Abstract

Manure and inorganic fertilizer help to meet crop nitrogen demand by supplementing soil nitrogen (N). However, excessive N losses reduce soil fertility and crop yield and can impair water and air quality. The objectives of the research were to compare different forms of fall-applied N for (1) the change in soil nitrate (NO3-N) over the growing season and (2) the aerial ammonia (NH3) and nitrous oxide (N2O) fluxes during the fall and early growing season. Treatments included solid beef cattle manure with bedding (BM), solid beef cattle manure only (SM), urea (UO), and no fertilizer (NF). The two-year plot-scale study took place in Brookings County, South Dakota, under rain-fed conditions in a silty clay loam. Manure and urea were applied at equal plant-available N rates of 130 and 184 kg·N·ha−1 in Y1 and Y2, respectively, according to the South Dakota nutrient management planning process. The average total (i.e., 0–0.60 m soil depth) soil NO3-N for Y1 (83 kg·ha−1) was significantly higher than Y2 (67 kg·ha−1), whereas surface (i.e., 0–0.15 m soil depth) soil NO3-N was not significantly different between years. The average surface soil NO3-N (33.5 kg·ha−1) and total soil NO3-N (105.0 kg·ha−1) for UO were significantly higher than the remaining treatments (). Soil water NO3-N concentrations, leaf-N, corn-grain-N, and yield measurements did not indicate any significant differences between treatments. Based on the two-year average, the highest NH3-N flux occurred from the BM (3.4 g·ha−1·h−1); however, this flux was only significantly higher than NF (1.4 g·ha−1·h−1). The NH3-N fluxes from UO (2.2 g·ha−1·h−1) and SM (1.7 g·ha−1·h−1) were similar to both BM and NF. The N2O-N flux from UO (0.79 g·ha−1·h−1) was significantly greater than NF (0.25 g·ha−1·h−1), while BM- (0.49 g·ha−1·h−1) and SM-produced (0.33 g·ha−1·h−1) N2O-N fluxes were not significantly different than neither UO nor NF. The three fall-applied N sources had similar aerial-N fluxes even though urea application resulted in significantly higher soil nitrate.

1. Introduction

Soil nitrogen (N) is critical for crop yield [1, 2]. As an N source, manure can also increase soil organic matter and improve soil health [3, 4], which translates to higher productivity. However, an excess amount of soil N loss can lessen nutrients for crop production and pollute surrounding air and water bodies. Over application or mismanagement of manure and N fertilizer sources can promote N losses from the soil [5, 6] through volatilization, denitrification, leaching, runoff, and erosion [710]. Nitrous oxide (N2O) loss to the atmosphere contributes to global warming as well as depletion of the ozone layer [1114]. EPA [13] indicated that agricultural activities including manure/fertilizer application and cropping practices were responsible for about 79% of total United States (US) N2O emissions in 2014. European studies suggest 10 to 50% of total agricultural ammonia (NH3) loss is from land-applied manure fertilizers [15, 16]. Ammonia has a wide variety of environmental impacts including soil acidification, acid rainfall, and eutrophication of ecosystems [17]. Also, aerial-NH3 can react with atmospheric gases such as sulfur dioxide or nitrogen oxides (in the presence of water) to form fine particulate matter that is very harmful for human and animal health and the environment [18, 19]. In order to reduce aerial and other N losses, we must first understand the conditions that promote N loss to the environment.

Eastern South Dakota is part of the Northern Great Plains and the Upper Missouri River Basin, with a semiarid climate [20]. In 2012, 22% of the US beef cows and 20% of corn production occurred in the Northern Great Plains [20], with many producers integrating livestock manure and crop production management decisions.

Previous studies have demonstrated the effect of land application of solid manure regarding nutrient availability [21], greenhouse gas emission [2224], NH3 emission [22, 23], and leachate concentrations [25]. Several studies have also compared the effects of slurry or liquid manure application of different manure types (swine, dairy, poultry, feedlot, etc.) for similar factors [8, 9, 2634].

Bedding is used on animal farms for animal comfort, to reduce animal injury and to aid in manure handling [35, 36]. Corn or soybean stover, wheat straw, or corn cobs are common bedding materials for beef cattle in the Midwest and Northern Great Plains because they are locally available [37]. Bedding with manure may impact soil properties, N and phosphorus (P) uptake, and N mineralization rates [4, 9, 28, 38, 39]. Carbon-rich bedding can temporarily immobilize manure N in the soil, delaying the release of plant-usable forms of N. However, as soil microbes use carbon as an energy source, they also help N mineralization [40]. Miller et al. [28] found that soil inorganic N, soil P, and soil mineralizable N were significantly affected by manure applications and the effects changed with year, bedding, rate of application, and their interactions.

In the Northern Great Plains, fall application of manure is a common practice—this allows emptying of manure storages prior to the winter period and avoids soil compaction associated with manure application during the often wet spring season. A study by Loecke et al. [8] reported that N use efficiency on corn was higher for fall-applied manure than spring-applied manure. Also, manure has sufficient time to decompose while it is applied in the fall and make nutrients more available for crop uptake in the spring [41]. Best management practices recommend fall application of fertilizers once the soil temperature is less than 10°C to limit N losses [42].

The objectives of the research were to compare different forms of fall-applied N (solid beef cattle manure with bedding, solid beef cattle manure only, and urea) and no fertilizer for (1) the change in soil nitrate (NO3-N) over the growing season and (2) the aerial ammonia (NH3) and nitrous oxide (N2O) losses during the fall and early growing season. This study was conducted near Brookings, South Dakota, following a nutrient management plan guide for the area. This study does not account for all N losses from the crop system, but these data add to the understanding of factors affecting aerial nitrogen losses. In conjunction with other research, this type of data helps in the future refinement of management and nitrogen loss factors used in manure nutrient management planning tools. Similarly, concurrent research occurred in North Dakota [23, 43] and Nebraska under different climatic and soil conditions.

2. Materials and Methods

2.1. Site Description

The research was conducted at the South Dakota Felt Farm in Brookings County (44°22′07.5″N and 96°47′35.7″W, and 516 m above mean sea level) between October 2015 and October 2017. The research site area was 0.11 ha (45.7 × 24.7 m) with average slope <1%. The soil was silty clay loam, classified as Udic Haploborolls [44]. Daily temperature and precipitation records for the research site were collected from the South Dakota State University Climate and Weather Station located about 7 km from the site.

2.2. Experimental Design

The experimental design was a randomized complete block design (RCBD). The four treatments included solid beef cattle manure with bedding (BM), solid beef cattle manure without bedding (SM), urea (UO), and no fertilizer/control (NF). There were four blocks with four plots (experimental units; 3.3 m × 9.1 m) per block. A plot was assigned in each block to each treatment (Figure 1). The project periods from nitrogen application in November of 2015 and 2016 to harvest in October of 2016 and 2017 were designated as Year 1 (Y1) and Year 2 (Y2), respectively.

Based on soil and manure tests, N-based application rates were determined using the South Dakota Fertilizer Recommendations Guide EC-750 [45] (Table 1) for a corn yield goal of 11.3 Mg·ha−1 (180 bu·ac−1). There was no credit prescribed for the soybean crop preceding Y1. Aggregated soil samples (0–0.6 m) from the research site area showed an average soil nitrate nitrogen concentration of 113 kg ha−1 prior to Y1 application, and the postharvest soil nitrate nitrogen levels in Y1 (Table 2) were used for Y2 application rates (methodology described in Soil Parameters). Manures and urea were manually applied to each plot in November each project year. The entire research site was tilled within 24 h of nitrogen application with two passes of a disk plow. There was no irrigation. In Y1, the corn plant date was May 2. In Y2, corn was first planted on May 6 and replanted on June 2 following insufficient seed emergence.

2.3. Sample Collection and Analysis
2.3.1. Manure Parameters

The SM and BM were from beef cattle manure stockpiles at the South Dakota State University Ruminant Nutrition Center, Brookings, SD. The BM included corn stover bedding. Prior to application, ten random shovel samples of manure were collected from stockpiles of each type in a pail and mixed thoroughly and a subsample used for manure characterization following recommended practices by Peters et al. [46]. Total N analysis was by dry combustion Dumas method, ammonium-N (NH4-N) analysis was by distillation, and phosphate (P2O5), potassium oxide (K2O), sulfur (S) were determined using microwave-assisted digestion and inductively coupled plasma spectroscopy.

2.3.2. Soil Parameters

Soil samples were collected prior to manure application and prior to planting, six-leaf vegetative stage (V6), and postharvest stage from each plot. On each sampling day, composite soil samples were collected for 0–0.15 m (0–6 in.) and 0.15–0.60 m (6–24 in.) depths in each plot using a probe auger. Shallow soil sample (0–0.15 m) analyses included ammonium nitrogen (NH4-N), electrical conductivity (EC), organic matter (OM), phosphorus (P) concentration, total N, total C, and pH. The NO3-N for the 0.15–0.60 m deep sample in each plot was added to the shallow sample measurement for total soil NO3-N. The pH, EC, and OM were analyzed using North Central Extension Research Activities guidelines [47]. Soil NO3-N and NH4-N were analyzed using the flow injection analysis QuikChem method 12-107-04-1-B with 2 M KCl extraction [48]. Similarly, Olsen-P was determined by the sodium bicarbonate method [49], whereas ammonium and total N and C in the soil samples were analyzed using the QuikChem method 12-107-06-2-F [50] and Dumas method [51], respectively.

A suction lysimeter (1270 mm in length and 22 mm diameter; Irrometer Company, Inc., CA, USA) was installed at 1200 mm soil depth in the north end of each plot (Figure 1). Soil water samples were collected on 17, 23, 31, 35, 44, and 50 d after planting in Y1 and 16, 24, 33, 39, 48, and 53 d after planting in Y2. The number of soil water sampling days depended on rainfall events and soil water availability. During sample collection, a hand pump applied a vacuum pressure between −60 and −70 kPa and the vacuum was maintained for 4 hours. Soil water collected in the lysimeters was extracted using a 60 mL polypropylene syringe, collected into a 50 mL polypropylene vial, and transferred to the laboratory for analyses. In the laboratory, NO3-N concentration in collected water was determined using an Automated Timberline TL2800 ammonia analyzer (Timberline Instruments, Boulder, CO).

The Aridlands Ecology Lab Protocol (modified 2009.01.19, S. Castle) was used to measure bulk density of the top 50 mm of soil. Soil surface (0–50 mm) samples were collected from each plot using the AMS bulk density soil sampling mini kit (50 mm dia. × 50 mm stainless steel ring). A composite soil sample was used for mechanical texture analysis.

2.3.3. Crop Parameters

The six most recently unfurled leaves below the whorl at the six-leaf vegetative stage (V6) and six leaves below the corn ear at tasseling (VT) and physiological maturity stage (R6) were collected and composited from six plants in each plot. We collected yield and corn-grain nitrogen concentration measurements during the Y1 harvest.

After drying and grinding, total N concentration of the plant and grain samples were measured using the micro-Kjeldahl procedure [52].

2.3.4. Ammonia Flux

The NH3 gas flux was collected at three locations in each plot using semistatic chambers with acidified (0.5 M H3PO4) foam strips as described by Jantalia et al. [53]. In Y1, flux was measured −4, 3, 7, and 13 d from the day of fall N-application, and −6, 10, and 30 d from the day of planting. In Y2, flux was measured −7, 1, 6, and 15 d from the day of fall N-application, and −35, 7, and 42 d from the day of replanting. Collection periods ranged from 1 to 2 d before fall N-application, 2 to 8 days after fall N-application, and 3 to 8 days in the spring. After each collection period, we transferred the acidified foam strips to a freezer. We extracted the acid trap solution from thawed sample traps with 250 mL of 2 M KCL solution. A subsample of the extraction solution was analyzed for ammonia using automated methods (Model TL2800 ammonia Analyzer, Timberline Instruments, Boulder, CO).

Ammonia concentration was obtained in g·N·ha−1 by multiplying NH3 concentration (μg·mL−1) and the total volume of solution (250 mL) and then dividing by the surface area of the soil covered by the respective chamber (79 cm2). The ammonia flux (g·N·ha−1·h−1) for each plot was determined by dividing ammonia concentration by elapsed time from installation to the removal of the NH3 traps.

2.3.5. Nitrous Oxide Flux

In this study, the static chamber method was used described by Parkin and Venterea [54] to measure N2O flux. Each chamber consisted of a polyvinyl chloride (PVC) collar (254 mm internal dia. × 150 mm) and a vented PVC cap. Each collar was installed 10 cm into the soil. The average headspace height was 142 mm with the cap in place.

Three collars on each plot (n = 48) were installed. In Y1, flux was measured −4, 7, and 13 d following fall N-application and −7, 1, 6, and 15 d following N-application in Y2. Flux measurements were collected −7 d in Y1 and −38 and −17 d in Year 2 relative to corn planting and subsequently at monthly intervals in both years. For each flux measurement, 10 mL gas samples were withdrawn from each chamber using a syringe after 0, 30, and 60 min of cap placement. We transferred samples to 12 mL preevacuated glass vials. A sample of ambient air during the sampling time for each block was collected to capture the ambient N2O concentration, and all samples were collected between 930 and 1600 h. The N2O concentrations in the vials were measured using gas chromatography (Model 14B, Shimadzu Corporation, Japan). Air and soil temperature (Model 00641W, AcuRite, Lake Geneva, WI) and soil moisture (Model ML2x, Delta-T Devices, Cambridge, England) were monitored in the vicinity of each chamber on each sampling day.

The average ambient concentration measurement was used in place of the 0 min samples for each sampling day. The N2O fluxes were determined from N2O concentrations relative to elapsed time. Flux calculations were not performed if (a) the 30 min (T30) and/or 60 min (T60) concentration(s) were less than (1 − error)  ambient concentration and (b) the quadratic curve through the 3 data points was concave down and T60  (1 + error) was less than T30  (1 − error) or (c) the quadratic curve through the 3 data points was concave up and a linear slope fit through the 3 points was not significantly different than zero. If the quadratic curve through the 3 points was concave down, the slope at time zero (the first-order coefficient of the quadratic equation fits through the 3 data points) was used to calculate the flux. If the quadratic curve through the 3 points was concave up, but the linear slope through the 3 points was significantly different than zero, the linear slope was used to calculate the flux. The allowable error (proportional to concentration) was 20%. Evaluated N2O fluxes were then converted into µg N2O-N m−2·h−1 using the ideal gas law equation. The resulted fluxes were corrected using soil properties (bulk density, clay fraction, pH, moisture content, and soil temperature) using the method of Venterea [55].

2.4. Statistical Analysis

PROC GLIMMIX procedure [56] was used for mixed model analyses with repeated measures for all variables except corn yield and corn-grain nutrients. Treatment (BM, SM, UO, and NF) and growth stage (stages differed for the various dependent variables) were considered fixed factors for surface and total soil NO3-N and leaf-N concentration data, but time (sampling day) was considered a random factor for soil water NO3-N concentration, NH3 flux, and N2O flux data. Year was considered a fixed effect and block considered as a random replication factor for all variables. The normality of the residuals was reviewed using Q-Q plots, and if residuals were not normal, different distribution options (e.g., lognormal, exponential, and Poisson) available in PROC GLIMMIX were tested. Different covariance structures were used to assess the repeated measure data, including covariance component (VC), compound symmetry (CS), autoregression (AR (1)), unstructured (UN), and Toeplitz (TOEP). The covariance structure selected for each variable was based on the smallest Akaike information criterion (AIC) value. Table 3 summarizes the fixed and random effects, distribution type, and covariance structure used for each variable. Significant differences were considered at .

The obtained least squared means (LSMeans) from the lognormal distributions were back transformed for reporting purposes. For post hoc tests, Tukey’s honest significant difference (HSD) was used. We used correlation analyses to investigate relationships between yield, leaf-N, and soil nitrate data.

3. Results and Discussion

Herein, data for weather, soil, and crop data that relate to aerial N2O and NH3 fluxes for the experimental period were presented. Additional supporting data collected during the experiment are in Table 4, Mehata [57], and Mehata et al. [58] for future reference or modeling purposes.

3.1. Weather Conditions

Daily mean air temperatures were 13°C in Y1 and 7°C in Y2 on the day of manure/fertilizer application (Figure 2). In Y1, there was no rainfall for 12 d after N-application; however, light rainfall (1 mm) occurred 2 d after N-application in Y2. Annual precipitation in 2016 and 2017 was 5.3% higher and 3.9% lower, respectively, compared to average annual precipitation between 1981 and 2010 (618 mm). Wet and cool weather in May of 2017 contributed to poor emergence and the need for replanting in Y2. The 530 mm of precipitation received during Y1 during the corn growth period (May through October 2016) was 17% higher than the 442 mm received during the Y2 growing season (June to October 2017).

3.2. Soil Nitrogen

All fertilized plots received equal amounts of plant-available nitrogen, which includes ammonium-N and mineralized organic N from the manure (Table 1). Treatment significantly affected the total soil NO3-N (0 to 0.60 m) and surface soil NO3-N (0 to 0.15 m) during the project (Table 2). The mean surface soil NO3-N (33.5 kg·ha−1) and mean total soil NO3-N (105.3 kg·ha−1) for UO were significantly higher than manure and no-fertilizer plots (, Table 2). The average surface soil NO3-N between Y1 (23.1 kg·ha−1) and Y2 (21.8 kg·ha−1) was not significantly different, but total soil NO3-N was significantly greater in Y1 (83.4 kg·ha−1) compared to Y2 (67.5 kg·ha−1). It is possible more soil NO3-N moved from the lower depths by leaching in Y2 because rainfall in Y2 (preplanting to V6 stage) was 226 mm, compared to 172 mm for Y1.

For the surface soil nitrate, the interaction between treatment and growth stage was significant. Carbon and organic matter can reduce N mineralization in manure plots, influencing soil nitrate. Qian and Schoenau [21] found that N mineralization decreases significantly with an increase in the C/N ratio in cattle manure.

The interaction between year and growth stage was a significant effect for surface and total soil NO3-N (, Figure 3). Nominally, but not significantly, the Y2 preplant surface and total nitrate-N were 25 and 18% greater than Y1, respectively. In this study, the recommended soybean credit [45] of 44.8 kg·ha−1 was neglected in the Y1 nitrogen application rate calculations (Table 1). Nitrogen fixation in the soil likely increased the Y1 preplant soil nitrate levels. The Y2 preplant nitrate-N concentration for the NF, BM, and SM treatments increased approximately 125 kg·N·ha−1 from the postharvest levels; the UO treatment showed a similar increase in addition to the urea-N applied in the fall. Mineralization of organic N in the research site area over winter and spring could account for some of the additional nitrate-N. Mineralization of N from soil organic matter and residual nitrogen immobilized from different fertilizer sources after snow melt may have contributed to the increased nitrate level preplant. The wet spring conditions and soil moisture levels may have also promoted upward movement of nitrate from lower soil profiles. Despite the higher Y2 preplant levels for all treatments though, nitrate-N levels decreased faster between preplant and V6. Surface nitrate decreased between preplant and V6 each year and increased between V6 and postharvest. The decrease in Y2 between preplant and V6 was more significant than Y1. The significant effect of growth stage between Y1 and Y2 on soil NO3-N may be related to the replanting of the crop in Y2 and differences in precipitation during the growing seasons between Y1 and Y2. Wetter conditions during early spring in Y2 may have promoted nitrate loss via erosion and leaching prior to the growing season measurements.

Other soil parameters such as average total N concentration (0.2%), ammonium concentration (22.68 kg·ha−1), and total carbon (2.2%) were not significantly different among treatments for Y2 (Y1 data not available) (Table 4).

3.3. Soil Water Nitrate Concentration

Soil water samples from each plot between corn planting and the V6 stage in both Y1 and Y2 growing seasons were indicators of leachate concentration under the root zone. Treatment was not a significant factor (; Table 2). The average soil water NO3-N concentration was significantly greater in Y1 compared to Y2. This difference may be due in part to soil nitrate and rainfall differences. The total soil NO3-N was higher at preplanting time but lower at the V6 stage in Y2 compared to Y1 at the same stages (Figure 3). The 148 mm of rainfall in Y1 between planting and V6 stage was greater than the 108 mm of rainfall in Y2 during the same period of corn growth. Allaire-Leung et al. [59] found a positive correlation between nitrate leaching and soil NO3-N. Nitrate leaching from soil also depends on soil type, N-application rate, types of N sources, cover crops cropping intensity, and crop N uptake [6062], and these factors influence translation of these research results to other fields and crop systems.

3.4. Crop Response

The average leaf-N concentration significantly differed based on year and growth stage (V6, VT, and R6; ; Figure 4). The leaf-N concentration at the V6 stage was higher in both years of study and the concentration of N in leaves decreased with corn growth stage. The N uptake from soil peaks between the vegetative and tasseling stages [63, 64]. The mean leaf-N concentration at the V6 stage was not significantly different between Y1 and Y2, whereas at VT and R6 stages, average leaf-N was significantly greater in Y1 compared to Y2 (Figure 4). The variation in leaf-N concentration over the growing stages may relate to differences in soil NO3-N over the corn growth period. The average leaf-N was 23.9 g·kg−1 for the Y1, whereas for Y2, it was 18.8 g·kg−1. The significant variation in leaf-N concentration among two years may be due to rainfall, soil moisture, and soil-available N for late corn planting in the Y2. Leaf-N has also been linked to rainfall during the growing season, and low rainfall can affect nutrient availability [65].

Yield and grain-N concentrations in Y1 did not differ based on treatment, including NF, despite differences in soil N. The BM and UO corn yields greater than 12.0 Mg·ha−1 exceeded the yield goal of 11.3 Mg·ha−1, whereas the SM and NF plots were 96% of the yield goal. Voss et al. [66] and Kovács and Vyn [67] found that ear-leaf-N concentration and corn yield were significantly correlated. The Y1 data showed that yield and leaf-N concentration for each plot at the corn tasseling stage (N = 16) were significantly related (r = 0.70, and ). The yield (r = 0.68) and grain-N (r = 0.71) also correlated with the average Y1 total NO3-N present in the soil (N = 16). The correlation result indicated that yield and grain-N are dependent on available soil N. These correlations suggest lower yield and grain-N concentrations in Y2 compared to Y1 were likely based on lower soil nitrate and leaf-N levels in Y2.

Crop response differences between manure- and urea-N sources were not expected because of the common plant-available N-application rates. However, differences in response between fertilized and NF treatments were expected. Limited replications and lack of Y2 yield data do not allow us to validate the nitrogen application recommendations as they pertain to corn growth.

3.5. Aerial Nitrogen Fluxes
3.5.1. Ammonia Flux

Table 5 shows the average NH3 fluxes for sampling days after N-application for Y1 and Y2. The measurements showed NH3 fluxes increased on the first sampling day after N-application in the fall for Y1 and Y2 relative to preapplication, and then decreased over the remainder of the fall sampling days (Table 5). Over the project, the analysis showed a significant effect of treatment on NH3 flux () but no significant effect of year or treatment by year interaction. The average NH3 flux from BM (3.4 ± 0.9 g·ha−1·h−1) was only significantly higher than NF (1.4 ± 0.4 g·ha−1·h−1), whereas SM and UO were not significantly different than either BM or NF. In the companion study by Niraula et al. [43] in North Dakota, the average NH3 loss rate from N-applied plots (solid beef manure with and without bedding and urea) was not significantly different, similar with the current study’s result. The daily mean NH3 loss rates measured in North Dakota ranged from 1.8 to 32.0 g·N·ha−1·h−1. Adviento-Borbe et al. [31] reported peaks as high as 15 g·N·ha−1·h−1 during manure application that decreased within 24 h, with fluxes typically below 1.07 g·ha−1·h−1 from liquid dairy manure and fertilizer N-treated plots under corn-corn and corn-alfalfa rotation. They stated that decreasing NH3 flux might be due to decreased total ammoniacal nitrogen at the soil surface, infiltration of slurry into the soil profile, and a drop in pH due to NH3 volatilization. The pattern of NH3 fluxes in this study was similar to their observation after N-application. Application timing, methods, N sources, and bedding material also affect reported soil NH3 flux data compared to this study. Within our study, the higher NH3 fluxes from BM treatment could be due to higher ammonium-N (NH4+-N) (Table 1) in the manure with bedding compared to the manure only (SM). Huijsmans et al. [68] observed that soil NH3 flux increased with an increase TAN in manure.

Lab-scale studies suggest bedding material can influence ammonia volatilization from stored manure with bedding (i.e., bedded pack manure). Ayadi et al. [69] found NH3 volatilization increased when corn stover bedding is used compared to soybean stubble during warmer weather because of rapid urea and protein hydrolysis and higher moisture content in corn stover. The simulation study by Spiehs et al. [70] showed using corn stover and different woodchip bedding produced a higher NH3 compared to green and dry cedar chips; the difference was associated with pH levels of the bedding materials. This study did not detect different ammonia losses between the bedded and nonbedded manure once land-applied and using data collected over multiple seasons.

The present study suggests that when manure or urea is applied to silty clay loam soils in late fall at agronomic rates based on plant-available nitrogen and incorporated within 24 hours, ammonia loss can increase slightly compared to no nitrogen application, but the source of nitrogen is not significant.

3.5.2. Nitrous Oxide Flux

The theoretical flux underestimation method as described by Venterea [71] used an average soil bulk density of 1.3 g·cm−3, a clay fraction of 18%, the measured surface soil pH (Table 4), and the soil temperature and moisture content (Table 6).

Table 6 shows the average nitrous oxide fluxes for each sampling day during Y1 and Y2. Treatment significantly affected the overall combined N2O fluxes. Year and interaction of year with treatment did not show any significant effects on N2O fluxes. The average (±SE) N2O flux from UO was 0.78 (±0.25) g N2O-N·ha−1·h−1 and significantly higher than the flux from NF of 0.25 (±0.08) g N2O-N·ha−1·h−1. The average flux (±SE) from manure treatments BM and SM were 0.49 (±0.15) and 0.33 (±0.10) g N2O-N·ha−1·h−1, respectively. The N2O fluxes from manure-treated plots were not significantly different from UO and NF (Table 6).

For the two-year study, the average (±SE) N2O flux was 0.44 (±0.14) g N2O-N·ha−1·h−1 for the silty clay loam soil plots. Miller et al. [4] compared the long-term land application of stockpiled feedlot beef manure with bedding (barley straw and woodchips) on C/N ratio, denitrification, and carbon dioxide emission in southern Alberta, starting in 1998. They annually applied stockpiled feedlot manure with bedding at the rate of 77 Mg (dry weight) ha−1·yr−1 for 13 to 14 years to a clay loam soil. The measurement of denitrification fluxes were taken in 2011 and 2012 (every 2 weeks between May and August). They found mean N2O fluxes for manure with straw bedding were between 0.04 and 44.92 g N2O-N·ha−1·h−1 (0.9 and 1078 g N2O-N·ha−1·d−1), from 0.03 to 13.58 g N2O-N·ha−1·h−1 (0.8 to 326 g N2O-N·ha−1·d−1) for manure with woodchip bedding, and 0.03 and 10.42 g N2O-N·ha−1·h−1 (0.6 to 250 g N2O-N·ha−1·d−1) for control. However, they observed that total N, daily denitrification flux, and daily carbon dioxide flux were not affected by bedding materials. The maximum mean N2O flux (0.80 (±0.25) g N2O-N·ha−1·h−1 in this study was very low compared to the maximum fluxes reported by [4]; it might be because of lower rate of solid beef manure application (about half application rate), different bedding materials used (corn stover vs. barley straw and woodchips), gas sampling techniques/methods, and weather conditions. Akiyama and Tsuruta [72] measured N2O flux for poultry manure (PM), swine manure (SM), and urea applied to soil using an automated flux monitoring system. They found the total fluxes were 0.21, 0.07, and 0.05 g N2O-N·ha−1·h−1 (184, 61.3, and 44.8 Mg N2O-N·m−2·y−1) from PM, SM, and urea, respectively. This study showed higher N2O fluxes than the fluxes reported by Akiyama and Tsuruta [72]; this difference likely can be attributed to soil properties (silty clay loam vs. Andisol (volcanic ash soil)), types and rate of N sources, sampling method, and different climatic conditions. In contrast, the N2O flux obtained for broadcast-incorporated N placement by Engel et al. [73] was similar to N2O fluxes from urea-treated plots from this study. Engel et al. [73] studied the effect of urea placements (broadcast, band, and nest) on N2O emission from a silt loam soil. The rate of urea application was 200 kg·N·ha−1. They found maximum N2O fluxes for the broadcast surface, broadcast incorporated, band, nest, and control were 0.62, 0.55, 1.03, 1.17, and 0.13 g N2O-N·ha−1·h−1 (14.8, 13.2, 24.1, 28.1, and 3.1 g N2O-N·ha−1·d−1), respectively. In the Red River Valley, North Dakota, Niraula et al. [43] found the N2O fluxes varied from 0.022 (NF) to 1.91 (UO) g·N·ha−1·h−1 (2.2 to 191 μg·N·m−2 hr−1). They did not find any significant difference between N-applied plots, which was similar to the current study. However, they observed very high N2O flux from the urea-treated plot, which was more than 2-fold as compared to our study. This may be due to urea application timing (spring versus fall), soil characteristics, and local weather conditions.

The literature suggests N2O flux normally varies strongly with the degree of water-filled porosity [74, 75], but this relationship was not apparent in this study. This may be, in part, because the flux correction method [71] reduces the influence of soil moisture on estimated flux rates in treatment comparisons. Nominally, the N2O fluxes generally decreased between preplant and the growing season which may due to the active crop N uptake and N losses, similar to Niraula et al. [43].

Our study suggested that application of solid beef manure with or without corn stover bedding can reduce the potential N2O release from the soil, but other parameters such as methane emission or capture, carbon dioxide emission, and nitrate leaching also need consideration when investigating nitrogen fertilizer sources.

4. Conclusion

In this study, manure and urea were applied to plots at equal plant-available N rates using nutrient management planning guide recommended rates. Fall application of beef manure with and without bedding and no fertilizer produced significantly lower soil nitrate levels compared to urea application during a two-year plot-scale study with the silty clay loam soil type in the Northern Great Plains region. No treatment differences in soil water nitrate concentration or corn leaf-N were observed. However, there were differences between years and growth stages for soil nitrate and leaf-N. Numerically, soil N2O and NH3 fluxes from applied N sources were UO > SM > BM and BM > UO > SM, respectively. The form of N-applied did not significantly affect fluxes. The UO and BM treatments produced significantly higher N2O and NH3 fluxes compared to NF, respectively. Because gas flux rates can vary by soil type, weather, and management practices, these data add to the body of knowledge for future modeling efforts and manure management decisions. Further study is required to refine N-application recommendations appropriate for increasing crop yield while minimizing soil N losses based on weather and soil characteristics.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Additional Points

Highlights. (i) Fall application of manure (with and without bedding) or urea did not result in different ammonia or nitrous oxide fluxes for the silty clay loam soil. (ii) Fall-applied urea produced significantly higher soil nitrate concentrations; however, crop response variables were statistically similar for all N source. (iii) Nitrate concentration in soil water differed between years, but not among N sources.

Disclosure

Any opinions, finding, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the U.S. Department of Agriculture.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This work was supported by a grant from the United States Department of Agriculture’s National Institute of Food and Agriculture (2015-67020-23453). The authors appreciate the statistical advice provided by Dr. Kathleen Yeater and the field work support from Ms. Suraiya Akter, Mr. Bryce Siverling, Mr. Doug Doyle, and Mr. Corey Moret.