Only a few studies are available on observed sinks for N2O. Table 15.1 presents selected examples of observed sinks for N2O. As explained earlier, most terrestrial and aquatic systems have the potential to act
Measurement technique Remarks
Fertilized grassland, poorly drained loamy soil, Berkshire, UK
Fertilized grassland, Siggen, south-west Germany
Unfertilized grassland, loam, poorly drained, Guelph, Ontario, Canada
Unfertilized grassland, clay loam, well-drained, Guelph, Ontario, Canada
Paragominas, Para State, Brazil
Fertilized rice paddy, early rice, top position on slope, clay soil, Yingtan, China
Spruce forest, sandy-loamy soil, Villingen, Germany
Annual mean sink of up to 3.2 mg N2O-N/m2/h
Some incidents of N2O uptake, with singular large uptakes of 41 mg N2O-N/m2/h Annual mean of 3 years
1.14 mg N2O-N/m2/h for April
1993 to March
Annual mean uptake 3 years of 0.8 (1996) to 1.7 (1997) mg N2O-N/m2/h
Net uptake in dry 2 years season of 2 mg N2O-N/m2/h Small seasonal sink 36 days of 0.63 mg/m2/h N2O-N in one of three plots 3-year mean sink of 3 years 1 mg N2O-N/m2/h with periods of emissions and uptake
Spruce forest, drained peat Mean sink of 0.3 mg 150 days soil, Scottish border, UK N2 O-N/m2/h
Open chamber, 3-6 gas N2O uptake at moderate to high water Ryden (1981, 1983)
samplings per week
Closed chamber, four replicates, 1-3 gas samplings per week content, low nitrate content (2-3 weeks after fertilization with 250 kg N/ha) and soil temperature >5°C It is not clear whether heavily fertilized grasslands with N2O uptake are an exception or not
Micrometeorological flux gradient method, one measurement per month
Micrometeorological flux gradient method, daily estimate based on hourly values Closed chamber, one measurement per month
Closed chamber, two replicates, sampling every 3-7 days
Closed chamber, eight replicates, hourly measurements in nine campaigns of 14 days in 3-year period in spring, summer, fall and winter Closed chamber, August-May with <1 measurement per week
Soil was a sink in 1993/94 possibly due to high rainfall in June; in 1992 and 1995 the site was a small source for N2O; fluxes were often below detection limit Soil was a net source in 1995
Glatzel and Stahr (2001)
Wagner-Riddle et al. (1997)
Maggiotto et al. (2000)
This 'active pasture' site is a net source Verchot et al. (1999)
of 2.85 mg N2O-N/m2/h Water management, N fertilizer
(122 kg N/ha) and rice straw (7 t with high C/N) application rate influenced N2O formation and uptake Unfertilized soil in N-limited spruce forest; N application (150 kg N/ha) caused system to turn into an N2O-N source of 0.9 mg/m2/h
N deposition of 48-96 kg N/ha/year caused system to turn into a source of N2O-N of 0.5-5.7 mg/m2/h
Deciduous forest, acid sand soil, Poppel, Belgium
Spruce-fir forest, podzol, Mt Ascutney, Vermont, and Mt Washington, New Hampshire, USA
Permanently water-covered freshwater riparian fen in Denmark
Mean 2-year sink of 7.4 mg N2O-N/m2/h
Mean sink of 1.12 mg N2O-N/m2/h (Mt Ascutney) and 0.23 mg N2O-N/m2/h (Mt Washington) In a 15 m transect from hill to fen, nitrate, oxygen and N2O concentrations declined to very low levels, while
21 months Closed chamber, six (631 replicates, weekly to days) event-based measurements
83 days Closed chamber, six
(May- replicates, one
August) measurement per
October Groundwater analysis 1 993 (N compounds) and sediment slurry incubations, including 15N experiments
High N2O-N uptake of 1.3 kg/ha over a 2-year period was attributed to low nitrification activity at pH H2O of 3.8; most negative fluxes were observed in wet periods (WPFS > 35%) Sink activity was attributed to low nitrification activity and thus low nitrate availability and high soil moisture content
Denitrification served as a sink for both dissolved N2O in groundwater recharging the fen, and the N2O produced within the riparian sediment
Goossens et al. (2001)
Castro et al. (1993)
Blicher-Matthiesen and Hoffmann (1999)
Three riparian sites in ;ium
15 Pre-alpine deep lakes
Antarctic surface waters (Bellighausen Sea and Drake Passage)
increased Closed chamber measurements show small fluxes, both negative and positive. Observed sink: -0.6 ± 0.4 mg N2O-N/m2/day Under complete anoxic conditions, water layers can be depleted of N2O North of the upwelling region, antarctic surface water formed from mixing of surface water and ice melt was moderately depleted of N2O
6 days in the period from August 2000 to Spring 2001
November/ December 2001
Closed chambers, three replicates, each box sampled five times over a total period of 72 min
Degree of N2O saturation based on measured concentrations in water measurements in water and atmosphere, using two sampling modes
In N-rich landscapes, riparian zones can act as (large) sources or small sinks for atmospheric N2O; optimal location for denitrification, relatively low nitrate
None of the lakes was found to act as a net sink for atmospheric N2O, because of poor mixing of anoxic layers and surface waters Source and sink areas were found in the studied area; overall there was a small negative flux
Dhondt et al. (2004)
Mengis et al. (1997)
N2O concentration as a temporary sink. In this overview, we focus only on studies in which a net sink was observed over a relatively long period of time.
The techniques and frequencies ofthe N2O flux measurements presented in Table 15.1 show a large variability. As has been discussed elsewhere (Bouwman et al., 2002a; Hofstra and Bouwman, 2005) this has a major effect on measured N2O and NO emissions and de-nitrification. Especially in experiments with a low frequency of gas sampling the uncertainty in flux estimates is large, and important flux 'events' may have been missed.
Sinks for N2O have been reported for terrestrial and aquatic systems, as well as for riparian zones. These systems are generally sources of N2O. However, when nitrogen availability and oxygen concentrations are low, soils, and in some cases lakes and certain regions of the ocean, could act as a sink for atmospheric N2O. We discuss a limited number of studies in which such N2O sinks were reported. In theory, soils are particularly powerful sinks, as illustrated by Blackmer and Bremner (1976), who showed by laboratory experiments that under conditions that are favourable for denitrification, the capacity of soils for N2O uptake is larger than their capacity for release.
Agricultural soils are usually fertilized, and therefore not considered to act as a sink for N2O. Nevertheless, some studies report episodes of uptake of atmospheric N2O in fertilized fields. For instance, uptake of N2O was observed in fertilized grasslands in the UK (Ryden, 1981, 1983) under the following conditions: high water content, very low NO3- availability and soil temperature exceeding 5°C. Sink activity was observed a few weeks after fertilization. Immediately after fertilization, the soils were considerable sources of N2O. Glatzel and Stahr (2001) also report on occasional N2O uptake in fertilized grassland soils, but do not answer to what extent these observations are exceptions. It is likely that, in general, fertilized soils are net sources of N2O on an annual basis. Wagner-Riddle et al. (1997) reported sink activity in unfertilized grassland in 1 out of 3 years. In the other 2 years, the site acted as a small source for N2O. Maggiotto et al. (2000) also observed sink activity in 2 out of 3 years in a grassland site. In a tropical 'active pasture', soils acted as N2O sinks during the dry season, and as sources during the wet season, leading to a net annual source (Verchot et al., 1999).
In rice paddy soils also, N2O uptake was observed by Xu et al. (1997; Table 15.1). They measured N2O fluxes in three rice paddy fields by closed chamber techniques. Of the three, a net uptake of N2O was measured in one field. Xu et al. (1997) suggest that this was related to the water management, the low fertilizer nitrogen inputs in this system and a high carbon/nitrogen ratio of rice straw return. Despite this observed sink, their results indicate that rice paddies are generally sources rather than sinks for N2O.
Forest soils are generally considered to be sources of N2O. However, some studies indicate that forest soils can in fact also be net sinks for N2O. For instance, Papen et al. (2001) report on net sinks in German forest soils and argue that increased atmospheric nitrogen deposition may change forest soils from net sinks to net sources.
Also Skiba et al. (1999) show that unpolluted forest soils can occasionally be a significant sink for N2O. In a drained peat area, they measured maximum uptake rates of -15 |ig N/m2/h. However, they do not extrapolate their hourly fluxes to longer-term fluxes, making it difficult to draw conclusions on the net uptake over a longer period of time. Also Goossens et al. (2001) report on relatively large N2O uptake in a Belgian forest soil. They observed negative N2O fluxes at low soil pH (3.8), indicating that nitrification is inhibited, during periods when the water-filled pore space exceeded 35%. Castro et al. (1993) observed occasional N2O uptake in forest soils during summer in north-eastern USA and attributed this to the low nitrification activity.
There are numerous examples of measurements where fluxes are below detection (Conrad et al., 1983; Slemr et al., 1984; Anderson and Poth, 1989; Papen et al., 1993; Jambert et al., 1994; Serca et al., 1994; Poth et al., 1995; Clemens et al., 1997; Butterbach-Bahl et al., 1998; Kilian et al., 1998; Petersen, 1999; Riley and Vitousek,
2000). In fact, these soils may have acted as small sinks as well as small sources. Similar to the observations of Wagner-Riddle et al. (1997) and Maggiotto et al. (2000), who measured sink activity in some years and emissions in other years, sites with small fluxes may change from sources to sinks depending on the environmental conditions.
N2O can be denitrified during leaching and runoff or in groundwater flows. For instance, Blicher-Matthiesen and Hoffmann (1999) studied NO-, oxygen and N2O concentrations in a 15 m transect from a hill to a permanently water-covered freshwater riparian fen. They observed that NO3-, oxygen and N2O concentrations in this transect declined to very low levels, while N2 concentrations increased. This indicates that, during transport from hill to fen, NO3- is denitrified and N2O reduced to N2. However, riparian zones can also be sources of N2O. In particular in nitrogen-rich regions, riparian zones may act as a source of N2O (Hefting et al., 2003). Dhondt et al. (2004) measured N2O fluxes in three riparian zones in nitrogen-rich landscapes using closed chambers. They observed small N2O fluxes, both positive and negative, and concluded that in these particular riparian zones the denitrification of NO3- was not leading to large emissions of N2O. However, they also argue that a better understanding of the controls of the N2O/N2 ratio in riparian zones that are meant to reduce nitrogen inputs to surface waters is important, to avoid increased N2O emissions as a result of water pollution control. In other words, riparian zones in eutrophic landscapes may be sources of atmospheric N2O rather than sinks. They also concluded that denitrification is controlled by the geomorphology of the river valley, rather than by the vegetation cover.
Whether lakes can act as a sink for atmospheric N2O is not clear. Many lakes have anoxic zones that are undersaturated with N2O. This is generally considered to be the result of N2O consumption by denitrifying bacteria. These zones could therefore act as a sink for N2O that is dissolved in the water. This N2O may be formed in the aquatic system or may have leached from terrestrial systems. It may also be of atmospheric origin and diffuse into the water from the atmo sphere. In the latter case, the lake can, in theory, be a sink for atmospheric N2O. This is why Knowles et al. (1981) hypothesized that eutrophic, anoxic lakes may be sinks for atmospheric N2O. However, observations by Mengis et al. (1997) do not support this hypothesis. Mengis et al. (1997) analysed oxygen and N2O concentrations in 15 prealpine lakes (Table 15.1). They observed that in the oxic waters, N2O concentrations typically increase with decreasing oxygen concentrations. They also observed that anoxic water layers are N2O-undersatu-rated and concluded that N2O is consumed in completely anoxic layers. Nevertheless, their results do not support the hypothesis that eutrophic anoxic lakes may act as a sink for N2O, because the surface layers in all the 15 lakes were supersaturated with N2O, because of poor mixing of the deep anoxic layers with the surface waters.
In the Antarctic Ocean, both source and sink areas were observed during an ocean flux study expedition in the spring of 1992 (Rees et al., 1997; Table 15.1). Whether or not an area acted as a source or sink was determined by hydrographical characteristics. Rees et al. (1997) hypothesize that upwelling water, supersaturated with N2O, can accumulate under the sea ice during winter. During spring, when the ice melts, this N2O-rich water is exposed to the atmosphere. The N2O-rich water is, however, mixed with N2O-depleted ice melt. As a result, a water layer can be formed that is undersaturated with N2O, acting as a seasonal sink for atmospheric N2O.
Further net N2O removal could occur in the relatively small volume of the oceans in which oxygen concentrations fall to near-zero levels at intermediate water depths, where denitrification is an important process. Results for the Somali Basin, however, suggest that upwelling regions may usually be a net source of N2O (De Wilde, 1999).
Was this article helpful?