**2. Materials and Methods**

#### *2.1. Soil Characteristics*

An incubation study was conducted on soils (0–15 cm depth) that were collected from six locations in Manitoba, Canada. The locations were Carman (Soil 1; 49◦29 6"N, 98◦02 2"W), Carberry (Soil 2; 49◦53 7"N, 99◦22 29"W), Deerwood (Soil 3; 49◦22 1"N, 98◦23 34"W), High Bluff (Soil 4; 50◦01 2"N, 98◦08 9"W), Beausejour (Soil 5; 50◦05 13"N, 96◦29 58"W), and Portage la prairie (Soil 6; 49◦57 9"N, 98◦16 0"W). These were the same six soils used in two previous studies [14,15]. In the Canadian soil classification system, all soils are classified as Chernozems (an equivalent of Chernozem in the FAO classification system) except Soil 4, which is classified as a Regosol (an equivalent of Regosol in the FAO classification system) [18]. The soils were air-dried and ground to pass through a 2-mm sieve. A subsample of each soil was analyzed (Table 1) for organic matter by the wet oxidation method [19], cation exchange capacity by ammonium acetate method [20], urease activity [21], soil texture by pipette method [22], field capacity [23], and pH (soil/water, 1:2) and electrical conductivity with a combined conductivity and pH meter (Orion versaStar, ThermoFisher Scientific Inc., Waltham, MA, USA).

## *2.2. Experimental Design and Treatment Applications*

The experimental setup was a completely randomized design containing two inhibitor treatments of six soils, a factorial layout for eight sampling periods, and was replicated three times for a total of 288 experimental units. The inhibitor treatments were NBPT (10 mg NBPT kg−<sup>1</sup> soil) and NBPT plus NI (DI; 10 mg NBPT + 2.5 mg NI kg−<sup>1</sup> soil). We used analytical grades of NBPT (CAS: 94317-64-3) and NI (3,4-dimethyl pyrazole phosphate; CAS: 202842-98-6) in this study.

Ten grams of each soil was weighed in 50 mL centrifuge tubes. The soil was wetted to 75% field capacity based on soil mass, capped, and left to equilibrate for 24 h at room temperature. Twenty-four hours after wetting, the soils in the centrifuge tubes were spiked with 0.5 mL of a solution containing either 200 mg NBPT L−<sup>1</sup> (NBPT inhibitor treatment) or 200 mg NBPT + 50 mg NI L−<sup>1</sup> (DI inhibitor treatment). The ratio of NBPT to NI in the DI inhibitor treatment was the same as the ratio of NBPT to NI in the double inhibitor formulation used in our previous studies [14,15]. However, the current study did not include urea with the inhibitors, as we discovered that the presence of urea interfered with the analytical procedure for NBPT. The tubes were recapped and placed in an incubator (Isotope Incubator, Model 304, Fisher Scientific, Hampton, NH, USA) set at 21 ◦C. On days

0, 0.5, 1, 2, 4, 7, 10, and 14 after treatment application, three replicates or samples of each soil by inhibitor treatment (i.e., six soils × two inhibitor treatments × three replicates for a total of 36 samples) were removed (destructive sampling) from the incubator for NBPT extraction and analysis. Day 0 was immediately after the soil was spiked with the inhibitor treatments.


**Table 1.** Selected soil (0–15 cm) properties.

<sup>a</sup> Canadian soil classification system.

#### *2.3. Extraction and Analysis of NBPT*

On each sampling day, 25 mL of deionized water was dispensed on the sampled centrifuge tubes and shaken on a reciprocating shaker for 30 min at 120 excursions per minute. After 30 min of shaking, the samples were centrifuged for 5 min at 10,000× *g* to allow soil residues to settle to the bottom. Immediately after centrifugation, about 4 mL aliquot was transferred using a 0.2 μm syringe filter (Basix™ Syringe Filters, ThermoFisher Scientific, Waltham, MA, USA) into a 20 mL vial. This was followed by transferring 1 mL of the filtered aliquot into a 2 mL high-performance liquid chromatography (HPLC) vial (9 mm surestop screw vial, ThermoFisher Scientific, Waltham, MA, USA) containing 0.1 mL dimethyl sulfoxide for NBPT analysis with HPLC-mass spectrometry (HPLC-MS), as described by Engel et al. [17].

The HPLC-MS (Bruker Compact QqTOF, Billerica, MA, USA) used was equipped with an electrospray source that operated in the positive ionization mode. The nebulizer pressure of the source was 0.3 bar with 5 L min−<sup>1</sup> of N2 drying gas at 200 ◦C. The capillary voltage was 3500 V, and the capillary exit voltage was 70 V. Reverse-phase chromatography was used to separate NBPT using an Intensity Solo C18 (100 × 2.1 mm, 2 μm) HPLC column (Bruker Daltonik, Billerica, MA, USA). The column was maintained at 35 ◦C with a flow rate of 300 μL min<sup>−</sup>1. The mobile phase consisted of formic acid 0.1% in Milli-Q water for Channel "A" and acetonitrile for Channel "B". A 2 μL aliquot of the sample was injected into the column and kept at 80% B from 0 to 3 min. From 3 to 4 min, the gradient was linearly ramped to 20% B, where it was kept for 1.5 min. Then, the gradient was linearly ramped to 80%, and it was held for 2.5 min at 80% for re-equilibration. The NBPT was eluted at approximately 3.3 min.

Data quantitation was performed using Bruker Daltonic QuantAnalysis (ver. 4.4) software (Billerica, MA, USA). The ion chromatograms for NBPT were defined as [M + H]+ (168.0719 *m*/*z*). The concentrations of NBPT recovered in soil were determined from a calibration curve of known standard solutions of NBPT and their corresponding peak

areas. The quantity of NBPT recovered was expressed as a percentage of NBPT applied to the soils.

### *2.4. Kinetics and Statistical Analysis*

Model fitting and statistical analysis were performed with SAS software (SAS Institute 2014, ver. 9.4 [24]). PROC NLIN was used to fit an exponential decay function (Equation (1)) to determine the degradation rate constant (k) of NBPT in the soils as follows:

$$\mathbf{Y} = \mathbf{b}\_o[\exp(-\mathbf{k}\mathbf{t})] \tag{1}$$

where Y is the % of NBPT recovered in soils at time t, t is the time in days, k is the NBPT degradation rate constant, and bo is an empirical constant.

For ease of interpretation, the generated k was used to calculate the half-life (t1/2) of NBPT in the inhibitor treatments using Equation (2):

$$\mathbf{t}\_{1/2} = \ln(2)/\mathbf{k} \tag{2}$$

We used PROC GLIMMIX (beta distribution) for repeated measure analysis to determine the significant effect of time, inhibitor treatments, and their interaction on the % of NBPT recovered in each soil. Furthermore, analysis of variance with PROC GLIMMIX (gamma distribution) was performed on the degradation rate constant and half-life of the NBPT across soils and inhibitor treatments. The fixed effects in the model were soil and inhibitor treatment. Mean comparisons were deemed significant at a probability level of 0.05 Fishers' protected least-significant difference. The goodness of fit for the exponential decay model was tested using the Nash–Sutcliffe model efficiency (ME) and root means square error [25]. Stepwise regression with PROC REG was used to analyze the influence of soil properties on the half-life of NBPT in soils.

$$ME = 1 - \frac{\sum\_{i=1}^{n} \left(\mathbf{Y}\_i^{\rm m} - \mathbf{Y}\_i^{\rm p}\right)^2}{\sum\_{i=1}^{n} \left(\mathbf{Y}\_i^{\rm m} - \overline{\mathbf{Y}}\right)^2} \tag{3}$$

where *Y<sup>m</sup> <sup>i</sup>* is the measured NBPT recovered in soil, *<sup>Y</sup><sup>p</sup> <sup>i</sup>* is the predicted NBPT recovered in soil, and *Y* is the mean of measured NBPT recovered in soil. When *ME* = 1, there is a perfect relationship measured and predicted NBPT recovery in soil; and when *ME* = 0, the model has the same precision as the mean of measured NBPT recovered.
