High-Frequency Nitrate required separate injections of monoammonium phosphate and Ca2 in order to avoid precipitation of Ca32. All fertigations lasted 24 h, starting and ending at 7:00 AM, and injection of the non-MAP fertilizers took place between 11:00 PM and 3:00 AM. The soil at the site is classified as a Milham sandy loam derived from granitic and sedimentary alluvium, deep and well-drained. In an existing experiment , five blocks are laid out as a randomized complete blocks design, with eight strip-plot treatments of 15 trees. Treatments assessed had K applications of 224.17 kg ha 1 and P applications of 74.46 kg ha 1. Four blocks were studied for N2O surface emissions, and three of these included soil gas and solute measurements.Static gas chambers were used, of 20.3 cm diameter PVC piping 11 cm high placed over PVC collars installed 5 cm deep directly after fertigation, with aluminum insulation, rubber bottom seals, and manual internal fans. Two drippers were sampled in each plot per event, and different drippers were studied at each fertigation event; those studied were situated about 2 m from the nearest tree trunk, to represent average temperature, moisture and root presence. Chambers were positioned at 4 distances from drippers, along a transect into the center of the alley, at 0, 20, 50 and 90 cm. The maximum extent of the wetting zone at surface rarely exceeded 60 cm from drippers. Emissions from each collar were used to describe emissions from rings around the dripper,hydroponic growing systems with borders midway between distances. Three gas samples were taken over 10-min intervals. To calculate fluxes the ideal gas law was used as by Parkin and Venterea including ambient temperatures from a CIMIS weather station 2.5 km from the orchard, and an elevation of 120 m, corresponding to 0.985 atm of pressure.
Nonlinearity was not apparent with these short flux times.Surface emissions of N2O were expected to derive almost entirely from N fertigations and the first subsequent irrigations in the Standard treatment , with strong seasonal dependence that informed the sampling schedule . Output rates of drippers were checked. Gas samples were taken around peak emissions on the first day after fertigation, and thenceforth from 8:00–14:00 to represent daily averages. Emissions for statistical comparisons were the sums of observed measurements per plot, considered as describing a whole day. However, in order to be comparable to Standard, emissions from the HF applications required interpolations of unmeasured events, which were averages of each measured event’s emissions with the previous or the next. Previous experience of the patterns of emission demanded a different approach in order to estimate total growing season emissions per treatment. First day emissions were calculated assuming that N2O emissions increased linearly from baseline at 7 A.M. until peak ambient temperature, which was close to the time of measurement, then a Q10 adjustment was made using hourly ambient temperatures for the remainder of the day and night . Both of these manipulations sought to avoid overestimation. Q10 values were derived from seasonal diurnal measurements. To estimate the emissions over two additional days after sampling, an exponential curve of decline was fit to measured points averaged by treatment, as emissions = y0 + a, which had the best fit to data. Baseline rates were calculated by treatment to describe time periods when the upper soil was dry, starting 5–7 days after each fertigation, where the curve of decline ended. These were extended until the first day of fertigation for emissions calculations.Soil gas, moisture, temperature and N-distribution profiles were assessed in 3 experimental blocks. 1/800 OD brass tubing was used to take gas samples from depths of 5,10, 15, 20, 30, 45, 60 and 80 cm in the soil, at 20 cm from drippers.
The tubing was reinforced with steel wire, for insertion, after which the wire was withdrawn and replaced with a second, close-fitting internal brass tube of 3/ 3200 OD to minimize internal air volume. Tubes were topped with septa consisting of Nalgene tubing of 1/800 inner diameter injected with a dual plug of silicone and butyl rubber; butyl rubber blocks N2O diffusion while silicone maintains a good seal through syringe use. A volume of air corresponding to the tube volume was extracted before sampling. 5-mL gas samples were taken from 5 and 10 cm depth, and 10-mL samples from 15 cm and greater depths. For samples at 5 and 10 cm, tubes with closed ends and lateral cuts near their tip were inserted at 45 to the soil surface, allowing the surface soil to be tamped down at the entrance point without compacting the soil to be sampled. On this sandy loam, samples could be taken 5 h after irrigation without apparent resistance due to soil water. Gas samples were stored in glass vials with silicone sealant applied over the butyl septum before evacuation to 45 mTorr. All samples were analyzed for N2O using a 63Ni electron capture detector , and larger vials were tested for CH4 and CO2 using a flame ionizing detector , both on the same gas chromatograph . Simultaneously, a custom-built thermo couple probe of diameter 3/1600 took soil temperature at 5, 10, 15, 20, 30, 45 and 60 cm . A neutron probe provided measurements of volumetric water content at 20, 30, 45, 60 and 90 cm, using aluminum access tubes, and a soil moisture count curve fitted to measurements from this soil . A Theta probe was used to measure uvol in the upper 10 cm. Texture and bulk density were tested at each site at 0–20, 20– 30, 30–45, 45–60 and 60–90 cm. To compare the vertical distribution of dissolved NH4 + and NO3 in the soil after fertigations, high-flow ceramic soil solution samplers were installed at 15, 30 and 60 cm depth, 20 cm from the dripper laterally. Evacuation down to 600 mm Hg of vacuum was accomplished with an automotive brake pump, and samples were collected 6 h later, but on the third days after fertigation, samples adequate for analysis were only extracted from 38% of samplers, with fewer in summer, and very few in general from 15 cm depth.
Beginning in June, exchangeable NH4 + and NO3 were also assessed using soil samples from 0 to 5, 12.5–17.5 and 27.5–32.5 cm on each day. Approx. 20 g fresh weight soil was extracted in 80 mL of 2 M KCl, with one hour of shaking. These and the samples of soil solution were run for colorimetric analysis of NH4 + and of NO3.N2O concentrations by depth were combined with moisture readings, bulk density and texture to estimate the soil diffusion coefficient for nitrous oxide using the Unified Diffusivity Model— Buckingham Burdine Campbell . The Campbell parameter “b” was estimated using the clay fraction . Net production or consumption of N2O was estimated for each measured point in the soil except the deepest, using the one-dimensional mathematical model of Yoh et al. and field measurements described above.The soils were collected in late August after a month of irrigations without fertilizer. One replicate per plot was used, and results analyzed as an RCB consistent with the field experiment. pH was tested with 1 g:2 mL slurry in deionized distilled water, using samples from 0 to 20 cm depth at 20 cm from drippers. Treatment effects on nitrification rate were tested with the same samples using an ammonium oxidation assay citing Berg and Rosswall. Treatment effects on denitrification were tested with the same samples using a modified denitrification enzyme activity procedure with N2 as a flushing gas, with glucose, and without chloramphenicol. 5% head space acetylene, generated from calcium carbide,hydroponic farming was added to a round of duplicate samples; in the first test on the soils, 10% head space acetylene had appeared to impede the entire denitrification process. Head space gas samples were nonlinear by 60 min, so data from 20 and 40 min were used. The microcosms were also tested at 24 h to describe Denitrification Potential ; by 48 h, N2O declined in some microcosms. Soils from UAN treatments were tested for N2O production over a 36-h incubation at 3% O2 and 50% WFPS according to the method of Zhu et al. with some modifications. Samples of 15 g dry weight underwent a 24-h pre-wetting at 25% WHC, then were put into 180 mL Erlenmeyer flasks with butyl rubber stoppers and twice evacuated to 1000 mTorr and flushed for a minute with N2. Flasks were then injected with O2 for a 3% O2 head space, then fertigated to 50% WHC with NH4 + or NO3, with and without acetylene, and pure water . 5 mL head space samples assessed N2O production rates through measurements at 5 and 15 h; by 25 h they had declined in NH4 + microcosms, although N2O production continued.Estimated growing season N2O emissions were greatest from HF UAN fertigation, followed by Standard UAN, and lowest from the HF NO3 treatment, but comparisons between the two UAN and the two HF treatments only showed a significant difference between the HF systems . The failure to find differences in the Standard UAN—HF UAN comparison may have been due to high variability within and between blocks for the Standard UAN plots; in the UAN comparison the significance of blocking was p < 0.69 vs. p < 0.12 in the HF comparison. CVs over the measured period were 52% in Standard, 24% in HF UAN and 27% in HF NO3. Emissions from the irrigations following Standard UAN fertigations totaled 26% of the treatment’s growing-season emissions. In previous work, little effect had been seen in subsequent irrigations, even in loam soils . In the HF treatments such residual effects were indistinguishable from effects of the next fertigation.
There were important seasonal effects; higher temperature caused earlier and higher peaks of N2O emission, as well as higher event totals .The average precipitation at the site is 18 cm. The winter of 2013–2014 which followed our work was the driest on record in California; the experimental site received only 2 cm of precipitation and so was not monitored for emissions. However, previous work in the orchard , monitoring UAN-derived N2O emissions from static sprinklers at similar application rates, included a winter with 12 cm precipitation. Of the 730 g N2O-N yearly average emission, those data allowed an estimate that 18.5 g N2O-N ha 1 were emitted in winter. The latter quantity, if it had been seen in all the treatments studied here, would increase estimated N2O emissions by 3.7–8.5%. At the same time, Mediterranean perennial crop systems usually see more winter rainfall, and are capable of supporting cover crops in the winter. In such cases, post-season emissions can be expected to constitute a larger portion of the total and cover crop management, can be an important control, as studied in vineyards .To analyze the transport and fate of applied fertilizer, NH4 + and NO3 in soil solution were sampled by suction lysimeters and in soil by KCl extraction . Urea from Standard UAN was hydrolyzed to NH4 + and further oxidized to mobile NO3 in the soil profile, which by the third day led to 14x higher total NO3concentrations at 60 cm depth under Standard UAN than under HF NO3 . Low dissolved NO3 in the HF NO3 treatment may be ascribed to higher diffusion to outer parts of the wetting zone , greater root uptake of nitrate through mass flow soon after fertigation, and the observed high retention of nitrate in the upper part of the soil . NH4 + in solution near surface agreed with differences in application strength , but such differences were not apparent in KCl-extractable NH4 + , which described much greater quantities.In general, a better understanding of the spatial origin and fate of N2O under different conditions should lead to improved fertilization and fertigation practices. In this investigation, net N2O production by depth was calculated from soil gas concentrations, soil diffusion parameters, temperatures and water contents along a one-dimensional vertical profile at 20 cm from the dripper. A 1-D model cannot give a mass balance for N applied with water from a point source; unaccounted lateral diffusion will probably amount to a net loss and lead to underestimates of production. However this problem was judged to be ameliorated by conditions observed before the experiment.