Maximum concentrations of N2O were observed in the soil at 10–15 cm, shallow depths where lateral diffusion away from the dripper could only account for a small loss of N2O compared with losses to the atmosphere. It was also observed that water tended to flow laterally, past the surface wetting front, through a sandy horizon above a clayey horizon which begins at around 50 cm depth, also diminishing net lateral gas diffusion. The model is highly sensitive to variability in N2O concentration data, and at the plot level the calculations showed fluctuations between production and consumption by depth which were not plausible or consistent.Where possible, curves were fit to the concentration data of the form N2O conc = ae, which provided the dc/dz terms. Furthermore, concentrations of N2O measured at 5 cm were generally much lower than at 10 cm and led to frequent estimates of net consumption near surface. It was deemed likely that these samples had been contaminated by atmospheric air. Production of N2O at 10 cm was therefore calculated with reference to ambient N2O concentrations at 0 cm instead of the 5 cm concentration data. Measurements spaced several hours apart determined that an average 4.4% of total production was accounted for by change in concentration over time during the measured days. Ultimately no treatment statistics could be reported with the profile production data, but the model revealed general patterns and treatment effects on the depths of N2O production. The highest emissions, which were seen in summer and fall,hydroponics growing system were associated with the most consistent patterns of N2O distribution in the soil profile. Results from Days 2, 3 and 4 after fertigation, not shown, had similar distributions to those measured on Day 1, although with a slightly higher fraction of N2O concentrated in the deeper soil, 40–60 cm.
The relative stability in depths of production was seemingly contradictory to the changes in N distribution seen in soil solution ; but it was notable that distribution of extractable N , showed less change at the same points . N2O concentration patterns under most days and treatments were bimodal, with a shallow peak at 10–15 cm and a deeper peak around 45–60 cm, in the zone of higher clay content. The deeper peaks were sometimes strong 3–4 days after fertigation, especially in the Standard UAN treatment, illustrating the deeper distribution of N under higher rates of UAN application. Calculations in UAN treatments after winter typically showed points of highest production at 10–15 cm depth, usually underlain directly by the points of greatest consumption, at 15–20 cm . The calculations for 20 and 30 cm might underestimate production, because of more significant lateral diffusion of N2O around 30 cm, where WFPS generally declined . Production was seen at the lower peaks around 45 cm, but calculations suggested that N2O produced in these lower peaks was generally consumed before reaching surface , consistent with the findings of Neftel et al. . This helps to explain why emitted N2O was less per unit applied in Standard UAN than in HF UAN. Calculations in HF NO3 profiles generally showed much lower net N2O production/consumption than the UAN treatments. This is credited to the more even distribution throughout the soil of applied NO3, vertically and laterally, which led to low concentrations. Production profiles further suggest that a high proportion of the N2O produced in this treatment was consumed before it could be emitted from the surface. Overall, surface emissions of N2O decreased more quickly over the days following fertigations than did soil gas concentrations and calculated in-soil production rates, suggesting greater importance of production near surface during the first and second days. Under the driest conditions, seen on Day 3 after fertigation in late summer, increased N2O concentrations at 60 and 80 cm were concurrent with the lowest post-fertigation surface emissions. Calculations of N2O production for that date showed consumption at 45 cm in both HF treatments , supporting the conclusion that N2O produced deeper was being consumed at points immediately above, as well as possibly diffusing downwards.
Although the averaging of soil gas profiles by treatment limited the options for statistical analysis, the factors driving N2O production in the soil could still be assessed. Multiple linear regressions of surface emissions and of production at 15, 30 and 60 cm were carried out using calculated N2O production per treatment per day at those depths, and the corresponding averaged NH4 + in solution, NH4 + in soil extracts, NO3 in solution, NO3 in soil extracts, WFPS and temperature. Treatments were pooled because the dataset was limited within each treatment and the differences seen when HF NO3 was separated were minor. Regressions had little predictive capability at 30 and 60 cm depth. Nevertheless, it was notable that WFPS had negative coefficients at both depths, indicative of more complete denitrification with greater soil moisture. At 15 cm, the Adjusted R2 was only 0.14 but several alternative analyses gave better predictions. When excluding negative production values, an Adj. R2 of 0.58 was seen, which rose to 0.68 when reduced to extractable NH4 + , WFPS and temperature. If production at 10 cm was averaged with that at 15 cm, most negative values were eliminated, and using all data and variables the Adj. R2 was 0.21, or 0.26 with extractable NH4 + , NO3 in solution, WFPS and temperature. These results caused some questioning of the calculations of N2O production and consumption, which were volatile even in averaged forms, so regressions were carried out with soil gas concentrations as well. At 15 cm, all variables regressed to Adj. R2 of 0.26; reduced to NH4 + , WFPS and Temperature, the Adj. R2 was 0.32. Concentration averaged between 10 and 15 cm had an Adj. R2 of 0.41, while reduced to NH4 + , WFPS and Temperature, the Adj. R2 was 0.49. Regression of surface emissions followed the same pattern, being compared to NH4 + and NO3 in soil extracts at 2.5 cm depth, WFPS and temperature, where NO3 was found insignificant. The adjusted R2 of this regression is not reported because it is less complete than the analyses above. The superior predictive capability of extractable NH4 + at 15 cm and near surface was unexpected, since it is usually assumed that only the NH4 + in solution is available for microbial consumption .
However, little relevant investigation has been done in soils and the question can be raised whether microbial foraging on clays can desorb ammonium .The persistence of input effects on the functioning of the soil microbial community is an important agro-ecological concern. Here several assays of nitrification and denitrification capacity tested for persistent treatment effects which could influence N2O emissions. Soils were collected in late August after a month of irrigations without fertilizer. Treatment differences were of interest, not the comparison of assay results to field rates. The most ready metric of a soil’s denitrification response to NO3 amendments is its denitrification enzyme activity ,hydroponic equipment designed to assess soil process rates before they are affected by the synthesis of additional enzymes. Since fertigation applications make a large amount of NO3 available in a short time, the preevent DEA of a soil may play a significant role in denitrification derived N2O emissions. Results showed very similar N2O production by the two HF treatments in a DEA assay, which were significantly higher than Standard UAN . Over 24 h, characterized as Denitrification Potential , this initial difference was persistent, although it lost statistical significance. Given that drip fertigation saturated zones are not entirely dissimilar from the conditions of these assays, it was expected that DEA and DP modified with acetylene might also suggest differences in the product ratio of denitrification in the field treatments. Results were inconclusive, with widely dispersed values. Rates of ammonium oxidation to nitrite, as an index of nitrification potential, supported the importance of frequency and rate of NH4 + application, HF UAN > Standard UAN > HF NO3, but differences were only significant between the HF UAN and HF NO3 treatments . Strict chemoautotrophs typically dominate nitrification in cropped soils , and their numbers are more likely to be affected by availability of NH4 + than are the heterotrophs largely responsible for N2O emission through denitrification. Higher amounts of available nitrite are known to stimulate nitrifier denitrification and associated N2O losses , so a persistent effect of NH4 + application on ammonium oxidation to nitrite could increase N2O emissions under HF fertigation. Ammonium oxidation and DEA assays are predicated upon standard conditions, the former being oxic, open, shaken slurry, and the latter completely anoxic.
Actual oxygen availability in drip zones may cover a wide range between those points, but is expected to be limited. Little data is available, but Gil et al. found 4.97% O2 in the sampled soil air of a clay loam in an avocado orchard under drip. It can be assumed that many surfaces within larger aggregates would have lower O2 , being well suited for nitrifier denitrification, which takes place at <5% O2, while denitrification requires <0.05% O2 . It was therefore deemed useful to test the persistent effects of HF fertigation on potential soil production of N2O at 3% O2 . The only treatment differences were in microcosms with NO3 amendments , where N2O was presumably derived mainly from denitrification inside aggregates, supporting DEA results . The lack of HF treatment effects with NH4 + may be due to high rates of adsorption on soil surfaces expected with this N source , leading to gradual liberation. Nevertheless, emissions of N2O with NH4 + amendments were higher than those with NO3, confirming the large potential contribution of nitrifier denitrification from drip zones. The alternative explanation, being a general, rapid turnover from nitrifier-produced NO2 and NO3 to denitrifier produced N2O, has not consistently been supported by isotopic studies in laboratory . Assessments of N2O/ product ratio using acetylene in DEA, DP, and 3% O2 incubation assays did not give robust support to the hypothesis that greater microbial capacity for nitrification and/ or denitrification should correlate to a higher portion of complete reduction of N to N2 . It must be noted that N2O is more likely to be reduced to N2 when NO3 is limited , which it was not in the DEA test and DP tests. Further, the reduction of N2O to N2 dominates under anoxic conditions , which were not prevalent in the 3% O2 test. The factors affecting the “completeness” of nitrifier denitrification to N2 have been little studied and may be distinct from those affecting denitrifier denitrification. Lastly, tests of residual NO3 suggested that acetylene may have slightly inhibited NO3 reduction. The comparison of N2O from HF UAN with a HFNO3 -based treatment including Ca2 raises the question of whether differences may be ascribable to the opposite pH effects of the fertilizers. HF Ca2 + KNO3 did produce a significantly higher pH than HF UAN within 6 months of the treatment’s inception . This could partly explain lower N2O emissions from the HF NO3 treatment , but the effectis likely not a strong one because all were in neutral range . Our observation of 2.0 greater N2O emissions from HF UAN than from HF NO3 agrees well with Abalos et al. , who saw 2.4 greater N2O emissions from urea than from calcium nitrate in a drip-fertigated melon field in Spain. The greater predictive capacity of extractable NH4 + over NO3 provided evidence of a high contribution of nitrifier denitrification to N2O emitted in the field. This was supported by laboratory tests of our field soils at 3% O2, and concurred with findings by Vallejo et al. , as well as by Sanchez-Martin et al. , who calculated that with dripfertigated ammonium sulfate, 45% of N2O came from nitrification. Considering both field and laboratory data, frequency effects in the application of UAN were only seen in nitrate denitrification rates and in N present at 60 cm depth. Nitrifier capacities do not seem to have been affected, due perhaps to the adsorption of fertilizer NH4 + and its gradual release over time. Still, rates of nitrifier denitrification in the field may have seen concentration effects, as a corollary of frequency differences.