systems. In this process, facultative anaerobic bacteria transform nitrate or nitrite into nitrogen gas (N2) that escapes into the atmosphere (Knowles 1982; Saunders and Kalef et al. 2001). Denitrification is considered to be the best way to reduce the N load in water bodies. However, most studies of sediment denitrification focus on rivers, lakes, reservoirs, estuaries, bays, inlets and gulfs and parts of the riparian zone (Barnes and Owens 1998; García-Ruiz et al. 1999; Livingstone et al. 2000; Tomaszek and Czerwieniec 2000; Saunders and Kalef et al. 2001). Studies on the denitrification potential of pond sediment in agricultural watersheds are rare (Yan et al. 1999).

Ponds are widely distributed in rice-based agricultural watersheds, particularly in southern China. Unlike lakes, ponds are usually smaller and located among cropland, and play an important role in the irrigation and drainage of rice cultivation. Therefore, the water in ponds is more dynamic than that in lakes. During the rainy season, much of the inline image originating from N fertilizer application is removed and runs into ponds, leading to inline image accumulation. The inline image of pond sediments can be reduced by denitrification, and the denitrification potential is influenced by the physical and chemical properties of the sediment and overlying water (Li et al. 2009). We measured the denitrification rate of pond sediment of the top 5 cm using the acetylene (C2H2) inhibition method and found that the denitrified N accounted for <1% of the total N applied to the rice (Li et al. 2009). Sampling depth and temperature are widely considered to be the most important factors affecting the denitrification potential of sediment (Groffman et al. 2002; Nowicki 1994). However, very little data are available on the effects of these two factors on the denitrification of pond sediment in agricultural watersheds. A better understanding of the denitrification characteristics of these pond sediments will improve predictions of the contribution of denitrification to the removal of inline image pollution and the management effects on soil N availability. The present study was undertaken to quantify the effects of sediment depth and temperature on the denitrification potential of ponds, in a typical rice-based agricultural watershed in China, using the C2H2 inhibition technique. We also compared the C2H2 inhibition technique and the inline image disappearance method.

Materials and methods

Sample collection

The study site was located in the Jurong agricultural watershed (21°57′–31°05′N, 119°10′–119°14′E), 40 km from Nanjing, Jiangsu Province, eastern China. The site receives a long-term average annual precipitation of 1050 mm, and the annual mean temperature is 15°C. Ponds are distributed widely in this rice-based agricultural watershed; we chose one typical pond to collect samples. The scale of the studied pond is approximately 856 m2; the water depth is approximately 2.2 m and the water temperature is 18°C. Twelve pond sediment cores (0–30 cm) were collected without disturbance using a custom-made corer on 29 March 2008. The corer was made of stainless steel and fitted with an organic vitreous transparent tube (8.4 cm in diameter and 35 cm long) for ease of observation. Unfiltered water samples for the incubation experiment were taken from the same location and the same depth using a Van Dorn sampler and placed in an acid-washed 1 L bottle. The sediment and water samples were transported to the laboratory as soon as possible. In the laboratory, the overlying water of the 12 samples was mixed fully and for the sediment cores (0–30 cm depth), each core layer at 5-cm intervals (0–5, 5–10, 10–15, 15–20, 20–25 and 25–30 cm) of the 12 samples was left in the same plastic pot, pooled and homogenized by gentle stirring to form one sediment slurry.

The average concentrations of inline image, ammonium (inline image) and total dissolved N in the overlying water were 0.65 ± 0.01, 0.28 ± 0.05 and 1.32 ± 0.08 mg N L−1 standard deviation (n = 3), respectively. The average organic matter and total N contents of sediment in the 0–5 cm layer were 5.64 ± 0.02% and 0.29 ± 0.01% (n = 3), respectively.

Denitrification potential of sediment at different depths

The denitrification potential of sediment is a measure of its capacity to reduce inline image to N2O or N2 (Jordan 1989; Klapwijk and Snodgrass 1982; Roy et al. 1994). The denitrification potential of sediment at different depths was measured in the laboratory using the C2H2 inhibition technique (Royer et al. 2004; Smith and Tiedje 1979; Tiedje 1982).

Approximately 14 mL of sediment slurry from each layer (0–5, 5–10, 10–15, 15–20, 20–25 and 25–30 cm) was dispensed into four replicated 250 mL conical flasks (total volume approximately 318 mL), and then 100 mL of KNO3 solution (dissolved in deionized water) was added at a concentration of 2 mg N L−1 as the culture medium. The conical flasks were sealed with a silicon rubber stopper and the headspace purged with He to create anoxic conditions. Pure C2H2 was added to the headspace of each tube to achieve a final gas mixture of 10% C2H2 and 90% He (David et al. 2006). The conical flasks were shaken to ensure complete infusion of C2H2 and incubated at 25°C for 4 h. Gas samples of approximately 5 mL were drawn from the headspace using a gas-tight syringe as the initial value. After incubation for 4 h, the conical flasks were shaken for at least 2 min to equilibrate the gas in the sediment and headspace, and then gas samples were collected.

Gas samples were analyzed for N2O on a Hewlett Packard 5890 Series II Gas Chromatograph (Hewlett Packard, Palo Alto, CA, USA) equipped with a Supelco 80/100 HAYESEPQ 3-m column (Supelco, Bellefonte, PA, USA) with an electron capture detector. The column and detector temperatures were 70 and 325°C, respectively. The sediment denitrification potential was calculated using the following equation:

image(1)
image(2)

where F is the denitrification potential (μg g−1 h−1), N is the flux of N2O (μg N2O-N g−1 h−1), M is the fresh weight of sediment slurry (g), Δt is the incubation time (4 h), ρ is the density of N2O under standard state (1.25 μg N2O-N mL−1), C is the concentration of N2O (p.p.b.v), Vg is the headspace of the conical flasks (mL), is the liquid volume (mL), α is a bunsen coefficient with a value of 0.544 at 25°C (Tiedje 1982), and T is the temperature when the gas samples were collected.

Denitrification potential of sediment at different temperatures

Approximately 14 mL of 0–5 cm sediment slurry with 100 mL of overlying water was dispensed into three replicated 250 mL conical flasks and incubated at 6, 12, 18, 24 or 30°C for 4 h using the C2H2 inhibition technique described above.

Comparison of the C 2H2 inhibition technique with the inline image disappearance method

The inline image disappearance method has been used to measure denitrification rates for many years (Andersen 1977; Yan et al. 1999). We assumed that the disappearance of inline image resulted only from denitrification and so was expressed as N removal capacity (μg N g−1 sediment h−1, fresh weight). We designed an experiment to compare the C2H2 inhibition technique with the inline image disappearance method. Three conical flasks with 14 mL of 0–5 cm sediment slurry and 100 mL of 2 mg N L−1 KNO3 were incubated at 6, 12, 18, 24 or 30°C for 4 h using the C2H2 inhibition technique described above. After collecting gas samples at 0 and 4 h, approximately 14.9 g of KCl was added to form 2 mol L−1 KCl to extract inline image from each flask. The inline image concentration was determined using a Continuous Flow Analyzer (Skalar, Breda, Holland). Denitrification rates using the inline image disappearance method were calculated using Eq. 3:

image(3)

where F is the denitrification rate (μg g−1 h−1), C1 and C2 are the respective concentrations of inline image at 0 and 4 h after incubation (μg mL−1), V is the liquid volume (mL), M is the fresh weight of sediment slurry (g) and Δt is the incubation time interval (4 h).

Statistical analysis

Relationships between the soil characteristics and the denitrification potential were examined in SPSS (version 13.0) using Spearman’s rank correlation.

Results and Discussion

Response of denitrification to sediment depth

At 0–15 cm depth, the sediment denitrification potential declined greatly with increasing depth; the denitrification potential changed little below 15 cm (Fig. 1). The highest denitrification potential was in the upper 5 cm of sediment (Fig. 1), which accounted for approximately 34% of the total for the 0–30 cm depth.

Figure 1.

 Response of the denitrification of pond sediment to depth over a 4-h incubation at 25°C. Error bars indicate the standard deviation of the mean of four replicates.

Previous studies have reported that denitrification in sediment occurs mainly in the top 2–5 cm (Livingstone et al. 2000; Saunders and Kalef et al. 2001; David et al. 2006; Allen et al. 2007), where there is abundant inline image, organic carbon and anoxic conditions (Seitzinger 1988). There was a similar phenomenon in the present study. With increasing depth, the denitrification potential usually decreased owing to the declining organic carbon content in the sediment (García-Ruiz et al. 1998; Groffman et al. 2002); however, denitrification still occurred in sediments at greater depths, although to a lesser degree. The denitrification potential in the upper 5 cm only accounted for 34% of that over 0–30 cm depth, comparable to Inwood et al. (2007) who found that denitrification activity was highest in the top 5 cm of benthic sediment, which accounted for >88% of the total denitrification. Thus, we concluded that the sediment denitrification potential would be underestimated if only the upper 5 cm sediment was used to determine sediment denitrification. Although the denitrification potential of sediment has been studied at different depths, comparisons of sediment depth and denitrification are rare (Inwood et al. 2007; Livingstone et al. 2000). In the present study, there was a significant correlation between denitrification and sediment depth, and the relationship fitted a logarithmic curve (Fig. 1; y = −0.01 ln x + 0.04, R2 = 0.86, n = 24, P < 0.01).

Response of denitrification to temperature

As the temperature increased from 6 to 18°C the denitrification rates showed a weak response and did not significantly change (Fig. 2). However, a further temperature increase from 18 to 30°C dramatically increased the corresponding denitrification rates, significantly above those recorded at 6–18°C. There was a significant correlation between denitrification potential and temperature, fitting an exponential curve (Fig. 2; y = 0.0007 e0.4131x, R2 = 0.83, n = 15, < 0.01).

Figure 2.

 Response of the denitrification potential of pond sediment to temperature over a 4-h incubation. Error bars indicate the standard deviation of the mean of three replicates.

Temperature has long been recognized as an important factor in regulating microbial denitrification. Nowicki (1994) indicated that denitrification rates initially increased slowly with increased temperature and then greatly over a range of 0–25°C, according to an exponential relationship, which is similar to our temperature response curve at 6–30°C. At low temperature, microbial activity was weak and the temperature effect was not obvious (Elefsiniotis and Lib 2006); this phenomenon was also reflected in our study, suggesting that denitrification microbes in subtropical regions are adapted to high temperatures (Saunders and Kalef 2001). Our results showing that denitrification of sediment incubated at 18–30°C was significantly higher than that at 6–18°C may imply that denitrification plays an important role in the removal of inline image in pond sediment throughout most of the year, as the studied site is located in a subtropical region with a mean temperature >18°C from April to November.

Comparison of the C 2H2inhibition technique with inline image disappearance

The C2H2 inhibition technique was compared with the inline image disappearance method for calculating sediment denitrification over a temperature range of 6–30°C (Fig. 3). There was a significant correlation between the two methods (R2 = 0.608, n = 14, < 0.01). However, the C2H2 inhibition technique tended to give lower rates, approximately 25% of the rates recorded using the inline image disappearance method; this is consistent with the results of Kewei et al. (2008) who showed that the denitrification rate of river sediments determined using the inline image disappearance method was higher compared with the technique of nitrous oxide production with C2H2 blockage.

Figure 3.

 Regression of the C2H2 inhibition (y) and the inline image disappearance (x) methods in determining the denitrification potential. The data point in the square deviates excessively from the other two, owing to a leakage problem during incubation, and is therefore not considered in the regression.

Rudolph et al. (1991) reported that the C2H2 inhibition technique for denitrification measurement was useful only when inline image was >10 μmol L−1 and when 100% of inline image was recovered as N2O in the slurries. The inline image concentrations in the present experiment before and after 4 h incubation were approximately 109 and 78 μmol L−1, respectively, much higher than the 10 μmol L−1 recorded by Rudolph et al. (1991), but the denitrification potential estimated using the C2H2 inhibition method was still lower than estimated using the inline image disappearance method. A possible reason for this difference is that there was a lag phase between inline image disappearance and N2O production. Ellis et al. (1996) showed that the inline image concentration significantly and rapidly decreased within 2 h of anaerobic incubation, whereas N2O began to evolve after 2 h. A significant portion of inline image was probably converted to NO at an early phase of incubation, and might be largely responsible for the higher inline image disappearance rate compared with the N2O production rate using the C2H2 inhibition technique (Kewei et al. 2008). An alternative possibility leading to the underestimation of the denitrification potential by the C2H2 inhibition technique may be leakage of N2O during incubation. Gases have been known to penetrate silicon rubber stoppers, although the amount is likely to have been small. Moreover, the C2H2 inhibition technique does not always effectively block conversion of N2O to N2 during denitrification (Tomaszek and Czerwieniec 2000). The main possibilities include: 1. C2H2 scavenging of NO, leading to underestimation of denitrification (Bollmann and Conrad 1997). It has been reported that oxidation of NO to NO2 is enhanced by the presence of C2H2 at concentrations >0.1%. If this happened in our study, then the C2H2 may have resulted in scavenging of part of the NO produced as an intermediate in the denitrification sequence and thus could not be further reduced to N2O. Consequently, the denitrification rate was underestimated. 2. Slow diffusion of C2H2 into fine-textured and/or saturated soils and sediments (Jury et al. 1982). As C2H2 has a low solubility in water, we may not obtain perfect inhibition effects in pond sediment, which may lead to underestimation. 3. Rapid decomposition of C2H2 by C2H2-degrading microbes (Flather and Beauchamp 1992). C2H2 can act as an organic carbon source for microbes in soil, resulting in the scavenging of C2H2 and incomplete inhibition of N2O reduction. In addition, under anaerobic conditions, dissimilatory inline image reduction to inline image and microbial immobilization can also be responsible for inline image consumption (Kaspar et al. 1981), which may result in overestimation of the denitrification potential for the inline image disappearance method.

The C2H2 inhibition method allows for a large number of samples to be analyzed, is easily manipulated and has high sensitivity. Previous studies have also used C2H2-based methods to learn much about the environmental regulation of denitrification, for example, control by oxygen, inline image, carbon, soil moisture, pH and other factors. C2H2 can be 50–100% effective in blocking conversion of N2O to N2 in aquatic ecosystem sediment (Binnerup et al. 1992; Seitzinger et al. 1993). Our study indicated that the C2H2 inhibition method was only 25% effective, suggesting that it was not useful for studying the denitrification of pond sediment.

Acknowledgments

This work was financially supported by the National Natural Science Foundation of China (No. 40721140018).

Ancillary