Nitric oxide and nitrous oxide emission from Hungarian forest soils; linked with atmospheric N-deposition

Studies of forest nitrogen (N) budgets generally measure inputs from the atmosphere in wet and dry deposition and outputs via hydrologic export. Although denitrification has been shown to be important in many wetland ecosystems, emission of N oxides from forest soils is an important, and often overlooked, component of an ecosystem N budget. During 1 year (2002–03), emissions of nitric oxide (NO) and nitrous oxide (N₂O) were measured from Sessile oak and Norway spruce forest soils in northeast Hungary. Accumulation in small static chambers followed by gas chromatography-mass spectrometry detection was used for the estimation of N₂O emission flux. Because there are rapid chemical reactions of NO and ozone, small dynamic chambers were used for in situ NO flux measurements. Average soil emissions of NO were 1.2 and 2.1 μg N m⁻² h⁻¹, and for N₂O were 15 and 20 μg N m⁻² h⁻¹, for spruce and oak soils, respectively. Due to the relatively high soil water content, and low C/N ratio in soil, denitrification processes dominate, resulting in an order of magnitude greater N₂O emission rate compared to NO. The previously determined N balance between the atmosphere and the forest ecosystem was re-calculated using these soil emission figures. The total (dry+wet) atmospheric N-deposition to the soil was 1.42 and 1.59 g N m⁻² yr⁻¹ for spruce and oak, respectively, while the soil emissions are 0.14 and 0.20 g N m⁻² yr⁻¹. Thus, about 10–13% of N compounds deposited to the soil, mostly as and , were transformed in the soil and emitted back to the atmosphere, mostly as greenhouse gas (N₂O).     

Differentiation of nitrous oxide emission factors for agricultural soils

Nitrous oxide (N₂O) direct soil emissions from agriculture are often estimated using the default IPCC emission factor (EF) of 1%. However, a large variation in EFs exists due to differences in environment, crops and management. We developed an approach to determine N₂O EFs that depend on N-input sources and environmental factors. The starting point of the method was a monitoring study in which an EF of 1% was found. The conditions of this experiment were set as the reference from which the effects of 16 sources of N input, three soil types, two land-use types and annual precipitation on the N₂O EF were estimated. The derived EF inference scheme performed on average better than the default IPCC EF. The use of differentiated EFs, including different regional conditions, allows accounting for the effects of more mitigation measures and offers European countries a possibility to use a Tier 2 approach.     

Indirect nitrous oxide emission from a nitrogen saturated spruce forest and general accuracy of the IPCC methodology

The relevance of indirect N₂O emission is a controversial topic which is subject to much uncertainty. Only a small number of studies measure the indirect N₂O emission at the interface from soil to stream. In addition, the majority of studies undertaken only cover a short-term period (<1 year). Therefore, limited information is available regarding the influence of seasonal or event effects, nor is there much information as to whether indirect N₂O emissions are reflected by N₂O in soil solutions. The present study aimed at clarifying these two questions along with the general relevance of dissolved nitrous oxide. A wetness gradient involving soil solutions of different soil types and surface waters within an N-saturated forest catchment (3.2 ha) was monitored over a period of 1 year. N₂O concentrations in soil solutions (0.09–16.6 μg N l⁻¹) were affected by events such as dry–wet cycles but did not reflect to the actual, indirect N₂O emission at the soil-stream interface. It was assumed that N₂O emission was due to N transformation processes. The N₂O concentration at the spring was three times higher than the N₂O concentrations in the soil solutions. Nevertheless, indirect N₂O emission was still subordinate (<1%) to the direct emission of N₂O. The weekly amount of indirect N₂O emissions depended only on the stream flow rate (62% of the total annual amount). For this reason it was necessary to measure indirect N₂O emission at short intervals and at the interface between soil and stream over a longer time period. Our results and the results of the reviewed studies show that the default IPCC emission factor (EF5-g=1.5%) overestimates the indirect N₂O emission from ecosystems. The emission factor should therefore be lowered to about 0.1–0.3%. In addition, the results indicate that indirect N₂O emission is an insignificant pathway in the N cycle of most ecosystems. However, final judgement will depend on long-term studies.     

The effect of diet manipulation on nitrous oxide and methane emissions from manure application to incubated grassland soils

Changes to agricultural management, particularly of the nitrogen (N) input to farms, have great potential for mitigating emissions of N containing gases, especially the greenhouse gas nitrous oxide (N₂O). Manipulating diets fed to livestock is a potential method for controlling N excretion and emissions of greenhouse gases (GHG's) to the atmosphere. We selected three slurries derived from sheep that had been fed, either ensiled ryegrass (Lolium hybridicum), lucerne (Medicago sativa) or kale (Brassica oleracea) and applied them to a grassland soil from the UK in a laboratory experiment using a special He/O₂ atmosphere incubation facility. The resulting fluxes of N₂O, CH₄ and N₂ were measured, with the largest total N fluxes generated by the ryegrass slurry treatment (14.23 ryegrass, 10.84 lucerne, 13.88 kale and 4.40 kg N ha⁻¹ from the control). Methane was emitted only from the ryegrass slurry treatment. The isotopomer signatures for N₂O in the control and lucerne slurry treatments indicated that denitrification was the main process responsible for N₂O emissions.     

Intramolecular distribution of stable nitrogen and oxygen isotopes of nitrous oxide emitted during coal combustion

The intramolecular distribution of stable isotopes in nitrous oxide that is emitted during coal combustion was analyzed using an isotopic ratio mass spectrometer equipped with a modified ion collector system (IRMS). The coal was combusted in a test furnace fitted with a single burner and the flue gases were collected at the furnace exit following removal of SO(x), NO(x), and H2O in order to avoid the formation of artifact nitrous oxide. The nitrous oxide in the flue gases proved to be enriched in 15N relative to the fuel coal. In air-staged combustion experiments, the staged air ratio was controlled over a range of 0 (unstaged combustion), 20%, and 30%. As the staged air ratio increased, the delta15N and delta18O of the nitrous oxide in the flue gases became depleted. The central nitrogen of the nitrous oxide molecule, N(alpha), was enriched in 15N relative to that occupying the end position of the molecule, N(beta), but this preference, expressed as delta15N(alpha)-delta15N(beta), decreased with the increase in the staged air ratio. Thermal decomposition and hydrogen reduction experiments carried out using a tube reactor allowed qualitative estimates of the kinetic isotope effects that occurred during the decomposition of the nitrous oxide and quantitative estimates of the extent to which the nitrous oxide had decomposed. The site preference of nitrous oxide increased with the extent of the decomposition reactions. Assuming that no site preference exists in nitrous oxide before decomposition, the behavior of nitrous oxide in the test combustion furnace was analyzed using the Rayleigh equation based on a single distillation model. As a result, the extent of decomposition of nitrous oxide was estimated as 0.24-0.26 during the decomposition reaction governed by the thermal decomposition and as 0.35-0.38 during the decomposition reaction governed by the hydrogen reduction in staged combustion. The intramolecular distribution of nitrous oxide can be a valuable parameter to estimate the extent of decomposition reaction and to understand the reaction pathway of nitrous oxide at the high temperature.     

A UK inventory of nitrous oxide emissions from farmed livestock

A UK inventory of the nitrous oxide (N₂O) emissions from farmed livestock was compiled to identify areas where potential abatement practices may be effective. Where possible, emission factors based on direct experimental data gathered under UK conditions were used, but published data were used when this was not feasible, together with statistical information, which included details of numbers of animals within each category of a species, animal liveweights, number of days housed, excretal rates and volumes of manures in stores. Total N₂O emissions were calculated for each component of livestock production systems, i.e. animal houses, manure stores, following application of manures to land and during grazing. Emissions were also estimated from land used for forage conservation and tillage. Total annual N₂O emissions from UK farmed livestock, based mainly on 1996 animal census data, were estimated to be 38.27 kt. The two main terms were 22.66 kt N₂O from mineral fertilisers after application to soils and 5.61 kt N₂O from stored manures (mainly in the form of farmyard manure). Within buildings, poultry were the largest contributors of N₂O, 2.97 kt, followed by cattle, 1.62 kt. Within the total emissions from stored manures, cattle were the largest contributors of N₂O, 3.58 kt, followed by poultry, 1.86 kt. Dietary manipulation and a move from solid manure based systems to slurry based systems appear to be promising abatement practices.     

Biogenic nitric oxide emissions from cropland soils

Emissions of nitric oxide (NO) were determined during late spring and summer 1995 and the spring of 1996 from four agricultural soils on which four different crops were grown. These agricultural soils were located at four different sites throughout North Carolina. Emission rates were calculated using a dynamic flow-through chamber system coupled to a mobile laboratory for in-situ analysis. Average NO fluxes during late spring 1995 were: 50.9±47.7 ng N m⁻² s⁻¹ from soil planted with corn in the lower coastal plain. Average NO fluxes during summer 1995 were: 6.4±4.6 and 20.2±19.0 ng N m⁻² s⁻¹, respectively, from soils planted with corn and soybean in the coastal region; 4.2±1.7 ng N m⁻² s⁻¹ from soils planted with tobacco in the piedmont region; and 8.5±4.9 ng N m⁻² s⁻¹ from soils planted with corn in the upper piedmont region. Average NO fluxes for spring 1996 were: 66.7±60.7 ng N m⁻² s⁻¹ from soils planted with wheat in the lower coastal plain; 9.5±2.9 ng N m⁻² s⁻¹ from soils planted with wheat in the coastal plain; 2.7±3.4 ng N m⁻² s⁻¹ from soils planted with wheat in the piedmont region; and 56.1±53.7 ng N m⁻² s⁻¹ from soils planted with corn in the upper piedmont region. An apparent increase in NO flux with soil temperature was present at all of the locations. The composite data from all the research sites revealed a general positive trend of increasing NO flux with soil water content. In general, increases in total extractable nitrogen (TEN) appeared to be related to increased NO emissions within each site, however a consistent trend was not evident across all sites.     

Impact of carbon source on nitrous oxide emission from anoxic/oxic biological nitrogen removal process and identification of its emission sources

Wastewater treatment is an important source of nitrous oxide (N₂O), which is a strong greenhouse gas and dominate ozone-depleting substance. The purpose of this study was to evaluate the effect of carbon source on N₂O emission from anoxic/oxic biological nitrogen removal process. The mechanisms of N₂O emission were also studied. Long-term experiments were operated to evaluate the effect of three different carbon sources (i.e., glucose, sodium acetate, and soluble starch) on N₂O emission characteristics. And batch experiments, in the presence or absence of specific inhibitors, were carried out to identify the sources of N₂O emission. The ammonia-oxidizing bacteria (AOB) and denitrifiers community compositions under different circumstances were also analyzed based on which the underlying mechanisms of N₂O emission were elucidated. The conversion ratios of N₂O in reactors with glucose, sodium acetate, and soluble starch were 5.3 %, 8.8 %, and 2.8 %, respectively. The primary process responsible for N₂O emission was nitrifier denitrification by Nitrosomonas-like AOB, while denitrification by heterotrophic denitrifiers acted as the sink. Reactor with sodium acetate showed the highest N₂O emission, together with the highest nitrogen and phosphate removal ratios. Carbon source has a significant impact on N₂O emission quantity and relatively minor effect on its production mechanism.     

Deciphering biodegradable chelant-enhanced phytoremediation through microbes and nitrogen transformation in contaminated soils

Biodegradable chelant-enhanced phytoremediation offers an alternative treatment technique for metal contaminated soils, but most studies to date have addressed on phytoextraction efficiency rather than comprehensive understanding of the interactions among plant, soil microbes, and biodegradable chelants. In the present study, we investigated the impacts of biodegradable chelants, including nitrilotriacetate, S,S-ethylenediaminedisuccinic acid (EDDS), and citric acid on soil microbes, nitrogen transformation, and metal removal from contaminated soils. The EDDS addition to soil showed the strongest ability to promote the nitrogen cycling in soil, ryegrass tissue, and microbial metabolism in comparison with other chelants. Both bacterial community-level physiological profiles and soil mass specific heat rates demonstrated that soil microbial activity was inhibited after the EDDS application (between day 2 and 10), but this effect completely vanished on day 30, indicating the revitalization of microbial activity and community structure in the soil system. The results of quantitative real-time PCR revealed that the EDDS application stimulated denitrification in soil by increasing nitrite reductase genes, especially nirS. These new findings demonstrated that the nitrogen release capacity of biodegradable chelants plays an important role in accelerating nitrogen transformation, enhancing soil microbial structure and activity, and improving phytoextraction efficiency in contaminated soil.     

Mitigation of methane and nitrous oxide emissions from drained irrigated rice fields

One of the important cultural practices that affect methane and nitrous oxide emissions from tropical rice plantations is the water drainage system. While drainage can reduce methane emissions, it can also increase nitrous oxide emissions, as well as reduce yields. In this experiment, four different water drainage systems were compared in a rice field in central Thailand including: (1) continuous flooding, (2) mid-season drainage, (3) multiple drainage and (4) a local method (drainage was done according to local cultural practice) in order to find a system of drainage that would optimize yields while simultaneously limiting methane and nitrous oxide emissions. Methane and nitrous oxide emission were observed and compared with rice yield and physical changes of rice plants. It was found that drainage during the flowering period could reduce methane emission. Interestingly, nitrous oxide emission was related to number of drain days rather than the frequency of draining. Fewer drain days can help reduce nitrous oxide emission. The mid-season drainage and the multiple drainage, with 6.9% and 11.4% reduction in rice yield, respectively, had an average methane emission per crop 27% and 35% lower when compared to the local method. Draining with fewer drain days during the flowering period was recommended as a compromise between emissions and yield. The field drainage can be used as an option to reduce methane and nitrous oxide emissions from rice fields with acceptable yield reduction. Mid-season drainage during the rice flowering period, with a shortened drainage period (3 days), is suggested as a compromise between the need to reduce global warming and current socio-economic realities.     

Assessment of nitrous oxide emission from cement plants: Real data measured with both Fourier transform infrared and nondispersive infrared techniques

Nitrous oxide (N₂O) is the third most important greenhouse gas after carbon dioxide and methane, and contributes about 6% to the greenhouse effect. Nitrous oxide is a minor component of the atmosphere, and it is a thousand times less than carbon dioxide (CO₂). Nevertheless, it is much more potent than CO₂ and methane, owing to its long stay in the atmosphere of approximately 120 yr and the high global warming potential (GWP) of 298 times that of CO₂. Although greenhouse gases are natural in the atmosphere, human activities have changed the atmospheric concentrations. Most of the values of emission of nitrous oxide are still obtained by means of emission factors and not actually measured; the lack of real data may result in an underestimation of current emissions. The emission factors used for the calculation of N₂O can be obtained from the “Guidelines for the implementation of the national inventory of emissions” of the Intergovernmental Panel on Climate Change, which refer to all nations for the realization of their inventory. This study will present real data, measured in several Italian cement plants with different characteristics. The work also shows a comparison between N₂O concentration measured with in situ Fourier transform IR (FTIR) and the reference method EN ISO 21258 based on nondispersive IR (NDIR), in order to investigate the interfering compounds in the measurement with NDIR.

Implications:N₂O may arise as an unwanted by-product of nitrogen oxide (NO_x) abatement systems, in particular selective noncatalytic reduction (SNCR). Since it is applied in the cement plants, N₂O emission from cement industry is evaluated, with both FTIR and NDIR instrument. Several considerations emerged from the results. First of all, the emission from this industrial sector is not negligible, and for that reason N₂O concentration should be regulated; another observation is that the reference method based on the NDIR technique is not as selective as FTIR could be.     

Methane and nitrous oxide emissions from an irrigated rice of North India

Upland rice was grown in the kharif season (June-September) under irrigated condition in New Delhi, India (28 degree 40'N and 77 degree 12'E) to monitor CH4 and N2O emission, as influenced by fertilizer urea, ammonium sulphate and potassium nitrate alone (at 120 kg ha-1) and mixed with dicyandiamide (DCD), added at 10% of applied N. The experimental soil was a typic ustochrept (Inceptisol), clay loam, in which rice (Oryza sativa L., var. Pusa-169, duration: 120-125 days) was grown and CH4 and N2O was monitored for 105 days by closed chamber method, starting from the 5 days and 1 day after transplanting, respectively. Methane fluxes had a considerable temporal variation (CV=52-77%) and ranged from 0.05 (ammonium sulphate) to 3.77 mg m-2 h-1 (urea). There was a significant increase in the CH4 emission on the application of fertilizers while addition of DCD with fertilizers reduced emissions. Total CH4 emission (105 days) ranged from 24.5 to 37.2 kg ha-1. Nitrous oxide fluxes were much lower than CH4 fluxes and had ranged from 0.18 to 100.5 g m-2 h-1 with very high temporal variation (CV=69-143%). Total seasonal N2O emission from different treatments ranged from 0.037 to 0.186 kg ha-1 which was a N loss of 0.10-0.12% of applied N. All the fertilizers significantly increased seasonal N2O emission while application of DCD reduced N2O emissions significantly in the range of 10-53%.     

Decomposition of nitrous oxide on pillared clays

Alumina-pillared smectites have been found to abate nitrous oxide in the presence of methane. The results indicate that the yield of the reaction (N20 --> N2 + (1/2)O2) increases when pillared clays are exchanged with transition metals, single-pass conversion rates of >70% being attainable. In particular, when double exchanged (calcium and subsequently copper) alumina pillared montmorillonite/beidellite is used as a catalyst, de-N2O activity reaches a maximum, which is maintained even after 4 h of work at a space velocity of 5.5 h(-1). A mechanism for the reaction is suggested, which implies that N2O is first adsorbed by the catalyst and then decomposes through two different paths: catalyst oxidation and catalyst reduction. Such a redox process explains the kinetic data.     

COD/N对潜流人工湿地脱氮效能及N2O排放的影响

人工湿地因其低能耗、易管理的优点被广泛应用于水处理和生态修复,尤其在发展中国家有其适用性。为探讨COD/N对潜流人工湿地脱氮效能及氧化亚氮（N2O）的排放影响,分别构建5组微型人工湿地CW1~CK,进水化学需氧量/总氮（COD/N）分别为20：1、10：1、7：1、4：1和0：1。结果表明,湿地出水DO浓度均低于0.5 mg·L-1,不同的COD/N下无显著差异,但pH表现为随COD/N的升高而降低。进水COD/N为10、7和4的湿地中COD去除率在80%以上,而进水COD/N=20的湿地中COD去除率不足70%。5组COD/N湿地中氨氮（NH4+-N）的平均去除率依次为（61±16.6）%、（23±14.1）%、（30±12.5）%、（28±14.5）%和（74±7.0）%,在碳源充足或无碳源的条件下有利于NH4+-N的去除。除COD/N=0外,其余湿地中出水未检测到硝态氮（NO3--N）。湿地中总氮（TN）的去除变化与NH4+-N相似。5组湿地系统中,N2O的平均释放通量和累积排放量,分别在0.83~11.84 mg·（m2·h）-1和30.65~490.80 mg·m-2之间。COD/N=4的人工湿地系统中,N2O的平均释放通量和累计排放量显著高于其他处理（p2O形态去除的氮占系统去除TN的百分比为4.87%。COD/N对潜流人工湿地的脱氮和N2O减量具有调控作用。     

An inventory of nitric oxide emissions from soils in China

The emission of NO was parameterized using empirical relationships with landuse type, fertilization rate and soil temperature. Eight landuse types (including four arable lands) were considered. Fertilization rates were distinguished for different regions and crops. A typical summer period of July in 1999 was chosen for detailed calculations. The total NO emission in the July is 141.1 Gg N, with 73.7% from arable lands and 22.0% from grasslands. The highest emission intensity can be more than 40 ng N m(-2) s(-1) in the heavily fertilized North China Plain, and the average of the whole lands is 6.5 ng N m(-2) s(-1). The annual emission was roughly estimated to be 657 Gg N, about 11.7% of the global total (5600 Gg N, reported by IPCC in 2000), and about 12.5% of the anthropogenic origin in China. Our results were compared with some earlier findings, and uncertainties were discussed.     

Gaseous nitrogen losses from a forest site in the North Tyrolean Limestone Alps

Microorganisms are responsible for the mineralisation of organic nitrogen in soils. NH ⁺₄ can be further oxidised to NO₃ ⁻ during nitrification and NO₃ ⁻ can be reduced to gaseous nitrogen compounds during denitrification. During both processes, nitrous oxide (N₂O), which is known as greenhouse gas, can be lost from the ecosystem.

The aim of this study was to quantify N₂O emissions and the internal microbial N cycle including net N mineralisation and net nitrification in a montane forest ecosystem in the North Tyrolean Limestone Alps during an 18-month measurement period and to estimate the importance of these fluxes in comparison with other components of the N cycle. Gas samples were taken every 2 weeks using the closed chamber method. Additionally, CO₂ emission rates were measured to estimate soil respiration activity. Net mineralisation and net nitrification rates were determined by the buried bag method every month. Ion exchange resin bags were used to determine the N availability in the root zone.

Mean N₂O emission rate was 0.9 kg N ha⁻a⁻, which corresponds to 5 % of the N deposited in the forest ecosystem. The main influencing factors were air and soil temperature and NO ⁻₃ accumulated on the ion exchange resin bags. In the course of net ammonification, 14 kg NH ⁺₄ −N ha⁻ were produced per year. About the same amount of NO ⁻₃ −N was formed during nitrification, indicating a rather complete nitrification going on at the site. NO ^t-₃ concentrations found on the ion exchange resin bags were about 3 times as high as NO ^t-₃ produced during net nitrification, indicating substantial NO ^t-₃ immobilisation. The results of this study indicate significant nitrification activities taking place at the Mühleggerköpfl.

     

Effect of topography on nitrous oxide emissions from winter wheat fields in Central France

We assessed nitrous oxide (N₂O) emissions at shoulder and foot-slope positions along three sloping sites (1.6–2.1%) to identify the factors controlling the spatial variations in emissions. The three sites received same amounts of total nitrogen (N) input at 170 kg N ha⁻¹. Results showed that landscape positions had a significant, but not consistent effect on N₂O fluxes with larger emission in the foot-slope at only one of the three sites. The effect of soil inorganic N (NH₄⁺ + NO₃⁻) contents on N₂O fluxes (r² = 0.55, p < 0.001) was influenced by water-filled pore space (WFPS). Soil N₂O fluxes were related to inorganic N at WFPS > 60% (r² = 0.81, p < 0.001), and NH₄⁺ contents at WFPS < 60% (r² = 0.40, p < 0.01), respectively. Differences in WFPS between shoulder and foot-slope correlated linearly with differences in N₂O fluxes (r² = 0.45, p < 0.001). We conclude that spatial variations in N₂O emission were regulated by the influence of hydrological processes on soil aeration intensity.     

Nitric oxide emissions from soils amended with municipal waste biosolids

Land spreading nitrogen-rich municipal waste biosolids (NO₃⁻-N<256 mg N kg⁻¹ dry weight, NH₃-N∼23,080 mg N kg⁻¹ dry weight, Total Kjeldahl N∼41,700 mg N kg⁻¹ dry weight) to human food and non-food chain land is a practice followed throughout the US. This practice may lead to the recovery and utilization of the nitrogen by vegetation, but it may also lead to emissions of biogenic nitric oxide (NO), which may enhance ozone pollution in the lower levels of the troposphere. Recent global estimates of biogenic NO emissions from soils are cited in the literature, which are based on field measurements of NO emissions from various agricultural and non-agricultural fields. However, biogenic emissions of NO from soils amended with biosolids are lacking. Utilizing a state-of-the-art mobile laboratory and a dynamic flow-through chamber system, in-situ concentrations of nitric oxide (NO) were measured during the spring/summer of 1999 and winter/spring of 2000 from an agricultural soil which is routinely amended with municipal waste biosolids. The average NO flux for the late spring/summer time period (10 June 1999–5 August 1999) was 69.4±34.9 ng N m⁻² s⁻¹. Biosolids were applied during September 1999 and the field site was sampled again during winter/spring 2000 (28 February 2000–9 March 2000), during which the average flux was 3.6±1.7 ng N m⁻² s⁻¹. The same field site was sampled again in late spring (2–9 June 2000) and the average flux was 64.8±41.0 ng N m⁻² s⁻¹. An observationally based model, developed as part of this study, found that summer accounted for 60% of the yearly emission while fall, winter and spring accounted for 20%, 4% and 16% respectively. Field experiments were conducted which indicated that the application of biosolids increases the emissions of NO and that techniques to estimate biogenic NO emissions would, on a yearly average, underestimate the NO flux from this field by a factor of 26. Soil temperature and % water filled pore space (%WFPS) were observed to be significant variables for predicting NO emissions, however %WFPS was found to be most significant during high soil temperature conditions. In the range of pH values found at this site (5.8±0.3), pH was not observed to be a significant parameter in predicting NO emissions.     

Winter runoff losses of phosphorus from paddy soils in the Taihu Lake Region of South China

A winter wheat field plot experiment was conducted on two types of paddy soils, from November, 2000 to June, 2001 to assess P losses to its surrounding watercourses by runoff in the Taihu Lake Region. Commercial NPK compound fertilizer and single superphosphate fertilizer were applied to furnish 0, 20, 80, and 160 kg P ha(-1). The experiments consisted of six replicates of each treatment in Changshu site and four replicates in Anzhen site, with a plot size of 5x6 m2 in a randomized block design. Results revealed that the average concentration of dissolved P (DP), particulate P (PP), and total P (TP) in runoff water during the winter season was 0.13, 0.90 and 1.04 mg P l(-1) respectively, from P20 plots in Anzhen site. While it was 0.67, 1.08 and 1.75 mg P l(-1) respectively, from P20 plots in Changshu site. The seasonal TP load (mass loss) from P20 plot ranged from a low of 290.88 g P ha(-1)season(-1) to a high of 483.54 g P ha(-1)season(-1), with a mean of 382.29 g P ha(-1)season(-1) in Anzhen, but from 444.92 to 752.21 g P ha(-1)season(-1), with a mean of 539.13 g P ha(-1)season(-1) in Changshu. Both in Anzhen and Changshu PP represented a major portion of the TP lost in runoff, the average PP/TP was about more than 80% in P0 and P20 plot, but it was decreased with the increase of P rate. The average seasonal P loads (DP, PP, and TP) in Changshu were greater than in Anzhen although runoff volume in Anzhen (45 mm season(-1)) was more than in Changshu (36 mm season(-1)). This was probably associated with the differences of soil physical and chemical properties between the two sites. Phosphate fertilizer rate significantly affected P concentrations and P loads by runoff. Both the mean concentrations and the average seasonal P loads from the P80 plots were lower than from the P160 plots, but obviously higher than from the P20 and P0 plots. There was no significant difference found between the P20 plots and the P0 plots both in Anzhen and Changshu sites. It indicated that P loads by runoff would be greatly increased in 5-10 years due to the accumulation of soil P if 20 kg P ha(-1) applied each wheat season in this area.