Impact of Annual Ryegrass on Nitrate-N Losses during One Growing Season of Maize in the Midwestern United States —An On-Farm Case Study

Winter cover crops have been shown to reduce nitrate-N (NO 3 -N) losses in runoff water and are recommended by the Illinois Nutrient Loss Reduction Strategy (NLRS) for reducing nutrient losses from agricultural fields. With an estimated 80 percent of the NO 3 -N load in Illinois coming from agriculture, the NLRS stresses the importance of farmers’ voluntary implementation of best management strategies in order to reach these goals. This study com-pares the difference in NO 3 -N losses from tile drainage water from an annual ryegrass (AR, Lolium multiflorum) winter cover-cropped treatment to a conventional tillage (CT) control (fall chisel and spring field cultivation). Throughout the maize (Zea mays L.) growing season, tile drainage water was collected and analyzed for NO 3 -N concentrations. Despite the AR treatment having a 29% lower mean daily NO 3 -N concentration, there was no significant difference in total daily NO 3 -N flux between AR and CT for this study period of April-July 2015. The cumulative losses of NO 3 -N were calculated at 11.65 and 10.56 kg ha −1 NO 3 -N for the CT and AR treatment, respectively, or a 9.4% reduction in the AR treatment during the period of study. When the season was divided based on growing season periods, the NO 3 -N flux values were less for the cover crop while the AR was actively growing, greater for the cover crop for the period following annual ryegrass termination through maximum crop canopy, and lower for the cover crop in the late stages of vegetative growth through relative maturity.


Introduction
Cover cropping utilizes the fallow period of an annual cash crop to grow a secondary crop, which can scavenge residual fertilizers [1] [2] [3], reduce erosion [4], and improve the overall health of the soil [5]. Benefits of cover crops include reduction in nutrient losses from erosion, leaching, or denitrification. Cover crops increase the infiltration of water, while the added surface residue can reduce evaporation losses, resulting in more water available to the subsequent cash crop [5] [6]. The losses of NO 3 -N through leaching, as N 2 or N 2 O through denitrification, NH 3 through volatilization, or as organic and ammonium nitrogen through erosion are not only economic losses in the field, but they also result in downstream consequences such as eutrophication or hypoxia [7], increase water treatment costs [8], and become heat trapping greenhouse gases [9].
Year to year NO 3 -N flux depends on the nitrogen use efficiency and the amount of rainfall.
The highest fluxes will happen when a wet year follows a dry year with little nitrogen uptake and when nitrogen management practices remain unchanged. In order to prevent NO 3 -N from leaving agricultural fields, cover crops can be implemented to utilize the NO 3 -N remaining in the soil after harvest of the cash crop, compete with winter annual weeds, and can be used as an alternative to fall herbicides [10].
Increases in modern maize production in order to feed an ever-growing population depend on maximizing nutrient use efficiency with soil testing, improving drainage, and utilizing better hybrids. Currently, fertilizers are not always effectively managed and natural soil fertility is too commonly unaccounted. This combination leads to nutrient losses by overland flow or drainage, both natural and man-made, and their losses are becoming more of a concern [11]. The improvement in managing fertilizer and soil fertility, specifically nitrogen and phosphorus, in row crop agriculture is necessary to meet nutrient loss reduction strategies proposed in Illinois [11]. In addition, fertilizer used in an unnecessary, excessive, or untimely manner will result in either loss into the environment and/or loss in profitability [12] [13] [14]. Given the dynamic interaction among weather, agronomic practices, and soil composition, the success of reduction losses will be challenging. However, a better understanding of how these systems interact is necessary if farmers are to reduce the negative impact on the environment while being able to maintain crop production at a profitable level. We cannot afford (financially or environmentally) to simply apply "more than enough' N" [15]. Service (USDA-NASS) census [16], there were approximately 19.65 million hectares (48.56 million acres) of drained land in the United States. Of this, 16.143 million hectares (38.89 million acres) are grain and oilseed production. Illinois has 3.41 million hectares (8.43 million acres) of subsurface-drained grain and oilseed land [16]. Detailed reviews presented by King et al. [17] [18] [19] summarize the necessity of tile drainage through historical expansion of arable land, increases in efficiencies by extending the time available to complete field work in the busy spring and fall seasons, maximizing productivity in the lands by stimulating mineralization, limiting denitrification and increasing nutrient availability to growing crops, minimizing crop stress from anoxic root conditions, and reducing the extent of surface erosion by allowing infiltration. However, along with its many benefits, there are inherent environmental risks associated with tile drainage. Most notably, subsurface tiles provide a shortcut from agricultural fields to nearby streams for infiltrated water. These alterations to natural soil processes pose risks to the environment due to the elevated concentrations of nutrients and pesticides in the drained water.
The Illinois Nutrient Loss Reduction Strategy (NLRS) [11] is a public-private partnership developed to address Illinois' nutrient contribution to the Gulf of Mexico, highlighting best management practices that reduce nutrient losses in Illinois, which will in turn help alleviate eutrophication in the Gulf of Mexico.
The Illinois NLRS has set Phase 1 reduction milestones for NO 3 -N and total-P of 15 percent and 25 percent, respectively, to the Mississippi River by the year 2025 with the ultimate goal of a 45 percent reduction for both nutrients when compared to average annual riverine loading for years 1980-1996. With estimated levels of 80 percent and 48 percent of nitrate-N and total-P nutrient loads coming from agriculture, respectively, the reduction strategy stresses the importance of farmers' voluntary implementation of best management strategies in order to reach these goals [11]. Among the specific management practices that can be used to reach these goals are the use of cover cropping, reduction in soil tillage, delayed timing and/or decreased rates of fertilizers, or bioreactors. Cover crops can be grown in rotation with cash crops. They are seeded during the growing season or following a cash crop and can be used for many processes in nutrient cycling. Cover crops can uptake residual nutrients from previous crops and also nutrients [26] or they can leave the ground too cool and wet to properly establish a cash crop if they are not terminated and rain continues to fall. Both scenarios may result in yield losses [27] [28] [29]. Cover crops are also an added cost to cropping practices and may not return added profits.
The objective of this study was to determine the impact of an annual ryegrass (AR) winter cover crop and tillage on NO 3 -N losses in tile drainage water. This study monitored the differences in daily cumulative drainage flow, water sample NO 3 -N concentrations, and daily NO 3 -N fluxes from subsurface tiles underlying the newly converted CT treatment and the continuous no-till AR winter cover crop treatment. Due to limited access and resources for this on-farm study, water samples were only collected during the maize growing season.

Materials and Methods
The research site and data collection were made possible through the conversion of a private field under normal agricultural production. Access to this field site was limited to one growing season of cover crops and maize, due to budget limitations and availability of the land and personnel. In addition, as the conventional tillage occurred in late fall when the soil temperature was quite cold (i.e. during snow cover), it is unlikely that nitrogen mineralization in the soil would have had any substantial impact on nitrogen transport before spring water quality monitoring began.
The site was located in Crawford County, IL, which met the requirements of 1) four consecutive years of no-till or minimum tillage, in which macropore drainage channels had not been compromised, 2) subsurface drainage in parallel pattern, 3) relatively level ground to maximize uniformity, 4) productive soils with drought tolerance and flood protection, and 5) previous crop harvested prior to October to allow for the proper cover crop fall establishment period.
Soil, geological and climate information were found in the Soil Survey of Crawford County, Illinois [30].

Tillage and Cover Crop Establishment
All site preparation was completed with commercial farm equipment and the field study was treated as if it was a field under normal production. The site was planted with an AR cover crop with a single coulter, no-till, air seeder drill at a depth of 1.3 cm (0.5 inch) and a rate of 15.1 kg ha −1 (13.5 lb ac −1 ) on September 27, 2014. No-tillage practices were carried out on the AR treatments. A period of warm and dry weather following planting allowed the AR to quickly germinate and emerge. Planting direction was 0 degrees, North, and perpendicular with tile lines. The AR was terminated at elongation stage with herbicide on April 17,

2015.
Primary tillage was completed on the CT plots November 18, 2014 with a chisel plow fitted with straight points. The working depth was in the range of 28 -

Cover Crop Biomass Collection
Aboveground biomass samples were collected from each of three AR plots on

Spring Fertilizer Application
Calcium carbonate agricultural limestone was surface broadcast at a rate of 4.5  [31]. Side dress applications were made with a 23-row single blade opener at depths of 75 mm (3 in).

Maize Establishment
Maize was planted with a double-disk no-till planter at a target population of 83,980 plants per hectare (34,000 plants per acre) with a 116-day hybrid (Dekalb ® 66-40) on May 6, 2015. Planting population was recorded with planter instrumentation and a final stand count was determined at relative maturity.

Flow Monitoring
The tile water flow was continuously monitored using in-line v-notch weirs and submerged pressure transducers for flow ( Figure 2). Teledyne ISCO ® (Lincoln, Nebraska, USA) 6712 automated water samplers were used for water collection.
The upstream flow monitoring station on each line was used to measure upstream flow and grab samples were intermittently collected for NO 3 -N analysis.  Each flow monitoring weir system was tested prior to field installation. Following testing for each unique weir, it was determined one flow curve would best represent all v-notch weirs (Equation (1)). Care was taken in installing the flow monitoring stations in the field with the same standards as they were tested in the laboratory. The flow of water in L s −1 for each v-notch weir, for the purposes of the field study, was determined by the following calculation: 3.007e * 2.175e * 2.8233 * *0.06309 where, Q ALL = Flow from v notch weir (liters/second); H = water depth above the nappe of the v (mm).
Onset ® (Bourne, MA, USA) HOBO ® U20L-04 pressure transducers were used to determine the water depth (H) in the tile lines. These pressure transducers measure absolute pressure with an accuracy of 0.1% full scale or 4 mm, but in order to do this, they must have another pressure transducer reading for barometric am each day, giving approximately 150 mL "sips" at 00:00, 04:00, 08:00, 12:00, 16:00, and 20:00 in each bottle. Autosamplers were programmed to take sips that were composited versus one large daily sample in order to collect water at different stages of flow.

Flow Analysis
The flow of water in the subsurface tiles was analyzed for the entire season and then segregated into three periods: prior to cover crop termination (April 12-April

Maize Biomass Samples
On September 20, 2015, at approximate relative maturity for maize, harvestable stand count was recorded and select above ground portions of plants were collected and weighed. The stand count was done by measuring a length of 5.31 m (17 ft 5 in) and counting plants with a harvestable ear on rows 6, 5, and 4 of each treatment. Following a stand count, the 6 th , 16 th , and 26 th plants were cut off at ground level and collected for analysis. All samples had a fresh weight recorded and they were then stored at 4 ˚C until they could be chopped and dried. Plant samples were run through a tree chipper prior to drying. Samples were dried using a forced air oven at 60˚C for 7 days. They were then weighed again and chopped into 5 mm pieces using a knife mill in preparation for shipping and analysis. They were sent to Agvise Laboratories (Northwood, ND) for analysis for total P, total N, and total C.

Yield Measurement
The focus on yield analysis was to use the mass flow sensor and internal software in a John Deere TM combine to determine yield. This method creates a much larger yield dataset than would otherwise be possible using a weigh wagon for each plot or hand harvesting. Harvest was completed on September 28, 2015 at a constant speed of 4.7 km hr −1 (2.9 mph) despite treatment harvest appearance or yield monitor readings. Each 24-row treatment was harvested by two, 12-row passes (9.14 m). Yield monitors in the combine collect data at a 1-second interval, or every 1.3 meters, which creates a harvest area of 11.85 m 2 (127.6 ft 2 ) for each point collected.

Statistical Analysis
Differences between the AR and CT treatments were analyzed for significance using an F Test Two Samples for Variances to determine if the variances were equal, followed by either a Two Sample t Test for unequal variances or Two Sample t Test for equal variance, depending on the result of the F test (Microsoft Excel).

Results and Discussion
The results from this study are presented for the entire study period first, followed by closer looks at strategic periods within the study period.

Study Weather Conditions
This study spanned from September 27, 2014, when the AR cover crop was planted, through September 28, 2015, when the maize was harvested. Since this study spanned only one cover crop and maize growing season, the observed temperature and precipitation data are presented in Table 1, along with historical data for comparison. The weather conditions, in terms of temperature and precipitation, were well suited for high maize yields during this time. Following termination of the winter cover crop in mid-April, there was a period of warm and dry weather, which allowed planting and an even germination and emergence of maize plants. The growing season for the maize crop from emergence to maturity brought above normal rainfall patterns during the months of March, April, and June and slightly below average rainfall during the months of July and August. The rainfall events were frequent which minimized plant water stress.

Full Season Analysis
Water flowed through tile lines intermittently due to rain events from the date of installation on April 12, 2015 through the July 21, 2015. During this period, approximately 402,517 liters and 320,393 liters drained through the AR and CT treatments, respectively.
Subsurface drainage continued for a large portion of the growing season, with the final period of flow happening on July 21. This was largely a combination of adequate rainfall and the size of the events, which provided excess soil water.
Most daily rainfall events that triggered large flows were greater than 3.8 cm (1.5 in). During much of the growing season, flow ceased in the CT plot prior to the AR plot. It was not until mid-June that the flow patterns were more uniform and equal between the two treatments. Figure 3

Living Cover Crop
The NO 3 -N concentrations, flow, and flux measurements were collected and analyzed to look specifically at the period of cover crop growth, April 12-April 18, 2015 (Table 3). This period marks the portion of the growing season in Table 3  The NO 3 -N concentrations, flow, and daily flux values were numerically less for the AR treatment than the CT treatment during the period when the winter cover crop was actively growing. However, there was no statistically significant difference between the AR and CT treatments for the April 12-18 period. The variability in daily flow is likely the reason for the higher p-values (>0.05).
Therefore, it is still valuable to discuss the numerical differences. The AR treatment influenced a 16.9% reduction in daily NO 3 -N concentrations from tile water samples, a 42.3% reduction in daily flow, and 30.5% reduction in daily NO 3 -N flux. An actively growing grass crop may have reduced the NO 3 -N flux in two ways: 1) uptake of available nitrogen from the soil profile thereby reducing NO 3 -N leaching losses, and 2) the uptake of soil water and a greater net transpiration resulting in less tile water flow. In addition, there was likely an increase in soil mineralization in the CT plots. Even with a late tillage event and a cooler than average spring, subsurface NO 3 -N values were still elevated in the CT plot.
It is known that high levels of nutrients can be lost during winter months when the soil is not frozen and tile drainage runoff is occurring [32] [33]. Therefore, this first sampling period provides some insight into the difference in treatments prior to maize planting.

Early Vegetative Maize Growth Following Cover Crop Termination
Daily NO 3 -N concentrations flow, and flux measurements were collected and analyzed for the period which immediately followed cover crop termination up to point of assumed evapotranspiration equilibrium (full crop canopy of both treatments). This period was April 19-June 18, 2015 (Table 4).
There were no statistically significant differences (p > 0.05) between the AR

Late Vegetative and Reproductive Growth
The late season vegetative and reproductive growth period was June 19-September 30, 2015.
The daily flow measurements for the AR and CT treatments for late vegetative period through the end of the season are given in Table 5. The AR no longer reduced evaporation from the soil surface and the depth of the corn roots likely equalized the deeper infiltration of the two treatments. The result is the treatments concentrations fallen back to more equal levels for the two treatments, but the CT treatment had a lower concentration than the AR treatment and marked one of only three days throughout the study in which a lower nitrate concentration was obtained from the CT treatment. The other two occurrences were on the second to last day of water collection, and a rainfall measuring greater than 2.54 cm followed starter fertilization and cover crop termination.

Cover Crop Biomass Analysis
The cover crops in this study were used as a resource to scavenge available soil nitrogen and, in turn, reduce leaching and denitrification losses. Table 6 shows certain characteristics of the growing cover crop prior to termination and the expected plant available nitrogen (PAN) to be released to the maize crop.

Maize Biomass Analysis and Yield
The analysis performed on the aboveground portion of maize shows the potential variability of a hybrid when the growing conditions are varied by tillage and soil nitrogen fertility (Table 7). For the CT treatment, plant uptake of nitrogen exceeded the fertilizer application rate of nitrogen on each plot. These results are consistent with the elevated leaching losses observed in the drainage water analysis, which is likely due to increased mineralization in the conventionally tilled soil. The biomass-N value for the sample from Plot 12 (CT) of 425 kg ha −1 is much greater than the applied amount. This higher uptake is likely due to greater N supply and uptake from soil mineralization. The lower leaves in the other test plots had more visible senescence whereas there was very little observed in the high nitrogen rate conventional plot, likely a response of remobilizing nutrients to support grain fill.
CT treatments also resulted in a greater plant biomass when compared to AR treatments. The highest yield of 16,842 kg ha −1 accumulated 425 kg ha −1 of nitrogen. However, a yield of 16,653 kg ha −1 was observed with a nitrogen value of 219.7 kg ha −1 and is a reminder of the variability in nutrient uptake and ability of modern hybrids to repartition limited nutrients and still achieve high yield levels.

Conclusion
Excess rainfall created tile flow through mid-July. The cumulative values of NO 3 -N flux and average NO 3 -N concentrations were greater for the CT treatment. When Journal of Water Resource and Protection comparing the total tile drainage flow, the AR treatment had a greater cumulative flow than the CT treatment. When the season was divided based on growing season periods, the NO 3 -N flux values were less for the cover crop while the AR was actively growing, greater for the cover crop for the period following annual ryegrass termination through maximum crop canopy, and lower for the cover crop in the late stages of vegetative growth through relative maturity. NO 3 -N concentrations were greater for the CT treatment than the AR treatment until the final two dates of water analysis. While the AR cumulative NO 3 -N flux was lower than that from the CT treatment, maize yields were also lower from the AR treatment. This may signify that nitrogen was either tied up in forms not available to the maize in the AR treatment, or perhaps the tillage in the CT treatment allowed for a higher rate of mineralization of nitrogen in the soil, which was available to the maize during the growing season. Additional research is needed to confirm the findings in this study and better understand some of the effects that require longer term evaluation.