Size and Persistence of Nitrous Oxide HotSpots in Grazed and Ungrazed Grassland

Nitrous oxide (N2O) emissions from agriculture contributed an estimated 60% of the global total in 2005. In the UK, grassland soils account for 30% of total emissions, 22% of which are estimated to come from urine and dung patches. These patches are possible sources of ‘hot-spots’ (area ca. 1 m2) of N2O fluxes. Spatial and temporal heterogeneity of N2O hot-spot fluxes were investigated in three grassland fields (grazed with dairy cows (DG), grazed with young stock (YG) or cut for silage (SC)) using gas sampling chambers surrounding historic hot-spots to establish their size. Fluxes from old dung and urine patches were measured, as well as freshly applied dung and urine to simulate the creation of hot-spots. Potential chemical and physical drivers were also measured. Large spatial variability of N2O fluxes was seen in all three grassland fields. Mean N2O fluxes for the historic hot-spots in the grazed fields (DG and YG) were significantly greater than (SC). The mean N2O fluxes in DG and YG (117.9 and 243.5 ng N m-2 s-1) were 15 to 30% greater than for SC. Soil temperature (15 20 °C) was the most significant driver of N2O production with a 1°C rise in soil temperature increasing emissions under DG and YG. N2O fluxes were enhanced by the fresh dung but not by urine. However, in the urine treatment, the nutrient input increased the microbial respiration response for the CO2 flux. Hot-spot N2O emissions from old urine and dung patches were persistent several months after application.


Introduction
Nitrous oxide (N 2 O) has been recognised as a major contributor to anthropogenic warming (Houghton & Ding, 2001), with a global warming potential estimated at 298 times that of carbon dioxide (CO 2 ) (IPCC, 2007).N 2 O is also broken down in the stratosphere in reactions that deplete stratospheric ozone (Crutzen, 1981;Ravishankara, Daniel, Portman, 2009).Agricultural sources are mainly from soil and contributed ~60% to the global total in 2005 (Smith et al., 2007;Reay et al., 2012), with 30% of the UK agricultural emissions from grassland (Fowler, Hargreaves, Skiba, & Bower, 1999;Skiba et al., 2012;Sozanska, Skiba, & Metcalfe, 2002).Production and emissions of N 2 O are increased by the addition of nitrogen (N) in the form of mineral or organic fertilisers and residues.Under grazing, concentrations of N from cattle urine and dung deposits are high and both increase emissions (van der Weerden, Lu, de Klein, Hoogendoorn, Littlejohn, Rys, 2011); the dung also contains a readily available carbon source.Within the UK, Yamulki, Jarvis, and Owen (1998) estimated that 22% of emissions of N 2 O from grassland originated from urine and dung patches.
The main N component of urine is urea (NH 2 CONH 2 ) making up 70% of its N composition (Oenema, Velthof, Yamulki, & Jarvis, 1997).Urea from urine deposited on grassland soil is transformed to ammonium (NH 4 + ) by hydrolysis within 24-48 hours through several microbial pathways that provide important sources of N 2 O emissions (Clough, Ledgard, Sprosen, & Kear, 1998;Monaghan & Barraclough, 1993;Wrage, Velthof, van Beusichem, & Oenema, 2001).An estimation of N 2 O losses from urine deposition to grassland by Williams, Ineson, Cowards (1999) indicated that approximately 8% of the annual urine-N was lost within the first 24 hours.Soil denitrification from urine additions is considered to be dependent on the amount of NH 4 + produced (Carter, Klumpp, & Le Roux, 2006).Denitrification can also be stimulated by carbon compounds, as C is mobilised from roots scorched by the urine (Monaghan & Barraclough, 1993;Ambus, Petersen, & Soussana, 2007).A small number of studies of the variation of N 2 O emissions as a result of dung and urine patches on grassland indicate that soil type (as a combination of physical or chemical properties) could also influence emissions (Ball, Horgan, Clayton, & Parker, 1997;Velthof, Jarvis, Stein, Allen, & Oenema, 1996).However, most investigations have focused on the affect of a limited number of driving variables.Van Groenigen, Velthof, van der Bolt, Vos, and Kuikman. (2005a) considered soil compaction and seasonal emissions, both in the field and in pot experiments, from urine and dung patches and found that water filled pore space (WFPS) enhanced N 2 O emissions in the short term and that soil compaction would increase emissions over a season (van Groenigen, Kuikman, de Groot, Velthof, 2005b).These factors of compaction and/or urine and dung patches gave 'hot-spots' of N 2 O emissions of 0.5 -2 m in diameter.
Increased soil electrical conductivity from the accumulation of soluble salts to the soil, as occurs under urine patches, has been investigated as a proxy for N 2 O emissions.Adviento-Borbe, Doran, Drijber, Dobermann (2006) found that emissions of N 2 O increased with soil salt concentrations when WFPS was 90%.Soil surface pH has been shown to be less important.Clough and Kelliher (2005) found, after the addition of urine, an initial rise in soil pH but there was no significant difference in N 2 O flux or soil pH compared to the control eleven days after application.In contrast, Orwin et al. (2010) found that urine maintained the soil surface pH above 7 for 29 days after application and thereafter the soil pH decreased below the control.Even though increased soil temperature should result in an increase in soil microbial activity, Yamulki et al. (1998) observed a correlation between N 2 O emission and soil temperature only at 10 cm depth during the autumn (mid-September).
The stimulation of N 2 O flux relative to CO 2 flux has not been a common focus in other work associated with grassland.Scott, Ball, Crichton, and Aitken (2000) considered the effect of sewage sludge application to grassland on N 2 O and CO 2 emissions and found short term temporal variability related to rainfall events and diurnal temperature change.Owing to the marked temporal variability of both gases, we considered it useful to measure these fluxes during this work.Our investigations had three aims.The first was to measure spatial heterogeneity of N 2 O and CO 2 fluxes from gas sampling chambers in three grassland fields (one grazed by dairy cattle (DG), one grazed by young stock (YG) and one cut for silage (SC)) around locations of regular, elevated N 2 O fluxes, to establish the size and extent of any historic hot-spot and the soil properties that may account for the fluxes.The second was to identify possible recent sources of hot-spots of N 2 O fluxes from measurements taken under old dung or urine patches, along with measured soil physical and chemical properties.The third was to create fresh hot-spots by applying dung and urine to grassland to establish any immediate change of N 2 O and CO 2 flux.

Site Description
The measurements took place in three grass fields on the Crichton Royal Farm situated on the south side of Dumfries, south west Scotland (55°02.5'N,3°35.3'W).One was grazed by dairy cattle (DG), high intensity management, another was grazed by young stock (YG), medium intensity of grazing management and the third was cut for silage (SC).The three fields all received different amounts of N during the six months prior to the study (Table 1).All were on freely draining sandy loam brown forest soils of the Crichton series (Eutric Cambisol; FAO classification).In each field there was already a group of 6 closed static chambers, 40 cm diameter, in place for the regular measurement of N 2 O fluxes.Average annual totals of fluxes from these sites were very high, 21.2 kg N 2 O-N ha -1 yr -1 , with the main driving variable being total N applied (Rees et al., 2013).During our investigation all the gas flux measurements and soil sampling were taken over the course of two days in June 2007.
Table 1.Fertiliser application (kg ha -1 ) to the three fields (Dairy Grazed (DG), grazed by Young Stock (YG) and Silage Cut (SC)) in the previous six months and the rainfall monthly totals (mm)

Spatial Variation around a Historic Hot-Spot Chamber
In the first experiment, two of six regularly-assessed static gas sampling chambers (labeled 'a' and 'b') that gave consistently the highest N 2 O fluxes (taken to be on 'hot spots') in each field were surrounded with 2 concentric rings of 5 chambers at a radius of 1 m and 2 m (Figure 1).This configuration was chosen in an attempt to distinguish whether hot-spots of N 2 O emission were more associated with historic N deposition from dung and urine residues than with hot spots due to variations in topography that can cover areas up to several m 2 (Ball et al., 1997).The configuration would also help to establish the size and location of any hot-spot.The pattern of the chamber placement was the same around each original hot-spot chamber in all three fields.Figure 1.Gas sampling chambers pattern used around an historic hotspot chamber (unnumbered).1 m and 2 m refer to the measured distances between chamber centres.

Old Urine and Dung Patches
In the field grazed by young stock (YG), four chambers were placed on localised areas of enhanced grass growth that had not been grazed but showed no indication of dung deposit and were assumed to indicate areas of old urine deposition, four chambers were placed on similar non-grazed grass that showed residual dung and four chambers were on areas that had average grazing.

Fresh Urine and Dung Patches
Three gas sampling chambers were embedded in the soil with freshly collected dung or synthetic urine applied.A further eight chambers had no additions and were placed around the field grazed by dairy cows, as controls.
The rates of application of the dung and urine represented realistic deposition rates by cattle.The synthetic urine was made up according to Carter et al. (2006) as a solution of 0.7g N l -1 , and was applied to the inside of the gas sampling chambers at a rate equivalent to 244 kg N ha -1 .The dung was applied at a rate equivalent to 200 kg N ha -1 (McGechan & Topp, 2004).
All emission rates were measured using manually closed static chambers 0.2 m high polypropylene cylinders of 0.4 m diameter enclosing an area of 0.13 m 2 embedded to a depth of 5 cm in the soil (Scott, Crichton, & Ball, 1999).The chambers were left for 4 hours after installation before any flux measurements were taken.Prior to each sampling, several samples of ambient air were taken at the time of chamber closure and samples were taken from each chamber at the end of the closure period (1 h) in gas tight vials.Gas samples were analysed for N 2 O and CO 2 in the laboratory using gas chromatography (Scott et al., 1999).
After the initial gas sampling, soil was collected from within each chamber to a depth of 0-20 cm and used for the determination of soil mineral N (NO 3 -and NH 4 + ), water content and pH.Nitrate and NH 4 + were measured by continuous flow colorimetric analysis of 1M KCl extracts prepared from field moist soil using a soil to extractant ratio of 1:5.Soil pH values were determined on suspensions of 10 g fresh soil in 25 ml water.The soil moisture measurements were made with a Delta-T capacitance probe to a depth of 5 cm, and soil temperature was measured at a depth of 10 cm.
The bulk density of the soil at 0-10 cm soil depth was measured by knocking metal rings of 7 cm diameter and 10 cm depth vertically into the ground and digging out the intact core.These soil cores were dried at 105°C for 24 hours and the bulk densities were used in conjunction with the soil moisture measurements to calculate the Water Filled Pore Space (WFPS).

Statistical Methods
Means were compared using two-sample t-tests.Pearson correlations were used to assess the association between variables.Data were assumed Normally distributed except for N 2 O fluxes, which were log-transformed before evaluation.All tests were performed at the 5% significance level.

Spatial Variation around a Historic Hotspot Chamber
The measurements from the spatial variation sampling around the original hot-spot gas sampling chambers in the three fields are shown in Table 2 for dairy grazed (DG), Table 3 for silage cut (SC) and Table 4 for grazed by young stock (YG).Fluxes of N 2 O measured from the chambers surrounding the original hotspot chambers in each field were highly variable (Figure 2).The largest individual N 2 O fluxes of 2883.3 and 819.3 ng N m -2 s -1 , were from DG and were accompanied by corresponding high CO 2 fluxes (Figure 3).The mean N 2 O fluxes for all the sampling chambers in the two grazed fields DG (243.5 ng N m -2 s -1 , P<0.05) and YG (117.9 ng N m -2 s -1 , P<0.001) were significantly higher than the ungrazed field SC (7.05 ng N m -2 s -1 ).
Table 2. Measured and mean values from the two sets of 10 gas sampling chambers (a and b) surrounding a chamber with high flux (original a + original b) from the dairy-cattle grazed field (DG)   The coefficient of variation (CV%) of N 2 O flux measurements from the pooled data within each field was greatest for the two sets of chambers in DG (241.6% and 179.0%) and least for the set of chambers 'a', in YG (95.0%).The CV of N 2 O flux among the other set of chambers in YG (170.5%) was closer in magnitude to that in the other grazed field, DG.Although the fluxes in SC were much lower, the CV's for the sets of sampling chambers in SC were also close to those in the grazed fields (157.9 and 128.1%).
In most sets of chambers in the three fields the greatest N 2 O flux (Figure 2) was not the original hot spot.The number of chambers giving higher N 2 O fluxes than the original hot-spot chamber varied from one (DG 'a') to four in YG 'b'.Chambers DG 'a', SC 'a', SC 'b' and YG 'a' had higher fluxes at 1m radius from the original chamber, in contrast with chambers DG 'b', and YG 'b' where N 2 O fluxes were greater in one or more chambers at 2 m radius from the original chamber.N 2 O fluxes from the chambers around YG 'a' also indicated a possible second hotspot of N 2 O for chamber 'a.7', opposite the cluster of chambers around the original hotspot.
CVs for CO 2 fluxes for YG (35.0 and 41.2%) were significantly smaller than DG (70.4 and 89.3%) and SC, (67.4 and 110.0%), but still indicated a high heterogeneity in relation to microbial activity.
The greatest mean CO 2 fluxes were from the YG chambers (177.9 ng C m -2 s -1 ) and were approximately double those in the DG chambers and 1.7 times greater than the mean for the SC chambers.
The mean soil temperature measurement was significantly lower at SC (P<0.001), the only other variable measured that showed a comparable pattern to the N 2 O fluxes, across the three fields.The only other significant correlation for the log of N 2 O was with NO 3 -(R 0.87, P<0.05) for the YG (chambers 'a').On average the highest mean N 2 O flux for this part of the study (Table 5) was from the gas sampling chambers placed over the old urine patches, 162.2 ng N m -2 s -1 , and was significantly greater (P<0.05),than 15.1 and 6.2 ng N m -2 s -1 for the normally grazed areas and dung patches, respectively.The mean CO 2 fluxes followed a similar pattern to the mean N 2 O fluxes, with the greatest mean values in the former urine patches, 150.1 ng C m -2 s -1 .The N 2 O fluxes from the normally grazed areas were most closely correlated, in general, with the other measured variables but were only significant (R 0.88, P<0.05) between log N 2 O flux and CO 2 flux.Nevertheless, the mean WFPS was lowest for the old urine patches (82.9%) compared to the normally grazed areas (89.8%) or dung chambers (90.4%).

Old Urine and Dung Patches and Effects on N 2 O Fluxes
As expected the mean soil NH 4 + (18.5 μg N g -1 ) concentration in the old urine patches was the highest, with the mean of the old dung areas being slightly lower (16.5 μg N g -1 ) compared to the normally grazed areas (12.6 μg N g -1 ).The greatest mean NO 3 -content was under the old dung chambers (8.6 μg N g -1 ) with the old urine and normal grazed chambers giving the same mean concentration of 6.5 μg N g -1 .
The range of mean soil pH for the three sets of chambers was narrow, between 5.50 and 5.60.Unsurprisingly, considering the potential N addition from urine and dung, statistically significant correlations between NH 4 + and NO 3 -were found in the normally grazed grass, old urine and dung patches (R 0.97, R 0.98 and R 0.98 P<0.05), respectively.

Effect of Fresh Dung and Synthetic Urine Application on N 2 O Fluxes
The addition of fresh dung to the gas sampling chambers gave mean N 2 O (1574.2 ng N m -2 s -1 ) and CO 2 fluxes (208.3 ng C m -2 s -1 ) that were significantly greater (P<0.01)than those from the urine addition chambers (53.7 ng N m -2 s -1 , 13.3 ng C m -2 s -1 ) (Table 6).
Table 6.Measured and mean values from gas sampling chambers and related soils variables on control, fresh synthetic urine addition or fresh dung addition to gas chambers (DG) field Fluxes of log N 2 O and CO 2 correlated significantly for both the fresh dung (R 0.95, P<0.05) and urine chambers (R 0.99, P<0.005).Contrary to expectation, the mean N 2 O and CO 2 fluxes for the chambers with no additions were greater than the fresh urine.The mean soil NH 4 + content in the fresh dung chambers (22.6 μg N g -1 ) was also more than double that for the fresh urine (Table 6).
Overall, the fresh urine chambers showed the largest number of statistically significant relationships between log N 2 O flux and the soil driving variables, including temperature (R 0.99, P<0.01), NO 3 -(R 0.99, P<0.01) and pH (R 0.97, P<0.05).
Within the fresh urine application chambers there were further statistically significant relationships between CO 2 flux and temperature (R 0.97, P<0.05), soil moisture (R 0.95, P<0.05) and NO 3 -(R 0.99, P<0.01).These relationships were not detected in the fresh dung or the chambers with no additions.

Historic Hotspot Chambers
Higher N 2 O fluxes were consistently observed in the 12.5 m 2 area surrounding the original hotspot chambers suggesting that the area of elevated N 2 O or hot-spot was greater than 2 m in diameter.The area within 1 m radius (~ 3 m 2 ) was rather larger than in previous work where a typical urine patch was between 0.2 m 2 (Hayes & Williams, 1993) and 1.1 m 2 (Moir, Cameron, Di, Fertsak, 2010), or a dung patch of ~ 0.2 m in diameter ( Van der Weerden et al., 2011).However the 2 m radius area 'hot spot' (~12 m 2 ) suggests that the N 2 O fluxes may result from either an amalgamation of contiguous hot spots formed from a number of historic urine and/or dung patches or that they have a topographical element that can produce large flux differences due to changes in soil carbon along with soil moisture and structure (Ball et al., 1997).
Fluxes were all positive though markedly different (Figure 2); the mean N 2 O fluxes for the DG and YG chambers were 34 and 17 times greater than the mean of the SC chambers.Mean N 2 O fluxes for these grazed fields were considerably greater than other grassland systems in Europe (Rees et al., 2013) and the maximum fluxes were comparable to data for another grazed site in the east of Scotland that also received large fertiliser N inputs (Flechard et al., 2007).Saggar et al. (2004) found that N 2 O emissions for ungrazed pastures were less than 10% of those from grazed pastures with the difference due to the addition of the excreta from the grazing animals.However, in these fields the reduction in fluxes for the ungrazed fields was much more dramatic with a 97% reduction compared to DG and 94% in comparison with YG.This could be due to N from excreted emissions, though, the overall mineral N content measurements show that as they were all > 10 mg g -1 dry soil and unlikely to be limiting in both the grazed and ungrazed fields (Conen, Dobbie, Smith., 2000).Nevertheless, there was a significant difference in N 2 O fluxes between the grazed (DG (P<0.05), the YG (P<0.001)) and the SC field.The mean soil temperature was the only driving variable that differed significantly between fields; the mean at SC was 1.67°C (P<0.001)lower than at DG and 2.45°C (P<0.001)lower than YG.The mean log N 2 O flux data were significantly correlated (R 0.83, P<0.01) with mean soil temperature in the three fields.With a good supply of mineral N, temperature would be a significant driver at these WFPS values (Conen et al., 2000).Alves et al. (2012) demonstrated that diurnal N 2 O fluxes were explained largely by the soil temperature changes.Scott et al. (2000), working on similar soils, found that soil surface temperature effected N 2 O emissions.The difference in temperature between the three fields could account for the differences in N 2 O flux as Dobbie & Smith (2001) showed increased emission rates (Q 10 ) of N 2 O fluxes of 2.3 as temperatures increased between 12 and 18°C in soil cores after nutrient addition.This has important implications for any soil temperature rises due to weather extremes or progressive warming in the future.
Unlike Schaufler et al. (2010), no significant correlation was found overall between CO 2 flux and soil temperature (R 0.2) for all the chambers, however after removal of the very high measurement from the SC set of 'b' chambers R increased to 0.37 and was significant (P<0.01)indicating that the microbial populations producing the N 2 O and the CO 2 are both influenced by temperature more than just the general respiration of the microbial population (Mathieu et al., 2006).
However, four out of the six sets of chambers gave significant (P<0.001)R values over 0.78 for correlations between the log of N 2 O and CO 2 (Figure 3); such correlations probably indicate enhanced soil microbial activity from nutrients, especially C inputs from the grazing animals' dung and urine, around carbon nuggets (2-10 cm diameter) as suggested by Parkin (1987).
Spatial variation of N 2 O emissions can increase as a consequence of anaerobic conditions, related to increased respiration of the soil microbes, following the introduction of a source of decomposable organic matter (Christensen & Tiedje, 1998), such as an old dung patch which can form C nuggets.
The large CV's of the N 2 O fluxes (95-242%) within sets of chambers have been observed in other experiments and have been linked to general soil heterogeneity (Yanai et al., 2003;Aarons, O'Conner, Hosseini, & Gourley, 2009).Choudhary, Akramkhanov, and Saggar (2002) also found large spatial variation in N 2 O fluxes, with a mean CV of 120%, for a grazed permanent pasture which again reflected the high soil heterogeneity and was attributed to highly localised (<1 m) concentrations of organic matter, nitrate level or water filled pore space.Mohammad, Roobroeck, Van Cleemput, and Boeckx (2011) also found that soil biological, physical and chemical properties (cores taken 25 m apart) were highly heterogeneous across a field and related this to inherent soil physical and chemical properties, mainly aggregate diameter, total C, bulk density and the microbial community structure (fungi/bacteria ratio).

Old Urine and Dung Patches
The highest mean log N 2 O flux (Table 5) was the old urine patches and was accompanied by the highest mean CO 2 flux, in contrast the lowest N 2 O flux was for the old dung patches.A lower mean soil temperature under the dung patches (Table 4) compared to the normally grazed chambers was accompanied by greater grass dry matter in the ungrazed old dung patches and this may help explain the reduction in emissions of N 2 O.The inverse correlations between log N 2 O and the CO 2 fluxes from the old dung patch chambers (R 0.89) may indicate that the carbon source from the dung reduced the microbial N 2 O flux activity as the dung patch aged.As Aarons et al. (2009) found, the addition of dung to soil increased the soil microbial biomass as a result of the increased addition of organic C and soil pH.This also occurred in the current study where pH was least acidic in the old dung patches (Table 4).
As Yamulki et al. (1998) found in measurements of N 2 O flux from urine and dung patches, the average flux from urine patches was more than five times greater than from dung, even though the total N in the dung was greater than urine.The difference in N 2 O flux between the old urine and dung patches could also have been influenced by the greater microbial availability of urea N from the urine, a form more easily used by the soil microbial groups responsible for N 2 O production (Mathieu et al., 2006).
The lower soil moisture contents for the old urine patches related to the more vigorous grass growth and depletion of available soil moisture in these areas due to avoidance by the grazing animals.Soil moisture for the old urine patches was significantly correlated with both soil NH 4 + (R 0.95) and NO 3 -(R 0.95), these were not correlated for the old dung patches.
Concentrations of soil NH 4 + were consistently greater than of NO 3 -for all the chambers, which was contrasted with the historic hot-spot chambers work reported above, where the NO 3 -concentrations were mostly greater.An explanation for this could be related to the greater age of the hot-spot patches and that the old dung and urine patches sampled in this experiment were more recent in comparison and there was still soil mineral NH 4 + to be transformed by microbial activity to mineral NO 3 -.

Effect of Fresh Dung and Synthetic Urine Application to Grassland
The mean N 2 O fluxes from chambers containing fresh moist dung were considerably greater than those containing fresh urine (Table 6).Allen, Jarvis, Headon (1996) observed a similar higher N 2 O flux from dung application than from urine but over a much longer time period (100 days) and attributed this to the more anaerobic conditions under the dung pat.Our data indicate that potential anaerobic conditions under dung can enhance N 2 O fluxes immediately after application.Overall, in nearly all chambers the concentration of soil mineral NH 4 + was less than the soil NO 3 -(Table 5).This is in conflict with the literature that suggests that soil NH 4 + would normally be greater in the urine application areas, due to more available and labile NH 4 + than from the dung (Yamulki et al., 1998, van der Weerden et al., 2011, Thomas, Logan, Ironside, & Bolton, 1988).However, as the measurements in this investigation were taken within hours of the urine and dung application, it is unlikely that the urea from the urine would be detected in that time.With no associated change in the pH with the application of the urine, as occurred in other studies generally after 1 day (Oenema et al., 1997), it would appear that the urea needs a greater time to move through the soil.The use of synthetic urine could also have suppressed the initial pH change due to a difference in the hydrolysis of the urea in the presence of hippuric acid in cow urine (Petersen, Stamatiadis, & Christofides, 2004).
Work by Saarijärvi and Virkajajärvi (2009) showed that the concentration of soluble organic N (SON) increased significantly within days after urine application, compared to a more modest increase and slower release for a dung pat.
Overall, the correlations of the variables measured from the addition of the dung and synthetic urine were more numerous and significant than from the historic hot-spot chamber measurements.This was as expected as there was a large immediate input of both N and C from the dung and urine that would effect both the microbial population and the soil chemistry.

Conclusions
The spatial heterogeneity of the N 2 O fluxes around the historic hot-spots was considerable and indicated hot-spot areas of up to at least 2 m in diameter.The smaller diameter patches indicated a contribution from historic urine or dung deposits but the larger diameter hot-spots would be too large to have been created by individual deposits alone and suggests either amalgamation of deposits and/or an influence of the micro-topography of the land surface that enhanced the N 2 O fluxes, this could include the aspect of the slope concentrating both N and C The differences in mean emissions between the grazing and silage conservation management was significant and should be an important consideration in the calculation of N 2 O emissions from grassland.The measurement of emissions with greater temporal and spatial replication from grassland with different management histories needs to be an important consideration for future inventory estimates.However, the enhancement of these differences by an increase in temperature between the fields of as little as 1.7°C to 2.5°C has possible implications for any increase in temperature through weather extremes or climate change by increasing radiative feedback by between 15% for young stock grazed and 30% for dairy stock grazed.
Historic hot-spots of dung and urine had more mineral NO 3 -than NH 4 + suggesting that potentially more of the NH 4 + was consumed by the soil microbial community and transformed to NO 3 -as they potentially have been primed by the addition of the large inputs of N and C (in the case of dung) with the microbial population having increased as a result.
The old visible dung patches were sites with higher CO 2 emissions and reduced N 2 O emissions compared to the old urine patches, indicating the organic C from the dung had enhanced the CO 2 emissions.The C had stimulated further activity as a result of the increased organic C input.
Additions of fresh dung of high moisture content and urine stimulated microbial activity as expected but the dung appeared to have a greater initial effect than the urine, potentially as a consequence of the NH 4 -N in the urine moving more slowly through the soil, coupled with an inhibitory effect of the NH 4 + and reduced transformation through nitrification to NO 3 -and a more rapid initial release of NH 4 + from the dung.

Figure 2 .
Figure 2. Magnitude of the N 2 O flux from the chambers located around the historic 'hot-spot' chamber in each of the three fields; Dairy-cattle grazed (DG) -flux ranges 4.3 to 2885 ng N m -2 s -1 , Silage Cut (SC) -flux ranges 0.2 to 128 ng N m -2 s -1 and grazed by Young Stock (YS) -flux ranges 3.2 to 570 ng N m -2 s -1 .The larger the area of the circle the greater the N 2 O flux

Table 3 .
Measured and mean values from the two sets of 10 gas sampling chambers (a and b) surrounding a chamber with high flux (original a + original b) from the Silage Cut field (SC)

Table 4 .
Measured and mean values from the sets of 10 gas sampling chambers (a and b) surrounding a chamber with high flux (original a + original b) from the Young Stock grazed field (YG) n/a samples not analysed.

Table 5 .
Measured and mean values from gas sampling chambers and related soils placed over normally grazed, old ungrazed urine patches (no indication of dung) and old dung patches(YG field)