*2.2. Nutrient Release Trial*

This experiment was a laboratory-scale leaching study to investigate N mineralisation kinetics using coarse-textured sand (Table 1). Nitrogen was added at a rate of 291 kg/ha as either CropUpTM (equivalent to 8.675 mg/g soil) or as ammonium chloride (NH4Cl) (equivalent to 0.9274 mg/g soil). This rate is higher than that used in the runoff trial (50–100 kg N/ha) to ensure that effects of dissolution and transformation were measurable. Sand with no N applied served as a control. Leaching was undertaken in 50 mL polypropylene centrifuge tubes with a small hole (5 mm diameter) drilled through the base to allow for drainage of the leachate. A disc of Whatman glass fibre paper was placed inside the tube to cover the hole so as to avoid soil loss during leaching.

Air-dried (<2 mm) sand (equivalent to 40 g on an oven-dry weight basis) was placed into the respective centrifuge tube. Initially, approximately 80–85% of the total soil mass was added. The respective treatments were then added to the sand surface, with CropUpTM as solid material and NH4Cl in 1 mL of solution. The remaining 15–20% of the sand was then placed above the soil/treatment interface. The sand was lightly compacted by dropping the tube 10–15 times vertically from a 2 cm height. This packing procedure resulted in the treatments residing ≈10 mm below the sand surface.

Each sand column was leached with the equivalent of 1.5 pore volumes of 0.005 M CaCl2 (≈12.7 mL) on days 1, 3, 10, 17, 24, 33 and 42 of the study. Dilute CaCl2 was chosen, rather than water, as the leaching solution to simulate the soil solution's ionic composition and ionic strength, and to minimise the likelihood of, albeit a low amount of, mobilised clay clogging the filter disc.

Leaching occurred over a 2–3 h period (under gravity/free drainage), and the leachates were collected in 70 mL polypropylene containers. After drainage ceased, each leaching tube was placed under a slight vacuum for 3 to 5 s to remove excessive solution from the base of the tube and to avoid generating anaerobic conditions. The drainage collected under vacuum was added to the gravity drainage, and the volume of the leachate was estimated by weighing.

All the leachates were filtered (<0.45 μm cellulose acetate filter membranes) and stored frozen before being analysed colormetrically for NOx and NH4 using a microtiter plate reader (BioTek EPOCH<sup>2</sup> Microplate Reader) at a wavelength of 625 and 540 nm, respectively. Following each leaching event, the tubes were re-capped (lids remained loose to maintain aerobic conditions) and incubated at 25 ◦C.

After the final leaching (day 42), the remaining available mineral N was extracted from the sand using 2 M KCl (2 h end-over-end shaking at a 1:5 soil-to-solution ratio). After extraction, the supernatant was removed, centrifuged (10,000 rpm for 20 min), filtered (<0.45 μm) and stored frozen prior to NOx and NH4 analysis. The concentrations of NOx and NH4 were corrected by subtracting the amount of each ion retained in the entrained solution.

## *2.3. Statistical Analysis*

General linear mixed models (GLMs) were used to analyse the rainfall simulator data, using restricted maximum likelihood (REML) in GenStat (2018). Residual plots were used to confirm the assumptions of homogeneous variances and low skewness, with flow rate as the standardising covariate. The treatments were initially analysed as 13 discrete levels, and subsequently as the factorial structure of N-rates by N-sources. A post hoc comparison between the adjusted means was performed using protected least significant difference testing at a significance level of 5% (*p* < 0.05). Nitrogen release data were analysed using analysis of variance (general linear model; *Statistix* version 10). The significant differences among the main treatments were separated by LSD (*p* < 0.05). The relationships between cumulative leachate NH4 and NOx with time were adequately described (r2 for NH4 = 0.85–0.95 and r2 for NOx = 0.75–0.89) by an equation of the form:

$$N\_i = k \times \ln t + N\_0 \tag{2}$$

where *Ni* = leachate N (mg NH4 or NOx) at time *i*, *t* = time (days), *N*<sup>0</sup> = leachate N (mg NH4 or NOx) at time 0 and *k* = first-order rate coefficient (d<sup>−</sup>1).

#### **3. Results**

A rainfall simulation trial was undertaken to evaluate the impacts of combined inorganic and organic fertiliser on nitrogen runoff losses compared with conventional inorganic fertiliser ((NH4)2SO4) alone. A nitrogen release experiment was concurrently undertaken to investigate N mineralisation kinetics.

## *3.1. Nitrogen Release from Organic and Inorganic Sources*

A leaching experiment was undertaken to investigate the N mineralisation kinetics in the coarse-textured sand used in the rainfall simulation trial. The results show that negligible NH4 and NOx were leached over the 42-day period in the control treatment (no N applied), confirming the very low mineral N status of the sand (Figure 1). The leachate from the CropUpTM treatment contained very low amounts of NH4 (<0.7 mg), which were often not significantly (*p* < 0.05) different from that found in the control. For example, leachate NH4 initially increased from 0.09 mg on day 1 to 0.67 mg on day 3, after which leachate NH4 decreased steadily to background (control) levels. Leachate NOx from CropUpTM remained very low over the initial 3 days, but increased to 0.17 mg by day 10, and 0.28 mg by day 17. By day 24, the amount of NOx in the leachate was not significantly different from that found in the control.

**Figure 1.** Leachate (**a**) NH4 and (**b**) NOx, and cumulative leachate (**c**) NH4 and (**d**) NOx as a function of time. Bars represent LSD values (*p* < 0.05). The solid and dashed lines in (**c**) and (**d**) were predicted using Equation (2).

The majority (>95%) of the mineral N leached from the NH4Cl treatment was in the NH4 form with little conversion to NOx (Figure 1). Of this NH4, ≈96% leached from the sand column within the initial 10 days of the study (corresponding to 2–3 pore volumes of leachate). Levels of NOx ranging from 0.11 to 0.16 mg were measured between day 17 and day 24 (corresponding to 3–4 pore volumes), but declined to background levels with increased leaching. The distributions of NH4 and NOx in the leachate showed that the majority of NH4 was readily leached from the sand, but a small amount of residual NH4 was nitrified during the later stage of the study.

Over the 42-day incubation period, 0.26, 1.26 and 9.64 mg of NH4 and 0.33, 0.68 and 0.58 of NOx leached from the control, CropUpTM and NH4Cl treatments, respectively (Figure 1c,d). The amount of cumulative mineral N (NH4 + NOx) that leached from the control, CropUpTM and NH4Cl treatments was 0.59, 1.94 and 10.22 mg, respectively (Table 3). The ammonium and NO3 release rate coefficients (Equation (2)) for the control, CropUpTM and NH4Cl treatments were 0.062, 0.302 and 0.884 d−1, and 0.073, 0.187 and 0.137 d<sup>−</sup>1, respectively.

After 42 days, much of the (2M KCl-extractable) mineral N was present as NH4, with negligible NOx, irrespective of the treatment (Figure 2). Mean NH4 concentrations ranged from 12.49 mg/kg in the control to 13.60 mg/kg in the NH4Cl treatment, but these were not significantly different from each other. Mean NOx concentrations ranged from ≈0 mg/kg in the control and NH4Cl treatments to 1.05 mg/kg in the CropUpTM treatment, with the majority of extracted NOx present in the pore water.

**Figure 2.** Mean 2M KCl-extractable NH4 and NOx, and mineral N (NH4 + NOx) concentrations for the control, CropUpTM and NH4Cl treatments. Parameters with the same letter are not significantly different at *p* < 0.05 from each other.

The amounts of mineral N (i.e., NH4 + NOx) recovered for the control, CropUpTM and NH4Cl treatments were 1.14, 2.58 and 10.84 mg, respectively (Table 3). Of these amounts, 2M KCl-extractable NH4 was ≈0.5 mg for all treatments, suggesting that little to no NH4 from either CropUpTM or NH4Cl was retained by the sand. This small quantity of NH4 may represent a background concentration that cannot be displaced from the cation exchange sites by K-NH4 or Ca-NH4 exchange.
