For tests conducted in 2013 (test 1), Teejet 8010 flat-fan nozzles were calibrated to compare three application volumes (252, 505, and 1038 L/ha⁻¹) at a pressure of 138 kPa (Table 1). There are no carrier volume recommendations on the label for azoxystrobin (Anonymous, 2013), however, these volumes represent a range that likely exceeds what a producer would be willing to use. The same nozzles were also calibrated to compare four different application pressures (138, 276, 414, and 552 kPa) at a similar carrier volume of approximately 281 L ha⁻¹ (actual range was from 252 to 299 L/ha, Table 1). A CO₂ pressurized application system with a two-row spray boom (nozzles were 101 cm apart centered on the beds) was placed on the quick hitch of a tractor at a height of 56 cm above the canopy, and driven at various speeds to establish the desired combination of application volume at a given pressure. The fungicide azoxystrobin (Abound FL, Syngenta Crop Protection LLC, Greenboro, NC) was banded (51 cm) over the rows (peanuts were planted in a single row/bed) at the same rate (0.45 kg a.i. ha⁻¹  =  0.225 kg a.i. banded ha⁻¹) for all treatments. Peanut foliage at the time of applications was continuous down the row and wider than the banded applications across the row.

In 2014, the application volumes of 131 and 524 L ha⁻¹ were compared at two pressure rates of 138 and 414 kPa in test 2 and carrier volumes of 131, 262, and 524 L ha⁻¹ were compared at 138 and 414 kPa in test 3. Teejet 8002VS flat-fan nozzles and Teejet 8015E flat-fan nozzles were used to calibrate the appropriate rates (Table 1). The same CO₂ pressured system was used and the same rate of azoxystrobin was applied in both years. The 2013 spray parameters were designed to explore the range of carrier volume and pressure that might be used in commercial applications, but did not permit testing for interactions between carrier volume and pressure. The 2014 application parameters still maintained a range of parameter values that showed differences in 2013, but allowed for carrier volume x pressure interactions to be tested.

In 2013, the test design consisted of six treatments (Table 1), that were applied at three different times during the season (24 July, 7 Aug, and 16 Aug, each plot was only treated one time during the season), in a randomized complete block design with four replications per application time. A spanish-type cultivar (Tamnut OL06) was planted in early May on single beds. The soil series was a Patricia fine sand with a texture of 80% sand, 6% silt, and 14% clay, with 0.3% organic matter and a pH of 8.2. Plots were four rows wide, where only the middle two rows were treated, and were 10.7 m in length. Soil moisture was monitored with a Watermark soil moisture (capacitance-type) sensor buried to a depth of 10 cm, and attached to a WatchDog 1200 series data logger (Spectrum Technologies, Inc., Aurora, IL). Volumetric soil moisture starting at the time of application was used to determine the approximate length of time between application and the first rain or irrigation event. All three application dates were followed within a few hours with an irrigation (24 July and 16 Aug) or rain event (7 Aug). This measurement method does not indicate the amount or intensity of water that impacts the plants. Hourly soil moisture can be seen in Fig. 1A. For the 24 July application, which was completed around 17:30 hr, there was an increase in soil moisture from 3.2 to 12.6% at 18:00 hr, and then a second irrigation event on 29 July at 5:00 hr when soil moisture changed from 4.3% to 12.6% in one hr. For the 7 Aug application, which was completed around 16:00 hr, there was an increase in soil moisture at 17:00 hr (from 8.6 to 20.5% due to rain), then on 14 Aug at 10:00 there was an increase from 5.2 to 8.6% in two hr. For the 16 Aug application which finished around 15:00 hr, there was an increase in soil moisture at 17:00 hr from 5.8 to 13.6% in one hr, and then on 21 Aug at 10:00 hr there was an increase of soil moisture from 5.1 to 13.2% in one hr.

In 2014, test 2 consisted of four treatments (two carrier volumes x two pressures; Table 1) applied on 24 July, and test 3 consisted of six treatments (three carrier volumes by two pressures, Table 1) applied on 28 July, each with four replications in a randomized complete block design. The two tests were located at different parts of the field. The field was planted with a virginia-type cultivar (Florida Fancy) in mid-May on a single bed. The soil series was an Amarillo fine sandy loam, and the texture was 68% sand, 10% silt, and 22% clay with 0.6% organic matter and a pH of 7.5. Plots were four rows wide with only the middle two rows treated and were 6.1 m in length.

Soil moisture was monitored for tests 2 and 3 similarly as in 2013. Average daily soil moisture can be seen in Fig. 1B. Test 2 was irrigated initially five d after application and test 3 was irrigated two d after application. Soil moisture in test 2 went from 7.1% at 0 hr to 22.7% moisture at 107 hr. Soil moisture in test 3 went from 15.1% at 0 hr to 24.6% at 40 hr after application. A second irrigation occurred with test 2 at 11 d after the original application and with test 3 at 8 d after the original application (Fig. 1B).

In 2013, plant and soil samples were taken when two irrigation or rain events had occurred after the fungicides were applied. Sampling was done 5, 8, and 6 d after the applications on 24 July, 7 Aug, and 16 Aug, respectively (Fig. 1A). An area in the middle of the plot consisting of 50 cm of row length was selected. Plants in this area were trimmed so that only the center 30 cm (width) remained, and then all the leaflets from top to bottom were removed, placed in a bag, mixed, and then placed in a cooler with ice, until they could be frozen at -20 C. The soil in this area to a depth of 10 cm was removed (4 cores in 50 cm of row), mixed, and a subsample of 20 g was placed in a plastic bag, and then placed in a cooler with ice until they could be frozen at -20 C. Samples were shipped overnight with ice packs to Omic USA Inc. (Portland, OR) to be analysed for azoxystrobin concentration using the European Standard EN 15662 (2008). In this procedure, the pesticide is initially extracted with a mixture of acetonitrile and water. The acetonitrile containing the pesticide is separated from water and cleaned using dispersive solid phase extraction. The cleaned extract is analysed using gas chromatography tandem mass spectrometer (GC-MS/MS) and/or ultra-performance liquid chromatography tandem mass spectrometer (UPLC-MS/MS). The pesticide detected in the sample is confirmed using multiple MSMS transitions. The quantification is done using a matrix matched calibration standard. The minimum detectable level for this assay was 0.02 ppm.

In 2014, sampling foliage was similar to 2013, but soil samples were taken at 0 to 8 cm, and 8-16 cm depths with a 2.5-cm diameter sampling probe (four cores/middle 50 cm of plot). In 2014, plots were sampled multiple times, with test 2 at 4, 8, and 15 d after fungicide application, representing sampling after 0, 1, and 2 irrigation events (no rainfall occurred, Figure 1B). Test 3 was sampled at 3 and 7 d after application which represented samples taken after 1 and 2 irrigation events. The expansion in sampling in 2014 both over time and at two depths was done in response to questions that were raised over the results in 2013, primarily where had the fungicide gone with the higher carrier volumes.

In 2013, the effect of application parameter combinations and date that the plots were sprayed were evaluated using analysis of variance with the PROC MIXED procedure of SAS (version 9.3, SAS Institute, Cary, NC). Significant treatment effects (P < 0.10) were separated using the least-squared mean function and pair-wise differences (PDIFF) at P < 0.10 in 2013. A square root transformation (√A) was used to normalize variance of azoxystrobin concentration on leaves, but was unnecessary with soil azoxystrobin concentrations.

Regression analysis (PROC REG, SAS) was conducted on subsets of the data for each application date, to examine the linear relationship between carrier volume and azoxystrobin concentration, and the linear and quadratic relationship between pressure and azoxystrobin concentration. A model was accepted if all parameters were significant at P < 0.10. The pattern of residuals was also examined to determine model suitability.

In 2014, the plots for each spray date were in two separate areas (not randomized within the same area as was done in 2013), so each spray date and sampling time was analysed separately for effect of carrier volume, pressure, and their interaction, with the PROC MIXED procedure. The dependent variables measured in 2014 were transformed azoxystrobin concentration in the foliage, soil from 0 to 8 cm depth, and soil from 8 to 16 cm depth. Significant interactions (P < 0.10) and main effects (P < 0.10) were separated using the least-squared mean function and pair-wise differences (PDIFF) at P < 0.10.

The individual treatment combinations were compared across each sampling date (sampling dates after 0, 1, and 2 irrigations for test 2 and sampling dates after 1 and 2 irrigations for test 3) with a t-test at P = 0.10. The purpose was to determine if the concentration of azoxystrobin changed with each subsequent irrigation event/time. Finally, the impact that initial irrigation timing had on azoxystrobin concentration was compared with a t-test (P ≤ 0.10) at each treatment combination (i.e. trt 1 was compared across test 2 and test 3 for differences after 1 and 2 irrigation events; trt 2 was compared across test 2 and 3 after 1 and 2 irrigation events, etc.).

There were a total of five trials conducted (three on spanish peanuts and two on virginia peanuts) that examined the impact of application parameters (carrier volume and pressure) on azoxystrobin concentration on foliage and in the soil. The focus of this work was on controllable application parameters, i.e. carrier volume and pressure. While irrigation/rainfall and plant size effects were not tested, they were measured and used in discussing results that may have differed between experiments.

The plant size (height [SD] x width [SD]) at the time of application was 45.1 [SD = 7.5] x 64.0 [SD = 5.8] cm, 49.4 [SD = 4.2] x 78.5 [SD = 6.9] cm, and 53.6 [SD = 8.1] x 73.6 [SD = 9.8] cm, for the first, second, and third application times, respectively. There was a significant interaction (P  =  0.034) between application parameters (six treatments) and date of application with respect to azoxystrobin concentration, therefore each application date will be presented separately. Carrier volume did not affect foliar √A for the 24 July application (Fig. 2A). For the 7 Aug application, √A in ppm was higher for the combination of 252 L ha⁻¹ + 138 kPa (√A  =  1.60) than for the carrier volumes of 505 L ha⁻¹ + 138 kPa (√A  =  0.69) and 1038 L ha⁻¹ + 138 kPa (√A  =  0.45). After the 16 Aug application, azoxystrobin concentration was highest (√A > 2.78) for those treatments with carrier volumes of < 300 L ha⁻¹ and/or < 414 kPa (Fig. 2A). The combination of 281 L ha⁻¹ and 414 kPa had a higher √A (2.25) than treatments with > 300 L ha⁻¹ and 138 kPa, which averaged √A between 0.97 and 1.09 ppm.

The √A on foliage could be adequately fitted with equations based only on carrier volume (Fig. 3), and there was no improvement in R² value by combining both carrier volume and pressure in the model. The models fitted were:

V was the carrier volume in L ha⁻¹. The slope value of the 16 Aug application was significantly different than that of the 24 July application (P < 0.10). Foliar √A decreased as carrier volume increased.

There was a significant interaction (P = 0.053) between application parameter combinations and date of application with respect to azoxystrobin concentration in the soil (Fig. 2B). There were no differences between treatments for the 24 July application. After the 7 Aug application, azoxystrobin concentration was lower (0.24 to 0.3 ppm) when carrier volume was > 300 L ha⁻¹ than for all treatments with a carrier volume < 300 L ha⁻¹ (concentration ranged from 0.54 to 0.82 ppm, Fig. 2B). When carrier volume was between 252 and 299 L ha⁻¹, but pressure differed, there was a higher concentration of azoxystrobin for 414 and 552 kPa (0.78 and 0.82 ppm, respectively) than with 138 kPa (0.54 ppm, Fig. 2B). Following the 16 Aug application, azoxystrobin concentration was higher (0.52 ppm) for the combination of 290 L ha⁻¹ and 276 kPa than for carrier volumes > 300 L ha⁻¹ (0.18 and 0.27 ppm for 505 and 1038 L/ha, respectively, Fig. 2B).

The concentration of azoxystrobin in the soil across all treatments averaged 0.17, 0.56, and 0.36 ppm for application dates of 24 July, 7 Aug, and 16 Aug, respectively, which are all significantly different (P < 0.001) from each other. Application date had more impact on the treatments where carrier volume was < 300 L ha⁻¹, than for the applications with 505 and 1038 L ha⁻¹ (Fig. 2B). The treatments with higher carrier volumes also had the lowest concentration of azoxystrobin overall (summed across foliage and soil). These results caused us to question where the fungicide actual went for the higher carrier volume treatments? Changes were made then in 2014 sampling to look at fungicide concentration over more time periods after application and to a greater depth in the soil.