Evaluating nitrogen losses, soil physical properties, and biological indicators in artificially drained fields in Northwest Minnesota A THESIS SUBMITTED TO THE FACULTY OF THE UNIVERSITY OF MINNESOTA BY Aaron L. Frankl IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE Dr. Lindsay Pease, Co-Advisor Dr. Anna Cates, Co-Advisor June 2022 © Aaron L Frankl 2022 Copyright i ACKNOWLEDGEMENTS I would like to thank my advisors, Dr. Lindsay Pease and Dr. Anna Cates. Throughout my time here, they have been incredibly supportive and dedicated to helping me succeed and grow as a scientist. I would also like to thank Dr. Fabián Fernández and Dr. Jeff Strock for serving on my committee and providing guidance throughout the process. Thank you to the staff and students in the Graduate Program in Land and Atmospheric Sciences. I would also like to thank Kyle Sherbine, who assisted with many aspects of this project. His friendship, guidance, and support made this work possible. Thank you to everyone else in the Cates-Pease lab group who helped shape and complete this project: Heidi Reitmeier, Hava Blair, Peyton Loss, Cameron Fleisher, and others. I would also like to acknowledge and thank the funding sources for this project: the Agricultural Fertilizer Research and Educational Council (AFREC) and the Minnesota Corn Growers Association. Finally, thank you to all my friends and family who supported me. Thank you to my parents and thank you to Joseph for guidance throughout the process. Thank you to my partner, Abby, who pushed me and opened my eyes to new ideas. ii ABSTRACT Subsurface drainage is a common practice to improve agricultural production in fields with poorly drained soils. This practice is becoming increasingly common in the Upper Midwest and NW Minnesota, where changing precipitation patterns are generating an increased need for artificial drainage systems. These systems increase yields, and over time, can reshape the soil ecosystem. By changing soil moisture throughout the profile, drainage can alter microbial activity and soil physical properties. In the first chapter of this study, we examined how drainage installation altered N cycling, which is largely mediated by soil microbes. In the two years following subsurface drainage installation, we observed several differences between the drained and undrained treatments: soil nitrate concentrations (NO3-), nitrous oxide (N2O) emissions, N mineralization, and edge of field losses. In the second chapter, we examined the longer-term changes facilitated by drainage to physical and soil health properties by sampling six sites. Three of these sites had their most recent drainage system installation before 2005; the remaining three had installations since 2015. The fields with older drainage systems had increased saturated hydraulic conductivity (Kfs) rates and increased biological activity (potentially mineralizable carbon, water-extractable organic carbon, and water-extractable organic nitrogen). Together, both studies identify the importance of understanding how drainage systems alter the soil ecosystem and demonstrate the importance of understanding drainage as an on-farm management decision. Additionally, they point to a need for more long-term drainage studies which focus on a wider range of soil properties. iii TABLE OF CONTENTS Acknowledgements ...................................................................................................... i Abstract ....................................................................................................................... ii List of Tables ................................................................................................................ v List of Figures .............................................................................................................. vi List of Abbreviations ................................................................................................. viii CHAPTER 1: Nitrogen Cycling in Recently Drained fields ............................................... 1 1.1. Synopsis ....................................................................................................................... 1 1.2. Introduction ................................................................................................................. 1 Soil N Cycle .................................................................................................................... 3 1.3. Materials and Methods ................................................................................................ 6 Site Description/Experimental Design .......................................................................... 6 Climate .......................................................................................................................... 7 Soil Sample Collection ................................................................................................... 7 Water Sample Collection .............................................................................................. 8 Discharge and Nutrient Loads ....................................................................................... 9 Water-Extractable Organic Nitrogen .......................................................................... 10 Greenhouse Gas Measurements ................................................................................ 10 Mineralization ............................................................................................................. 12 Nitrogen Uptake ......................................................................................................... 12 Statistical Analysis ..................................................................................................... 12 1.4. Results ......................................................................................................................... 13 Edge of field Losses ..................................................................................................... 13 Nitrous Oxide Emissions ............................................................................................. 16 Soil Nitrogen Transformations .................................................................................... 17 Grain Yield and Uptake ............................................................................................... 18 Loss Comparisons ....................................................................................................... 19 1.5. Discussion ................................................................................................................... 19 Management Implications .......................................................................................... 26 CHAPTER 2: The Long-term Impact of Subsurface Drainage on Soil Physical Properties and Biological Indicators ............................................................................................ 43 2.1. Synopsis ...................................................................................................................... 43 2.2. Introduction ................................................................................................................ 44 Soil Physical Properties ............................................................................................... 45 Soil Biological Indicators ............................................................................................. 46 2.3. Materials and Methods ............................................................................................... 48 Site Selection and Sampling ........................................................................................ 48 Infiltrometers .............................................................................................................. 49 Aggregate Stability ...................................................................................................... 50 Bulk Density ................................................................................................................ 50 Textural Analysis ......................................................................................................... 51 iv Potentially Mineralizable Carbon ............................................................................... 51 Water-Extractable Organic Carbon/Nitrogen ............................................................. 52 Statistical Analysis ....................................................................................................... 52 2.4. Results ......................................................................................................................... 53 Climate and Sampling Conditions ............................................................................... 53 Saturated Hydraulic Conductivity ............................................................................... 54 Aggregate Stability ...................................................................................................... 54 Bulk Density ................................................................................................................ 54 Potentially Mineralizable Carbon ............................................................................... 54 Water-Extractable Organic Carbon/Nitrogen ............................................................. 55 2.5. Discussion ................................................................................................................... 55 Soil Physical Properties ............................................................................................... 56 Biological Indicators .................................................................................................... 59 Management Implications .......................................................................................... 61 2.6. Conclusions ................................................................................................................. 61 References ................................................................................................................. 69 v LIST OF TABLES 1.1 Information about crops, N application, yield, and key dates at experimental site ..... 28 1.2: Average daily temperature and average precipitation of the 10 years prior to study and each year of the study ........................................................................................................ 28 1.3: Flow-weighted concentrations of NO3-N and TN in surface and subsurface drains. Upper case letters signify significance (p < 0.05) between time periods (e.g., Early 2020 Surface TN vs Late 2020 Surface TN), lower case letters for surface/subsurface comparisons (e.g., Surface Early 2020 TN vs Subsurface Early 2020 TN) ...................... 29 1.4: Nutrient loads for undrained and drained plots. Statistical significance (p < 0.05) is indicated only between undrained and drained plots for each N species within the same period. Organic N is non NH4-N organic species .............................................................. 29 1.5: Seasonal, yearly, and total cumulative loads from different sources over the period of study. Statistical significance (p < 0.05) indicates average daily load during time period, not comparison of cumulative load. Upper case letters signify significance between time periods (e.g., Early 2020 Surface TN Load vs Late 2020 Surface TN Load), lower case letters for surface/subsurface comparisons (e.g., Surface Early 2020 TN Load vs Subsurface Early 2020 TN Load) ...................................................................................... 30 1.6: Net ammonification, nitrification, and mineralization from the drained and undrained plots ................................................................................................................................... 30 1.7: Grain yield for 2019 to 2021 by drainage treatment. Wheat was planted in 2019 and 2021 and soybean in 2020. Polk County average yields are sourced from USDA NASS ........................................................................................................................................... 31 1.8: Total nitrogen uptake (TNUP) by drainage treatment in 2021 .................................. 31 1.9: N losses separated by year and treatment. Crop removal in 2020 was calculated based on a conversation factor in Fernández et al. (2009). Please note that leaching below subsurface drains was not conducted in this study, and that leaching in the subsurface in both drained and undrained treatment occurred ................................................................ 31 2.1: Site descriptions for sampling locations. All sampling locations used conventional tillage and management practices unless noted ................................................................. 63 2.2: Precipitation and temperature for the past 30 years as well as in 2021 at each sampling location .............................................................................................................................. 63 2.3: Summaries of soil physical properties and soil biological indicators. A star (*) indicates a significant difference (p < 0.1) ........................................................................ 64 vi LIST OF FIGURES 1.1: Lake Winnipeg Watershed and location of the field site, in Crookston, MN. Water flows north in the Red River to Lake Winnipeg in Manitoba, Canada. Adapted from Wassenaar and Rao (2012) ................................................................................................ 32 1.2: The soil N cycle on the drained portions of the research plots in Crookston, MN. There are also two lift stations that collect water at the edge of field (“Runoff” and “Leaching” mechanisms), which act as controlled drainage ............................................. 33 1.3: Layout of experimental site. Plots 1 and 3 are drained, while 2 and 4 are undrained. Tile drains run south to north in each plot, collected at the tile lift station (northern dot). Surface drains are not shown but run east to west and are collected at the surface lift station (southern dot) ..................................................................................................................... 34 1.4: Surface discharge, subsurface discharge, and precipitation from the experimental site (Midwest Regional Climate Center). In 2020, 1.5% of precipitation was discharged at the edge of field, where only 0.14% was in 2021 ................................................................... 35 1.5: Discharge from surface and subsurface drains in 2020 and 2021. A) Discharge in mm in 2020 and B) in 2021. Note that the y-axis varies between the two years ...................... 35 1.6: Monthly concentrations in both surface and subsurface drains in 2020 and 2021. A) NO3-N concentrations and B) TN concentrations. Samples collected from lift station water storage structures ............................................................................................................... 36 1.7: Drained and undrained daily NO3-N (open circles) and TN (closed circles) loads by year. To include all points, some 0.0 kg ha-1 loads appear below 0.0 ............................... 37 1.8: Relative percentages of loads during 2020 and 2021. Note that May 2021 is missing due to machine error. NH4-N was grouped with organic N due to no NH4-N data in 2020 ........................................................................................................................................... 37 1.9: TN and NO3-N loads from the experimental site in 2020 and 2021. A) NO3-N and B) TN loads from both surface and subsurface drains ........................................................... 38 1.10: N2O fluxes and cumulative emissions across the growing season in 2020 and 2021. A) Flux in 2020, B) Cumulative in 2020, C) Flux in 2021, D) Cumulative in 2021 ........ 38 1.11: Volumetric water content in the top 6 cm of soil for drained and undrained treatment. in 2020 (top) and 2021 (bottom) ....................................................................................... 39 vii 1.12: Cumulative A) ammonification, B) nitrification, and C) mineralization during the 2021 growing season averaged across drainage variables ................................................. 40 1.13: WEON separated by depth across the time of study. A) WEON from 0 - 30 cm, B) from 30 - 60 cm, and C) from 60 - 90 cm. There are no significant differences between treatments at all depths and seasons .................................................................................. 41 1.14: Soil NO3- by depth. Samples collected in spring, prior to planting and represent the impact of treatment from the previous growing season and winter. A star (*) indicates statistically significant results (p < 0.05) ........................................................................... 42 1.15: Grain yield comparisons by year in response to drainage treatment. Yield in 2019 was prior to drainage treatment, 2020 after one year of drainage, and 2021 after two years of drainage ......................................................................................................................... 42 2.1: Site locations in the Red River of the North Basin (RRB). Approximate location of the RRB in Minnesota, North Dakota, and South Dakota highlighted ............................. 65 2.2: Box plots of saturated hydraulic conductivity (Kfs) values of the three treatments. The >15 treatment has significantly higher Kfs than the <5 treatment ..................................... 66 2.3: Percent water-stable aggregates by size class for the different treatments ................ 66 2.4: Bulk density of three treatments, split by depth. There is no significant difference between drainage treatments at either depth ...................................................................... 67 2.5: A) Potentially mineralizable carbon (PMC), B) Water-extractable organic carbon (WEOC), and water-extractable organic carbon (WEON) separated by depth. Starred (*) depths indicate significant difference between >15 years and <5 years (p < 0.1) ............. 67 2.6: A) Potentially mineralizable carbon (PMC), B) water-extractable organic carbon (WEOC), and water-extractable organic carbon (WEON) separated by crop. Starred (*) depths indicate significant difference (p < 0.1). The undrained treatment has been removed from visual analysis to not skew data ................................................................. 68 viii LIST OF ABBREVIATIONS Kfs Saturated Hydraulic Conductivity N Nitrogen N2O Nitrous Oxide (gaseous) NH4-N Ammonium-Nitrogen NH4+ Ammonium Ion NO3- Nitrate Ion NO3-N Nitrate-Nitrogen O2 Oxygen (gaseous) P Phosphorous PMC Potentially Mineralizable Carbon RRB Red River of the North Basin TN Total Nitrogen WEOC Water-Extractable Organic Carbon WEON Water-Extractable Organic Nitrogen 1 CHAPTER 1: NITROGEN CYCLING IN RECENTLY DRAINED FIELDS 1.1. Synopsis Nitrogen (N) is a limiting nutrient for many crops and is supplied by fertilization. However, N losses from agricultural fields pose environmental concerns and economic challenges. At the same time, artificial subsurface drainage is an increasingly common practice to support agricultural productivity in the Upper Midwest. Drainage can also promote nutrient losses and alter N cycling, as it alters the functions of microbial communities in soils. In this study, we quantify changes in N cycling following subsurface drainage installation on a 24-hectare field in NW Minnesota. Soil, water, and gas samples were collected over a two-year period in both drained and undrained treatments to evaluate N losses and transformations. Following periods of high precipitation, soil nitrate (NO3-) decreased and nitrate loads increased in drained sections. In addition, the undrained treatment experienced greater nitrous oxide (N2O) losses, a potential pollution swap of drainage intensification. Other measured N metrics such as water-extractable organic nitrogen (WEON) showed no difference between treatments, and mineralization rates varied rarely between the treatments. During the first years following subsurface drainage installation, N losses mainly occurred through subsurface and surface drains. These losses via drains are variable and largely controlled by precipitation and soil moisture. 1.2. Introduction The importance of nitrogen (N) in Upper Midwest cropping systems is well known (Clark et al., 2019; Fernández et al., 2017; Rubin et al., 2016). In corn (Zea mays 2 L.) and wheat (Triticum spp. L) production, N is commonly considered the limiting nutrient in crop growth (Rubin et al., 2016). However, N losses from agricultural fields to both surface and groundwater is a serious environmental concern and an economic challenge to growers. Eutrophication, or the nutrient-enrichment of water bodies, occurs when nutrients such as N and phosphorus (P) enter waterways in high concentration, impairing marine and freshwater ecosystems. These nutrients facilitate algal growth and accumulation, producing algal blooms, which can damage ecosystems through oxygen depletion and blocking of sunlight for marine life. Eutrophication and its associated algal blooms can alter food webs and threaten aquatic ecosystems as well as contaminate the water humans uses for recreation and drinking (Wurtsbaugh et al., 2019). With a rising population that necessitates agricultural intensification and increased N fertilizer use, eutrophication is likely to intensify in the coming centuries without the implementation of nutrient load reduction strategies (Rabalais et al., 2009). The rapid eutrophication of Lake Winnipeg in Manitoba, Canada is associated with N and P loads delivered by the Red River flowing north from NW Minnesota (Figure 1.1). N and P loads leaching into the Red River from agricultural fields in NW Minnesota will likely increase with forecasted growth in the region’s corn production (Benoy et al., 2016; Bourne et al., 2002; Cormier et al., 2021; Rattan et al., 2017; Schindler et al., 2012; Yates et al., 2012). While P is the primary limiting nutrient for algal blooms in freshwater and an emerging issue in the RRB, N loading also selectively promotes toxic cyanobacteria blooms present in Lake Winnipeg (Gobler et al., 2016). 3 The management of N and P originating on agricultural fields in the Red River of the North Basin (RRB) is essential for reducing these algal blooms in Lake Winnipeg. Across the Midwest, artificial subsurface drainage is a common practice to support agricultural production in poorly drained soils. The negative impacts of poorly drained soils on crops in the Midwest is well documented (Evans and Fausey, 1999; Kaur et al., 2020; Mukhtar et al., 1990; Skaggs et al., 2012), as is the increased grain yield of crops in fields with subsurface drainage (e.g., Kladivko et al., 2005; Paiao et al., 2021). The adoption of subsurface drainage has generally been slower in northwest Minnesota than other regions of the Midwest due to economic limitations. However, subsurface drainage is becoming more common in the RRB, a region in which unplanted and unharvested hectares represent approximately 25% of total crop land due to poorly drained soils (Jin et al., 2008; Wiersma et al., 2010). Although the benefits of subsurface drainage on grain yield in the RRB remains unproven (Mehring et al., 2015; Mourtzinis et al., 2021; Wiersma et al., 2010), the installation of subsurface drainage is intensifying primarily due to the introduction of corn production into the region. At the same time, the RRB is increasingly experiencing more irregular and higher-intensity rainfall events due to climate change (Krichels et al., 2019; Pease et al., 2017). Both climatic change and the increase of subsurface drainage necessitate smarter water and nutrient management in crop production to mitigate degradation of water resources (Jiang et al., 2020; Morton et al., 2015). Soil N Cycle The soil N cycle (Figure 1.2) is largely influenced by water and its movement in the soil profile. This occurs two ways. First, the oxygen level in the soil determines 4 whether denitrification, nitrification, and mineralization are possible (Barakat et al., 2016, Zhu et al., 2018). Second, flow of soil water moves dissolved N species both through and from the system (dissolved N, leaching). Subsurface drainage is installed to aerate the soil and increase the rate of water movement through the soil, thereby altering the soil N cycle. The altering of N cycling by subsurface drainage is dependent on soil properties, as shown by variable results across soil textures (Borchers and Perry, 1991; Palmer et al., 2017; Sugihara et al., 2010; van der Werff et al., 1995). In addition to being lost rapidly through both subsurface drains and leaching, inorganic forms of N can often be reduced or oxidized by changes to soil water content. Waterlogged soils produce anoxic conditions. Thus, without O2 available, denitrifying bacteria in soil use NO3- as an electron acceptor, converting NO3- to gaseous forms of N, such as NO, N2O, and N2, which are then lost to the atmosphere (Reddy et al., 1984; Zumft et al., 1997). Drainage can intervene in this process by decreasing moisture and increasing oxygen in the soil profile, thus reducing nitrification (Colbourn and Dowdell, 1984; Gambrell et al., 1975; Gramlich et al., 2018; Grossel et al., 2016; Jungkunst et al., 2006). Leaching and crop uptake are two other important N cycle pathways affected by soil drainage. Nitrate leaching below the root zone can occur due to NO3- and soil’s repulsive negative charges, which results in weaker NO3- sorption to soil particles and its loss in soluble form. Ammonium leaching is less extensive in agricultural systems because it adsorbs strongly onto soil particles (Gramlich et al., 2018). Ultimately, crop uptake of inorganic nitrogen forms (NO3- and NH4+) is closely tied to mineralization. 5 The decomposition of organic forms of N to inorganic forms (NO3- and NH4+) through the process of mineralization is also altered by water content and O2. Mineralization is also connected to N application rates, crop residue quality, organic matter, and temperature (Paiao et al., 2021). This array of factors makes mineralization rates highly variable and difficult to predict. Studies focusing on mineralization rate and drainage have generally produced mixed results, suggesting that mineralization rates are variable with soil type or even management (Fernández et al., 2017; Mcisaac and Hu, 2004; Rice et al., 1987). In addition, accurately quantifying mineralization rates has proven to be challenging in agricultural fields in non-controlled environments (Eickenscheidt et al., 2014; Fernández et al., 2016). Quantifying changes to the N cycle are difficult due to the complexities of environmental systems. While fundamental changes caused by drainage to moisture levels and their theoretical changes to the N cycle are well understood, the directions, magnitudes, and interactions of pathways require more research due to the complex natures of these systems (Blann et al., 2009; Gramlich et al., 2018). As is particularly relevant for the current study, little work has examined N cycling directly following the installation of drainage. Additionally, the clay-rich soils and water quality concerns in the RRB make it a unique and critical environment as corn production moves northward (Ju Cho and McCarl, 2017; Motha and Baier, 2005). This research will provide baselines for quantifying N cycling in the initial growing seasons following drainage installation, where changes in N management may be needed to account for increased losses through the various pathways described above. In drained fields, we expected to see increases in edge of field losses, but increased N 6 crop uptake and reduced N2O emissions. By quantifying these pathways of loss, we can more accurately evaluate how N moves through the soil profile, providing better information to maximize yields and reduce the environmental impacts of subsurface drainage. 1.3. Materials and Methods Site Description/Experimental Design The study took place at the Northwest Research and Outreach Center in Crookston, MN (Figure 1.3) located in the Red River of the North Basin. The river flows north from North Dakota and Minnesota, discharging into Lake Winnipeg in Manitoba, Canada. The soils of the Red River of the North Basin are formed above clayey lacustrine sediments from glacial Lake Agassiz (Jin et al., 2008). The two major soil series in the study site are the Bearden-Colvin silty clay loams (Bearden: Fine-silty, mixed, superactive, frigid Aeric Calciaquolls; Colvin: Fine-silty, mixed, superactive, frigid Typic Calciaquolls) and the Bearden-Fargo silty clay loams (Bearden: Fine-silty, mixed, superactive, frigid Aeric Calciaquolls; Fargo: Fine, smectitic, frigid Typic Epiaquerts) Typical agricultural crops in the region include soybeans (Glycine max L.), sugar beets (Beta vulgaris L.), wheat, barley (Hordeum vulgare L.), potatoes, (Solanum tuberosum L.) sunflowers (Helianthus annus L.), and corn (Table 1.1). Historically, the field under study has been in annual soybean-wheat rotation, with occasional years of sugar beets, corn, and barley. The field is generally planted in late April to early May, dependent on spring moisture levels, and harvested in late summer (wheat) or fall (soybeans, corn, etc.). 7 For the study, a 24-hectare field was divided into four 6-hectare plots, designated as plots 1-4 from west to east (Figure 1.3). In October 2019, subsurface drainage (referred to as tile drainage) was installed in plots 1 and 3. In the drained treatment, subsurface drainage lines are spaced 15-meters apart and lie at an approximate depth of 1 meter from the surface with a 0.1% slope from south to north. Another set of drains collect surface water, which flows across the field east to west. Active use of the drainage system began on May 24, 2020. Agronomic information about the site is summarized in Table 1.1. Climate The mean annual temperature in Crookston is 4.4 °C, with an average temperature of -12.9°C in the winter and 20.1°C in the Summer (2012-2021, Midwest Regional Climate Center). The mean annual precipitation is 424 mm (Table 1.2; 1990-2019, Midwest Regional Climate Center). Soils typically freeze by November-December and generally thaw in March. Soil Sample Collection Soil samples were collected each October, April, and July between Fall 2019 and Spring 2022 (8 collections total). Soil samples were collected using a Giddings soil coring probe with a 27 mm diameter (Giddings Machine Company, USA). The sample density was approximately three per hectare, repeated at the same location each season. The 90 cm deep cores were split into three depths: 0-30 cm, 30-60 cm, and 60-90 cm. Samples were oven-dried at 60°C, ground, and passed through a 2 mm sieve prior to analysis. 8 Samples were evaluated for soil NO3- at the University of Minnesota Research Analytical Laboratory using a LACHAT QuikChem® 8500 Series 2 (Lachat, USA) in 2020 and 2021, and by a SEAL AQ 400 Discrete Analyzer (SEAL, United Kingdom) 2 M KCl (EPA-114-C method) in 2022. NO3- values were converted from mg kg-1 to kg ha-1 using soil bulk density. Water Sample Collection In the NW corner of the research plots, two lift stations collected subsurface drainage and surface water. Water samples were collected by AS950 Hach Portable Water Samplers (Hach, USA) at each station, collecting a representative sample from the water stored in the lift station. Each sample represents a 24-hour period; every 6 hours, the sampler pumped approximately 200 mL of water, generating an 800 mL daily composite sample. The two large lift stations collecting subsurface and surface water act as a controlled drainage structure, only pumping water when the water level hits a certain level (approximately 2.25 m3 during 2020 and 1.14 m3 in 2021), dependent on the amount of water entering the station. During this study, the tanks stored at maximum 2.25 m3 of water before a pumping event. Upon return to the laboratory, samples were either immediately prepared for analysis, or kept frozen (below 4°C) to prevent degradation. A portion of each sample was filtered with a 0.45 μm pore diameter membrane filter and analyzed for Nitrate- Nitrogen (NO3-N). Unfiltered samples were analyzed for Total N (TN) following alkaline persulfate oxidation (Patton and Kryskalla, 2003). In 2020, samples were analyzed using flow-injection analysis with a LACHAT QuikChem® 8500 Series 2 (QuikChem® 9 Method 10-107-04-4-C). In 2021, samples were analyzed using a SEAL AQ 400 Discrete Analyzer (EPA-126-C method). For missing concentration values, linear interpolation of concentrations was utilized as described by Williams et al. (2015), which has been used by others (Hanrahan et al., 2020; Van Esbroeck et al., 2016). Additional uncertainty for concentrations range between 5% and 15%, as outlined in the parameters of Williams et al. (2015). Although many studies report water data in the context of a water year (October – September), this study grouped the two-year study into growing seasons (2020 and 2021). Previous studies in similar climates also represent data on the growing season basis, as the upper soil profile is generally frozen between November-March (Kokulan et al., 2019a, 2019b). For seasonal analysis of flow characteristics (discharge, concentration, and load) each year was subdivided into two periods: between May-June (referred to as early growing season) and July-October (referred to as the late growing season). The two lift stations in this study were not operational between November-April due to negligible water flow. Discharge and Nutrient Loads Within each lift station, a HOBO Water Level Data Logger - UL2L-01 (Onset, USA) was deployed. Each logger records water pressure (PSI) and water temperature (°F) at five-minute intervals. In addition, a HOBO Water Level Data Logger - U20-001- 01 (Onset, USA) was deployed to record barometric pressure (PSI) and air temperature (°C) at five (2019 and 2020) or ten (2021) minute intervals. Barometric pressure (equivalent to water level) was used to calculate the change in water level at each recorded interval. This value was then converted to discharge and 10 nutrient loads for Total Nitrogen (TN), Nitrate (NO3-N), and Ammonium (in 2021; NH4+) were calculated using discharge and nutrient concentration. Drained plot load values were calculated by adding surface and subsurface nutrient loads. Undrained plots only included only surface nutrient loads. Water-Extractable Organic Nitrogen Water-extractable organic N (WEON), a measure of microbially-available substrate in soil, was conducted on 16 subsamples (Haney et al., 2012; method adapted from Haney et al., 2012). Samples of 4.0 g of ground, dried soil and 40 mL of deionized water were placed into centrifuge tubes. Samples were shaken for 10 minutes on a reciprocal shaker, then centrifuged for 5 minutes at 3500 rpm. The supernatant was filtered through Whatman 2V filter paper, placed into glass vials, and diluted with deionized water for analysis (samples were 1:4 diluted with deionized water) to create a 19.9 mL sample; 0.1 mL of 2 M hydrochloric acid was added to each sample. Samples were then mixed and measured in a Shimadzu TOC-L Laboratory Total Organic Carbon Analyzer (Shimadzu Scientific, USA) for total nitrogen (TN). Samples were also evaluated at the University of Minnesota Research Analytical Laboratory for inorganic forms of N (nitrates, nitrites, and ammonium; LACHAT QuikChem® 8500 Series 2, Lachat, USA). These values were then subtracted from TN to calculate WEON. Samples were mass corrected for soil water content. Greenhouse Gas Measurements Between April and October of 2020 and 2021, greenhouse gas fluxes were measured weekly in each of the four plots. In each plot, five greenhouse gas static chambers were used to collect samples (see Collier et al., 2014 for discussion of this 11 methodology). A cylindrical anchor was tamped down into the soil in five adjacent (approximately 1 meter apart) locations in each plot, preventing the lateral flow of gasses. During sampling, a chamber lid was sealed onto the anchor using a rubber slip, trapping gasses within the chamber, allowing for 10-mL gas samples to be collected at times 0, 20 and 40 minutes. The syringes were flushed with air between each sample. Additionally, soil moisture (ML3 Theta Probe, Dynamax, USA), soil temperature, and air temperature (HM Digital Thermometer, Fresh Water Systems, USA) were measured at each chamber, as well as the Barometric pressure at the field site. Notes on weather and soil conditions were also recorded. Additionally, standards of known gas concentrations were collected on each sampling date, to calibrate degradation of samples. Gas samples were analyzed using a G2308 Picarro Gas Concentration Analyzer (Picarro, USA). Nitrous oxide (N2O) concentrations were measured in each vial, along with three N2O standards of 1, 10, and 100 ppm. Concentrations in each sample were then adjusted to determine flux using both environmental conditions and concentrations of standards: first, concentrations of standards were used to generate non-linear regressions. Samples were then plotted along these curves using the T0, T1, and T2 timestamps. Here, these values were also adjusted based on barometric pressure, air temperature, and soil surface area to measure flux in μmol/m2min. Finally, the three timestamps were combined to determine one linear gas flux rate, based on the volume of the chambers. Final flux values were reported in g N ha-1 d-1. Cumulative N2O values were calculated using the trapezoidal integration method. 12 Mineralization Twice a month from April to October (2021), soil cores were collected at each greenhouse gas sampling location to conduct N mineralization measurements. An in-situ core sampling technique was used (Fernández et al., 2017). One schedule 40 PVC pipe (15-cm inner diameter and 13-cm height) was driven into the soil in each plot and capped and retrieved bi-weekly (“final” sample). An “initial” sample was also collected, to standardize readings. Soil samples were then lightly crushed and homogenized. 2.0 g of crushed soil were analyzed for NH4+ and NO3- using SEAL AQ 400 Discrete Analyzer 2 M KCl (AGR-231-A method). Inorganic nitrogen levels were then adjusted to kg ha-1 based on the volume of mineralization cores. Samples were mass corrected for soil water content. Nitrification and ammonification rates were determined by calculating the change in inorganic nitrogen between the “initial” and “final” soil samples, and mineralization was calculated by summing nitrification and ammonification. Nitrogen Uptake During the 2021 growing season, total nitrogen uptake (TNUP) was measured on the same 16 sample points measured for WEON. At each location, physiologically mature wheat plants were sampled by cutting the base of the stem a centimeter above the soil surface. These plants were then oven-dried at 50°C, ground, and weighed to pass through a 1mm sieve. Four subsamples were analyzed for N using the dry combustion method (Matejovic et al., 1995). TNUP was then calculated using biomass yield. Statistical Analysis All statistical analysis was completed using the R programming language (R Development Core Team, 2014). For edge of field data, non-parametric statistical tests 13 were utilized to identify significant differences in data (comparisons between treatments, time scales, etc.). Differences were evaluated using two-sample Wilcoxon Signed Rank Tests. For soil and gas measurements, random intercept mixed models were utilized to compare differences between drained and undrained treatments over time. Models which included both random slope and random intercept included WEON, soil NO3- , N2O, and N mineralization. Variance analysis (grain yields) was assessed using F-tests. Correlations (N2O) were determined using the Kendall Rank Correlation Coefficient. Prior to analysis, all data were evaluated for normality, and transformed if necessary (if it is a required assumption of that test; WEON was log-transformed). All statistical thresholds were at the 0.05 level of significance. 1.4. Results Edge of field Losses Discharge and Precipitation During the 2020 NW Minnesota growing season (approximately April 1 - September 30 but can shift dependent on crop and yearly climate), precipitation at the experimental site was 475.59 (Figure 1.4). During the 2021 growing season, precipitation at the experimental site was 209.30 mm, considerably lower than the 30-year normal of 424 mm for Crookston, MN (1990 - 2019; Midwest Regional Climate Center). During the period of measurement in 2020, 1.5% of precipitation became edge of field discharge through either surface or subsurface discharge (Figure 1.4). In 2021, 0.14% of precipitation became edge of field discharge. It is important to note that the lift stations stored approximately 1.14 – 2.25 m3 of potential discharge at any time, decreasing this percentage from expected values. 14 Annual subsurface discharge (plots 1 and 3) was significantly greater than annual surface (plots 1, 2, 3, and 4) discharge both years (p < 0.0001; Figure 1.5). Between 63 - 99% (median of 80%) of edge of field losses were from subsurface drainage with a larger share in 2020 compared to 2021 (p < 0.0001). In 2020, 75% of total edge of field discharge was from the subsurface. In 2021, 69% of edge of field discharge was from the subsurface. In 2020, 1.22 mm of discharge was expressed as surface drainage and 5.31 mm from subsurface drains (Figure 1.5A). In 2021, 0.076 mm of discharge left the field in surface drains, and 0.36 mm discharged through subsurface drainage (Figure 1.5B). Edge of field TN and NO3-N When comparing loss pathways, during periods of normal-high precipitation (2020), flow-weighted concentrations of TN and NO3-N were significantly greater in subsurface drains than surface runoff (Table 1.3 for full comparisons and p-values). However, during much of 2021, flow-weighted concentrations of TN and NO3-N were significantly greater in surface runoff. The exception was late 2021 TN, where there was no significant difference between subsurface and surface drains. Notably, NH4-N (evaluated in 2021 only) does not follow this pattern, with flow-weighted concentrations higher in subsurface drains than in surface runoff. Examining seasonal variation during each year, during normal to high precipitation, flow-weighted concentrations of both TN and NO3-N were significantly higher in surface runoff in the early portion than the later portion of the growing season (Table 1.3; Figure 1.7). However, for subsurface drainage, flow-weighted concentrations 15 for TN and NO3-N were greater in the later portion of the growing season than the earlier portion. The patterns are more complicated in 2021, when precipitation slowly increased over the course of the growing season: flow-weighted NO3-N concentration in both surface and subsurface drainage is significantly greater in the early portion of the growing season versus the later portion of the growing season (Table 1.3). Conversely, in surface and subsurface drainage, flow-weighted TN concentration is significantly greater in the later portion of the growing season versus the earlier portion of the season (Table 1.3). For flow-weighted NH4-N concentration, concentrations were greater in the later portion of the growing season than the earlier portions. Nutrient loading Over the course of the study, as well as individually in 2020 and 2021, the drained treatment had significantly larger loads of NO3-N, NH4-N, and TN than the undrained plots (see Figure 1.7 for representation of each loading event over the course of the study, Table 1.4; p < 0.0001). Subsurface drainage accounted for approximately 81% of the total load and was statistically greater than surface runoff (p < 0.01; Figure 1.8). Examining daily loads, during periods of higher precipitation, loads of most N species were significantly greater in subsurface than surface runoff (Table 1.5). This trend is true for all of 2020, and parts of 2021. However, there is one main exception: in both early and late 2021 (and overall in 2021), NO3-N loads were not significantly different between subsurface and surface runoff (p = 0.16 for early 2021 and p = 0.10 for late 2021). Although drought conditions 16 in 2021 generated little differences in trends of N loads, October 2021’s large precipitation totals generated substantially higher loads, more similar to the loads seen in 2020 (Table 1.5). Comparing seasonal variation, in 2020, daily loads were significantly greater in the early portion of the growing season versus the later portion of the growing season, largely following precipitation trends. This is true for both NO3-N and TN. Similarly in 2021, loads largely followed precipitation patterns, with larger loads in the late portion compared to the early portion of the growing season, apart from surface NO3-N loads, where the late portion of the growing season only had a marginally greater load than the early portion (p = 0.06). Over the course of the study, relative percentages of pathways and species for N loads varied (Figure 1.8; Figure 1.9). Approximately 65% of loading was NO3-N, compared to 35% of loading was NH4-N and organic forms of N. This varied between the two years: in 2020 with higher precipitation, 82% of loading was NO3-N, and in 2021 with lower precipitation, only 49% of loading was NO3-N (Figure 1.8). Nitrous Oxide Emissions In 2020 and 2021, there were no statistical differences for N2O fluxes between the drained and undrained treatments on each sampling date or across the growing season (Figure 1.10). However, in both years, there are clear trends: the undrained treatment had higher cumulative emissions across the entire growing seasons both years (Figure 1.10). In 2021, there are large emissions events at the start of the growing season particularly in the undrained treatments, and as moisture decreases, emissions plateau in both treatments. Emissions are greater in 2021 than 2020 for both treatments (Figure 1.10). 17 In 2020, N2O fluxes correlated with soil moisture (p < 0.01), but did not correlate with soil temperature, air temperature, or barometric pressure. In 2021, there were no correlations with any environmental factors. Soil Nitrogen Transformations Mineralization In 2021, both the drained and undrained plots had net negative ammonification throughout the growing season, with the undrained treatment having higher magnitude negative ammonification (Figure 1.12; Table 1.6). This was likely an intermediate step, since much of this N was then nitrified, and both treatments had net positive nitrification (Figure 1.12). For both treatments, ammonification and nitrification generally demonstrated inverse patterns throughout the growing season. Ammonification was greatest for both treatments at the start of the growing season and leveled off in later months. Nitrification showed a similar pattern, but also decreased in both treatments towards the end of the growing season. The undrained plots had higher magnitude nitrification than drained plots. Overall, cumulative net N mineralization was negative in all plots due to the larger magnitude of ammonification than nitrification. The undrained plots had a more negative cumulative mineralization than the undrained plots (Figure 1.12c). Although trends exist, there were no significant differences between drainage treatments. WEON At all three depth increments (0 - 30, 30 - 60, and 60 - 90 cm) there was no significant difference between drainage treatments over the course of the study (p = 0.75, 18 p = 0.81, p = 0.63; Figure 1.13). Sampling date was significant between 0 - 30 cm (p < 0.0001) and 30 - 60 cm (p < 0.0001), but not 60 - 90 cm (p = 0.24). Additionally, the interaction between drainage and sampling season is significant from 0 - 30 cm (p < 0.05), but only marginally significant from 30 - 60 cm (p = 0.06) and 60 - 90 cm (p = 0.23). Within any single season, the drainage effect was not significant. Soil Nitrate Soil NO3- varied throughout the course of the study in the two treatments. In 2020, there were no significant differences between the drained and undrained treatment at all depths (Figure 1.14). Following one year of drainage and high precipitation, the drained treatment had significantly (p < 0.05) lower NO3- than undrained at the 0 – 30 cm increment, and marginally lower (p < 0.1) at the 30 – 60 cm increment. After two years of drainage and the 2021 growing season with less precipitation, trends to a lesser extent demonstrated decreased NO3- at the 0 – 30 cm increment in the drained treatment, similar to 2020. Notably, trends demonstrated increased soil NO3- between 30 – 60 cm (p < 0.1) and 60 – 90 cm (p < 0.1) in the drained treatment compared to the undrained treatment, a trend not demonstrated in 2020. Grain Yield and Uptake The grain yields varied during the study years (Figure 1.15). In 2019 wheat yield was 4200 ± 210 kg ha-1, in 2020 soybean yield was 2200 ± 460 kg ha-1, and in 2021 wheat yield was 3300 ± 350 kg ha-1 (comparisons with average Polk County Yield can be found in Table 1.7). Additionally, there were differences between the drained and undrained plots, particularly in 2020 and to a lesser extent in 2021 (Table 1.7). For the 19 one year prior to drainage installation, there was no significant difference between yields in the drained and undrained plots (p = 0.92). In 2020, with average precipitation, there was a trend for greater years in the drained treatment (19.73% higher yield) than the undrained treatment. In 2021, with lower-than-average precipitation during the wheat growing season, there was a trend of increased yields in the undrained treatment (9.39% higher yield) than the drained treatment. While changes in yield between years are difficult to compare due to crop rotation, wheat in 2019 and 2021 can be compared: while the undrained treatment saw a 18.35% decrease from 2019 to 2021, the drained treatment saw a 25.79% decrease between the two years, demonstrating a potential interaction between drainage and precipitation. Between 2020 and 2021, the undrained treatment had greater variance in grain yield than the drained treatment (p < 0.001). Wheat TNUP in 2021 ranged from 43.02 to 75.64 kg ha-1, with an average of 59.33 ± 8.06 kg ha-1 (Table 1.8). There was no statistical difference between the drained and undrained treatments (p = 0.81). Loss Comparisons A comparison of N losses measured in this study can be found in Table 1.9. Notably, in 2020, losses were greater in the drained treatment than the undrained, mainly facilitated by subsurface drainage losses during periods of high precipitation. However, in 2021 both treatments had much similar N losses, despite N application (Figure 1.7). 1.5. Discussion Following subsurface drainage installation, decreases in soil NO3- (likely facilitated by subsurface drainage) appear to be the main mechanism for N losses from 20 the soil profile. However, variation in crop productivity and NO3- demand may also explain differences in soil NO3- between treatments, which occurred most prominently in the upper portion of the soil profile. Secondary and smaller pathways for N transformation included increased mineralization in the drained treatment, and increased denitrification in the undrained treatment. Aside from drainage treatment, differences between mineralization rates and N2O emissions may at least partially be explained by other factors such as seasonal, management, and climatic factors. It is important to note that while this study did not capture a complete representation of all nitrogen forms in these systems, drainage treatment comparisons are still meaningful. Although not statistically different, trends from the first two years supported our hypothesis that drainage treatment would decrease N2O emissions in the drained treatment. Reduced N2O emissions is often considered a potential positive environmental impact of subsurface drainage (Castellano et al., 2019; Fernández et al., 2016). N2O emissions are soil-moisture and soil-texture dependent. Ideal conditions for denitrification very by soil texture and the water-filled pore space (WFPS), but the range is typically ~60-95% moisture (Banerjee et al., 2016; Ciarlo et al., 2007; Dobbie and Smith, 2003 Marinho et al., 2004; Schindlbacher et al., 2004). Generally, poorly drained soils have higher rates of N2O emissions (Bandibas et al., 1994; Ciarlo et al., 2007; Marinho et al., 2004; Schindlbacher et al., 2004; Wang et al., 2022; Webster and Hopkins, 1995; Weitz et al., 2001). Likewise, studies have found that similar soils have lower emissions under drained than undrained conditions (Dobbie and Smith, 2006; Fernández et al., 2016; Krichels et al., 2019; Kumar et al., 2014; Snyder et al., 2009). N2O fluxes only correlated to soil moisture in 2020 but not in 2021, potentially explained 21 by the drought conditions in 2021 which generated low to negative fluxes between July and October. Interestingly, emissions did not increase with the increase in precipitation that occurred near the end of the 2021 growing season (Figure 1.10D) although soil moisture levels were higher in the undrained sections (Figure 1.11). The magnitude of difference in emissions between the drained and undrained treatments are less than other studies (Dobbie and Smith, 2006, Fernández et al., 2016, and Kumar et al., 2014). It is possible that the composition of microbial denitrifiers and legacy moisture may be a complicating factor altering N2O emissions in this study (Banerjee et al., 2016; Krichels et al., 2019; Wang et al., 2022). Banerjee et al. (2016) found that while current moisture conditions correlated with N2O emissions, soil moisture hysteresis can also change microbial activity, composition, and ultimately gaseous emissions. Groffman and Tiedje (1988) and Krichels et al. (2019) found that historical soil drainage can inform N2O emissions due to different stimulation of denitrifiers. Given that this study examines the first two years following subsurface drainage installation, it can be assumed that both treatments had consistent moisture conditions prior to drainage installation and may indicate similar microbial communities in both treatments. Because of this, it may take many growing seasons (Averill et al., 2016; Evans and Wallenstein, 2012) to substantially alter the microbial communities responsible for identification and demonstrate greater differences in emissions between the drained and undrained treatment seen in other studies. In 2021, N2O emissions were influenced by fertilizer application. N2O emissions are known to be closely related to the quantities of applied N fertilizer, and likely indicate the differences in emissions between 2020 and 2021 (Shcherbak et al., 2014; Gramlich et 22 al., 2018). Additionally, the large emissions spike in the undrained treatment in 2021 is likely related to the combination of high soil moisture following snowmelt and fertilizer application (Fernández et al., 2015; Venterea et al., 2012). Soil N transformations, including mineralization (ammonification and nitrification) have been commonly linked to changes in soil moisture (frequent wet-dry cycles can promote mineralization; Goncalves and Carlyle, 1994; Guntiñas et al., 2012; Sierra, 1997) as well as inputs of N fertilizers through a priming response (Kuzyakov et al., 2000), and are controlled by heterotrophic microorganisms which require O2 and moisture. However, studies have demonstrated a complicated connection between moisture and mineralization processes (Eickenscheidt and Brumme, 2013; Guo et al., 2014). Although not statistically significant, there were trends related to moisture and mineralization processes in this study. From mid-June onward, the undrained treatment had higher moisture content in the soil profile than the drained treatment, and higher O2 levels promoted more nitrification, less ammonification, and higher mineralization than the drained treatments (Figure 1.12). This may also be caused by higher soil N levels in the undrained fields (also seen in Fernández et al., 2017). Although we expected higher mineralization rate in the drained treatment, the low moisture content in 2021 greatly reduced the magnitude of nitrification and increased the magnitude of ammonification, causing overall net negative mineralization for both treatments, as drought conditions limited the microbial activity to complete these processes. Similar results have been observed in other mineralization studies under dry conditions (Cassman and Munns, 1980; Fernández et al., 2017; Fu-sheng et al., 2005). Under more typical precipitation, we 23 may expect lower ammonification levels and net positive mineralization (Paiao et al., 2021). The differences between drainage treatments are most variable at the start of the growing season, but as the soil profile dried, the differences remained fairly consistent (Evans et al., 1998). Negative cumulative mineralization indicates higher levels of immobilization, which occurred most extremely during the drought conditions. These differences in nitrification and mineralization likely resulted in differences in N uptake by the wheat plants throughout the 2020 and 2021 growing season. N plant uptake in 2021 demonstrated a trend of higher uptake in the undrained treatment than the drained treatment, although not statistically significant. In 2020, when neither N nor moisture were yield-limiting factors, the drained treatment had higher yields than undrained (Figure 1.12). Aeration from drainage likely caused increased mineralization, which facilitated both greater plant uptake and losses via drainage and is demonstrated by the decrease in soil NO3- relative to the undrained section, which had lower yields and losses. In 2021, moisture (not N) was the yield-limiting factor, and the higher moisture in the undrained conditions allowed for greater TNUP and yield despite drought conditions (Paul and Myers, 1971; Ruckauf et al., 2004), particularly because the clay-rich soils retained moisture from 2020 (at the end of the 2020 growing season, the undrained treatment had higher moisture; Figure 1.11). There are two pools of organic N quantified in this study: Dissolved Organic Nitrogen (DON; in this study, organics measured in water samples) that left the field through surface and tile lines, and WEON (also known as Soluble Organic N or Extractable Organic N), a measure of organically bound N (Ros et al., 2009). While there were significantly greater losses of DON in the drained than undrained treatments, there 24 were no significant differences in WEON pools between the two treatments. This is likely due to differences between WEON and DON pools. While they are controlled by similar processes (Kalbitz et al., 2000; Bregliani et al., 2010), WEON is controlled by biological N cycling and varies seasonally and with soil depth (Ros et al., 2009). Therefore, it is generally more stable outside of these temporal variations, whereas DON changes are more related to hydrology and water fluxes in the soil profile and therefore losses can occur more rapidly (van Kessel et al., 2009). This is consistent with our findings, where WEON changes over the course of this study did not vary by treatment, but instead by depth and season (highest in Spring and Summer and near the top of the soil profile; see Crookston et al., 2021 for discussion of temporal variability of soil health indicators including WEON), while DON varied more with precipitation changes as seen through the relative percentages of N species in loads throughout the two years. NO3-N loads through subsurface drains have long been recognized and understood (Baker et al., 1975; Baker and Johnson, 1981; Gast et al., 1978; Gilliam et al., 1979; Jacobs and Gilliam, 1985; Randall and Mulla, 2001; Skaggs and Gilliam, 1981). These losses occurred during times of high precipitation, particularly in most of 2020 and October 2021, consistent with other studies, as NO3-N leaches from soils most heavily during periods of high precipitation and high discharge (Drury et al., 1993; Randall and Goss, 2008; Randall and Mulla, 2001, Jaynes et al. 2001). Although short-term NO3-N load responses can be impacted by antecedent soil moisture levels (Baeumler and Gupta, 2020; Davis et al., 2014; Wolf et al., 2020), the substantially higher loads and discharge across 2020 were likely related to timing of precipitation events and field capacity: in 2020, the above-average precipitation in the spring-summer, versus the above-average 25 precipitation in the in late summer-fall resulted in a higher discharge in 2020 than 2021. In the spring-summer, evapotranspiration losses were minimized, and the field is closer to field capacity from snowmelt (Randall and Iragvarapu, 1995), compared to in the fall when evapotranspiration is higher, and the fields were dried out over the summer (Randall and Goss, 2008). Thus, although precipitation levels were relatively high in late summer-fall 2021, the soil was able to infiltrate and store precipitation with little discharge. As more precipitation occurred, the field eventually moved closer to field capacity in the later summer-fall, from a few higher load events that occurred in October 2021 (Figure 1.9). Overall, TN and NO3-N loads were controlled by precipitation, and less so by other factors. When looking at concentrations, TN followed precipitation patterns closely over both years, while NO3-N concentrations were impacted by antecedent NO3-N concentrations and fertilizer application (similar patterns observed in Gentry et al., 1998; Korsaeth and Eltun, 2000; Rankinen et al., 2017). In 2020, TN concentrations increased over the course of the growing season and peak in August, while NO3-N concentrations are highest at the start of the growing season. NO3- levels are high in the spring after winter thaw and early spring mineralization, then declines due to leaching, crop uptake, and immobilization (Baeumler and Gupter, 2020). In 2021, a different pattern was seen: TN concentrations peak with highest precipitation near the end of the growing season while NO3-N concentrations were highest at the start of the growing season, following fertilizer application (e.g., Jaynes et al., 2001). These changes could also be related to plant uptake of N, with crop uptake generally decreasing the quantity of NO3-N available to be leached from the soil profile (Randall and Goss, 2008). 26 The discharge and load rates in this study were much lower than many comparable studies, particularly in drought conditions in 2021 (Helmers et al., 2022). In a comprehensive study of controlled drainage across the Midwestern United States, Helmers et al. (2022) found significantly higher discharge and loading rates in all Upper Midwest sites (although all sites received more precipitation than this study). Helmers et al. (2022) found 28 - 49% of precipitation became the edge of field discharge, above the 1.5% and 0.14% in 2020 and 2021 of this study. Part of this may be explained by the clay-rich soils of the RRB, however some is likely also due to drainage design in this study (see below). A study in Ontario, Canada with more similar precipitation patterns more closely aligned with our discharge values, which were as low as 9 mm (approximately 3.4% of total precipitation expressed as edge of field discharge; Sunohara et al., 2014) from tile drains during a dry year. Management Implications We found that precipitation was a major factor in N losses in recently drained soils. Particularly in years with heavy precipitation, it is likely that drained fields will lose more nitrogen to subsurface drains, but drainage may also increase plant N uptake and reduce gaseous emissions. These differences are minimized in years with lower precipitation, as drainage treatment effect is only existent during normal to high precipitation. Utilizing water management systems such as controlled drainage may also help reduce nutrient loads and reduce yield variations during years with extreme precipitation. With more interest in promoting the sustainable intensification of agricultural drainage (Castellano et al., 2019), more N rate trials in drained fields are needed to 27 explore the possibility of being able to lower N rates with drainage. This study provides an examination of the N cycle in the first years following drainage installation. However, continued monitoring N cycling in these fields with integrated N rate trials will provide a more robust understanding of sustainable intensification of agricultural drainage with changing precipitation patterns in the decades to come. 28 Table 1.1. Information about crops, N application, yield, and key dates at experimental site. Year Crop Historical N Application Rate (kg ha-1) and Timing Yield (kg ha-1) Plant Date Harvest Date 2019 Wheat 162.5 Fall, 112 Spring 4210 5/9/2019 9/2/2019 2020 Soybeans N/A 2160 5/20/2020 10/2/2020 2021 Wheat 162.5 Fall, 112 Spring 3280 4/27/2021 8/3/2021 Table 1.2. Average daily temperature and average precipitation of the 10 years prior to study and each year of the study. Year Average Daily Temperature (°C) Average Precipitation (cm) 2009-2018 4.49 50.32 2019 2.56 68.76 2020 3.72 52.17 2021 6.11 43.09 29 Table 1.3. Flow-weighted concentrations of NO3-N and TN in surface and subsurface drains. Upper case letters signify significance (p < 0.05) between time periods (e.g., Early 2020 Surface TN vs Late 2020 Surface TN), lower case letters for surface/subsurface comparisons (e.g., Surface Early 2020 TN vs Subsurface Early 2020 TN). Period Surface Subsurface NO3-N NH4-N TN NO3-N NH4-N TN ____________________________________________ mg L-1__________________________________ Early 2020 3.05 ± 1.04A a N/A 3.53 ± 1.08A a 6.75 ± 3.55A b N/A 7.56 ± 3.64A b Late 2020 2.25 ± 0.44B a N/A 2.51 ± 0.49B a 8.72 ± 1.82B b N/A 11.62 ± 1.78B b Early 2021 1.69 ± 0.47A a 0.13± 0.14A a 2.29 ± 0.61A a 1.19 ± 0.74A b 0.24 ± 0.31A b 2.09 ± 1.20A b Late 2021 1.53 ± 1.44B a 0.54 ± 0.68B a 3.98 ± 3.03B a 0.78 ± 0.48B b 0.66 ± 0.75B b 5.13 ± 3.35B a 2020 Total 2.64 ± 0.89A a N/A 3.00 ± 0.97A a 7.78 ± 2.95A b N/A 9.68 ± 3.48A b 2021 Total 1.61 ± 1.06B a 0.33 ± 0.53a 3.12 ± 2.31B a 0.99 ± 0.66B b 0.45 ± 0.61b 3.58 ± 2.92B a Total 2.09 ± 1.11a 0.33 ± 0.53a 3.07 ± 1.82a 4.14 ± 3.97a 0.45 ± 0.61b 6.42 ± 4.41b Table 1.4. Nutrient loads for undrained and drained plots. Statistical significance (p < 0.05) is indicated only between undrained and drained plots for each N species within the same period. Organic N is non NH4-N organic species. Undrained Drained NO3-N NH4-N Organic N TN NO3-N NH4-N Organic N TN ____________________________________________ kg ha-1________________________________________ 2020 2.28a N/A 0.37a 2.65a 40.97b N/A 6.53b 47.51b 2021 0.16a 0.03a 0.16a 0.35a 0.56b 0.27b 2.57b 3.13b Total 2.45a 0.028a 0.53a 3.00a 41.52b 0.27b 9.10b 50.63b 30 Table 1.5. Seasonal, yearly, and total cumulative loads from different sources over the period of study. Statistical significance (p < 0.05) indicates average daily load during time period, not comparison of cumulative load. Upper case letters signify significance between time periods (e.g., Early 2020 Surface TN Load vs Late 2020 Surface TN Load), lower case letters for surface/subsurface comparisons (e.g., Surface Early 2020 TN Load vs Subsurface Early 2020 TN Load). Period Surface Subsurface NO3-N Load NH4-N Load TN Load NO3-N Load NH4-N Load TN Load __________________________________________ kg ha-1__________________________________________ Early 2020 2.20A a N/A 2.56A a 28.05A b N/A 31.40A b Late 2020 0.09B a N/A 0.10B a 10.63B b N/A 13.46B b Early 2021 0.06A a 0.007A a 0.08A a 0.06A a 0.029A b 0.11A b Late 2021 0.10A a 0.021B a 0.27B a 0.33B a 0.21B b 2.66B b 2020 Total 2.28A a N/A 2.65A a 38.68A b N/A 44.85A b 2021 Total 0.16B a 0.028a 0.35B a 0.39B a 0.24b 2.77B b Total 2.45a 0.028a 3.00a 39.08b 0.24b 47.63b Table 1.6. Net ammonification, nitrification, and mineralization from the drained and undrained plots. Undrained Drained _______ kg N ha-1 _______ Net Ammonification -137.48 -104.15 Net Nitrification 73.88 66.05 Net Mineralization -63.59 -38.10 31 Table 1.7. Grain yield for 2019 to 2021 by drainage treatment. Wheat was planted in 2019 and 2021 and soybean in 2020. Polk County average yields are sourced from USDA NASS. Year Undrained Yield Mean Undrained Yield Range Drained Yield Mean Drained Yield Range Polk County Avg. Yield ____________________________________________ kg ha-1 ________________________________________ 2019 4200 ± 200 3960 - 3720 4220 ± 220 3970 - 4730 3499.88 2020 1970 ± 450 930 - 2800 2350 ± 380 1410 - 2900 2165.30 2021 3430± 310 2530 - 3800 3130 ± 330 2010 - 3700 2679.51 Table 1.8. Total nitrogen uptake (TNUP) by drainage treatment in 2021. Undrained Drained _________________________ kg N ha-1 __________________________ TNUP Mean 60.02 ± 8.33 58.66 ± 7.87 TNUP Range 47.64 - 75.64 43.02 - 70.89 Table 1.9. N losses separated by year and treatment. Crop removal in 2020 was calculated based on a conversation factor in Fernández et al. (2009). Please note that leaching below subsurface drains was not conducted in this study, and that leaching in the subsurface in both drained and undrained treatment occurred. Year Treatment Runoff Subsurface Crop Removal N2O Emissions Total Losses __________________________________ kg N ha-1 _________________________________________ 2020 Drained 2.65 44.86 119.67 3.68 170.86 Undrained 2.65 0 99.93 4.62 107.20 2021 Drained 0.35 2.78 71.31 5.33 79.77 Undrained 0.35 0 80.12 5.70 86.17 32 Figure 1.1: Lake Winnipeg Watershed and location of the field site, in Crookston, MN. Water flows north in the Red River to Lake Winnipeg in Manitoba, Canada. Adapted from Wassenaar and Rao (2012). 33 Figure 1.2: The soil N cycle on the drained portions of the research plots in Crookston, MN. There are also two lift stations that collect water at the edge of field (“Runoff” and “Leaching” mechanisms), which act as controlled drainage. 34 Figure 1.3: Layout of experimental site. Plots 1 and 3 are drained, while 2 and 4 are undrained. Tile drains run south to north in each plot, collected at the tile lift station (northern dot). Surface drains are not shown but run east to west and are collected at the surface lift station (southern dot). 35 Figure 1.4: Surface discharge, subsurface discharge, and precipitation from the experimental site (Midwest Regional Climate Center). In 2020, 1.5% of precipitation was discharged at the edge of field, where only 0.14% was in 2021 Figure 1.5: Discharge from surface and subsurface drains in 2020 and 2021. A) Discharge in mm in 2020 and B) in 2021. Note that the y-axis varies between the two years. 36 Figure 1.6: Monthly concentrations in both surface and subsurface drains in 2020 and 2021. A) NO3-N concentrations and B) TN concentrations. Samples collected from lift station water storage structures. 37 Figure 1.7: Drained and undrained daily NO3-N (open circles) and TN (closed circles) loads by year. To include all points, some 0.0 kg ha-1 loads appear below 0.0. Figure 1.8: Relative percentages of loads during 2020 and 2021. Note that May 2021 is missing due to machine error. NH4-N was grouped with organic N due to no NH4-N data in 2020. 38 Figure 1.9: TN and NO3-N loads from the experimental site in 2020 and 2021. A) NO3- N and B) TN loads from both surface and subsurface drains. Figure 1.10: N2O fluxes and cumulative emissions across the growing season in 2020 and 2021. A) Flux in 2020, B) Cumulative in 2020, C) Flux in 2021, D) Cumulative in 2021. 39 Figure 1.11: Volumetric water content in the top 6 cm of soil for drained and undrained treatment. in 2020 (top) and 2021 (bottom). 40 Figure 1.12: Cumulative A) ammonification, B) nitrification, and C) mineralization during the 2021 growing season averaged across drainage variables. 41 Figure 1.13: WEON separated by depth across the time of study. A) WEON from 0 - 30 cm, B) from 30 - 60 cm, and C) from 60 - 90 cm. There are no significant differences between treatments at all depths and seasons. 42 Figure 1.14: Soil NO3- by depth. Samples collected in spring, prior to planting and represent the impact of treatment from the previous growing season and winter. A star (*) indicates statistically significant results (p < 0.05) Figure 1.15: Grain yield comparisons by year in response to drainage treatment. Grain yield comparisons by year in response to drainage treatment. Yield in 2019 was prior to drainage treatment, 2020 after one year of drainage, and 2021 after two years of drainage. 43 CHAPTER 2: THE LONG-TERM IMPACT OF SUBSURFACE DRAINAGE ON SOIL PHYSICAL PROPERTIES AND BIOLOGICAL INDICATORS 2.1. Synopsis Subsurface drainage is a common practice to support agricultural production and increase yields in poorly drained soils. Following decades of subsurface drainage installation, agricultural fields often have higher water discharge and increased nutrient losses. However, few studies have evaluated the changes in soil properties and soil health metrics associated with long-term drainage of the soil profile. In this study, we sampled three field-scale sites drained over fifteen years prior to sampling and three field-scale sites drained within the last five years. We evaluated three soil physical properties: saturated hydraulic conductivity (Kfs), bulk density, and aggregate stability, as well as three soil health metrics, water-extractable organic carbon (WEOC) and nitrogen (WEON), and potentially mineralizable carbon (PMC). The fields with older drainage systems had significantly greater Kfs, WEON (all depths), WEOC (between 15 - 30 cm), and PMC (15 - 30 cm). There were no significant differences in bulk density, aggregate stability, and WEOC (0 - 15 cm) and PMC (0 - 15 cm). We hypothesize that the increased Kfs is likely the result of increased preferential flow paths and cracks, typical of clay-rich soils found in NW Minnesota. These preferential flow paths could also be acting as biological “hot spots” with increased microbial diversity and activity, indicated by the higher biological indicators in the fields with older drainage systems. Evaluation of soil physical properties and soil health metrics should become a standard practice in drainage research. Understanding how subsurface drainage installation changes long-term 44 soil properties will be valuable to understand how drainage will alter soil characteristics and water movement in the soil profile over time. 2.2. Introduction In the Midwest, subsurface drainage is a common practice to support agricultural production in poorly drained soils. Drainage removes excess water from the soil profile, lowering water tables and increasing field trafficability. This allows for timely field operations, prevents water logging, and increases yields (Abid and Lal, 2009; Evans and Fausey, 2015). Drainage is also common throughout the United States; as of 2017, 22.5 million hectares of agricultural land were reported to have artificial subsurface drainage, a 12.7% increase from 2012 (USDA, 2017). In Minnesota alone, 3.3 million hectares (approximately 32% of statewide agricultural land) utilized subsurface drainage in 2017, a 20.0% increase from 2012 (USDA, 2012 and 2017). Drainage systems in the United States are also aging. In 1994, 12.6 million hectares in the United States had been artificially drained, making a large percentage of subsurface drainage systems are between 10 - 30 years in age (Skaggs et al, 1994). There is a need to understand to what extent older drained systems function differently from the more recently drained systems. Previous studies have tracked long-term nutrient losses and fluxes from subsurface drainage (Kladivko and Bowling, 2021; Kladviko et al., 2004), while others have utilized models to simulate changes in hydrology or nutrient losses due to subsurface drainage (Ahmed et al., 2007; Jin et al., 2003; Thorp et al., 2008). However, relatively few studies have evaluated changes in soil properties and soil health metrics associated with long-term subsurface drainage, particularly in Midwest cropping systems (Talukolaee et al., 2018; Welage, 2020). In their 31-year long-term drainage study, 45 Kladivko and Bowling (2021) note increasing trends in both volume and peak flow rates over time. They hypothesize that these changes may be due to changes in soil properties from drainage. In our study, we explore the potential soil mechanisms for observed changes in drainage discharge by evaluating soil physical and biological properties in fields with different drainage ages. In addition, this study’s location in the Red River of the North Basin (RRB) has a unique context with poorly drained, clay-rich soils. Soil Physical Properties Subsurface drainage removes excess water which has infiltrated the soil profile that cannot be held within the soil matrix or by capillary forces against gravitational forces (Fausey, 2005). These changes in water movement that drainage facilitates in the soil profile can alter the physical properties of soil over time. The impact of drainage on soil physical properties has been a long-standing interest (Hundal et al., 1976). Of particular interest has been the saturated hydraulic conductivity (Kfs) of soil, which measures the potential of a saturated soil to transport water and is an important component of drainage modeling (e.g., Alakukku and Turtola, 2010; Bouma et al., 1979; Hundal et al., 1976; Jia et al., 2008; Wellage, 2020). Understanding how saturated hydraulic conductivity changes over the lifespan of a drainage system informs growers on the timing and length of field saturation following large precipitation events. Additionally, preferential flow, or the rapid transportation of water and nutrients through the soil profile via macropores (Bouma, 1981), often leads to increased nutrient losses through drainage systems (Stone and Wilson, 2006). In systems with long-term drainage, we expect preferential flow paths and cracks to develop in soil over time with increased drainage, increasing infiltration and Kfs (Kladivko and Bowling, 2021). 46 Another property of interest is bulk density, a measurement of soil compaction which informs water and nutrient movement, as well as aeration and root growth (Hundal et al., 1976). Bulk density has been shown to decrease with the installation of subsurface drainage systems, although the effects are of varying magnitudes or not statistically significant (Baker et al., 2004, Baker et al., 2010; Chow et al. 1993; Hundal et al., 1976; Jia et al., 2008; Lal and Fausey, 1993; Wells and Williams et al., 1996; Wellage, 2020). Aggregate properties inform porosity, aeration, infiltration, crop growth, and other water-related soil properties, and is an important soil physical property when quantifying soil quality (Amézketa, 1999). The impact of subsurface drainage on aggregate stability has thus garnered interest, with some studies noting an increase in aggregate stability due to the installation of subsurface drainage (Abid and Lal, 2009; Baker et al., 2004; Hundal et al., 1976), and others finding its impact to be negligible or decreasing aggregation properties (Kumar et al., 2014; Lal and Fausey, 1993). Together, Kfs, bulk density, and aggregate stability provide a wide understanding of how soil plant growth, precipitation, and physical properties may change over years of subsurface drainage. Soil Biological Indicators As drainage changes the water table in fields, it can also alter soil biological processes in the soil ecosystem (i.e., oxic versus anoxic conditions) and microorganism activity. However, few studies have examined the impact of drainage on soil biological metrics. Some studies have noted the loss of carbon in dissolved form following subsurface drainage and the concurrent depletion of soil organic carbon stocks, though most discussions of nutrient losses have focused on nitrogen or phosphorus losses 47 (Jacinthe et al., 2001; Kumar et al., 2014; McTiernan et al., 2001; Ruark et al., 2009). Still, few studies have focused specifically on soil health metrics, including those commonly used by farmers in the Haney Test (also known as the Soil Health Tool, commonly used by many laboratories and the NRCS to evaluate soil health; Haney et al., 2012). These metrics include potentially mineralizable carbon (PMC), water-extractable organic carbon (WEOC), and water-extractable organic nitrogen (WEON). As interest in soil health has increased demand for soil biological testing, assessing baseline expectations for undrained and undrained conditions is critical for interpreting results. PMC measures part of the microbially available soil C and is likely to be connected to soil moisture content. While there has been interest in the relationship between drainage and PMC (Beauchamp et al., 1980), little research has studies PMC in drained fields at length or compared PMC in drained and undrained fields (Das et al., 2019, Jauhiainen et al., 2016, Kovalev et al., 2021, Libohova et al., 2014). Changes to WEOC and WEON, both measures of microbially available substrate, are commonly lost in drainage water with other forms of dissolved carbon; both have been rarely studied, particularly in relation to changes in management (Beauchamp et al., 1980, Grebliunas et al., 2016). Examining how PMC, WEOC, and WEON change over time may provide valuable information on how microbially available C pools evolve with continuous drainage. Providing more baselines and interpretations to the established Haney Test will help further the applicability of these indicators, generating a common framework for soil health practices (Stewart et al., 2018). In this study, we examined soil physical properties and biological indicators on- farm and at experimental sites in the RRB. We expected to see increases in Kfs, aggregate 48 stability, and bulk density, as seen in other locations in the Midwest. Because most studies focus on changes in discharge and loads, we focus specifically on the relationship between drainage age and soil properties, to measure specific changes to properties within the soil ecosystem and provide a greater understanding of the effects and dynamics of long-term drainage on the soil ecosystem. 2.3. Materials and Methods Seven agricultural fields were sampled for this study (Table 2.1). Of the seven fields, three were drained before 2005 (referred to as >15), three had drainage installed after 2016 (referred to as <5), and one has never been drained (undrained). All fields are located within the RBB, and within Polk, Marshall, and Traverse counties (Figure 2.1). These field locations were chosen to minimize variability and maximize comparison between similar drainage status, location, and crops. Selection criteria also included silty clay loam soils as defined by NRCS’ SSURGO and soils were evaluated for soil texture (Table 2.1). Site Selection and Sampling We selected three sites that had subsurface drainage installed over 15 years ago (NWROC 1, NWROC 2, and Farm 1) and three sites that were drained within the last 5 years (Farm 2, Farm 3, and NWROC 3), as well as one undrained field (NWROC 4) as a benchmark. At each field, eight sampling locations were identified for analysis. At each sampling location, 7 cm diameter cores were sampled from 0-15 cm and 0-30 cm using an AMS Soil Core Sampler (AMS, USA). These cores were used for bulk density, 49 texture, PMC, WEOC, and WEON. Additionally, aggregate stability samples and Kfs were measured at all sampling locations in each field. At NWROC sites, four of these locations were located within 2 meters of subsurface drainage line, and four were located equidistant between drain lines. However, at on-farm sites, the exact locations of subsurface drains were unknown, and samples were distributed approximately one and a half drain spacings apart (i.e., 27 meters for fields with 18 meter spacing) to capture variation. While Farm 2 was strip-tilled, the other fields were conventionally tilled on similar schedules. Infiltrometers At each sampling location, a SATURO Infiltrometer (METER Group, Inc., USA) was used to calculate field saturated hydraulic conductivity (Kfs). Kfs describes the rate at which water can move through pore spaces or fractures during field saturation (Meter Group, 2017). The SATURO dual head infiltrometer is a single ring infiltrometer. To measure Kfs , the SATURO ponded water onto the soil surface inside an insertion ring pounded 5 cm deep in the soil. The SATURO then automatically pumped air and water to achieve specific pressures on the soil surface. During each 95-minute run, two cycles were run at two heads (0 cm for low pressure, 10 cm for high pressure). At each pressure, once steady-state conditions were achieved, the infiltration rates were measured for 10 minutes, allowing the machine to calculate Kfs. The simplified equation for this calculation used by the SATURO machine to calculate Kfs is derived from Reynolds and Elnick (1990): Kfs = Δ(i1-i2)/D1-D2 (1) 50 Where D1 is the high-pressure head, D2 is the lower pressure head, Δ = 0.993*insertion depth of the infiltrometer + 0.578*radius of the infiltrometer, i1 is the infiltration rate at the high-pressure head, and i2 is the infiltration rate at the lower pressure head. Machine error was common during field analysis and runs that demonstrated inadequate equilibration were discarded (between 6-8 runs were used in statistical analysis from each site). Aggregate Stability At each site where Kfs was measured, soil samples for aggregate stability analysis were collected between 0-15 cm within a 15 x 15 cm area using a shovel (8 samples per site). Only intact aggregates were saved from the sample (Norris et al., 2020). Samples were then air dried, lightly mixed, and selected to preserve aggregates and avoid utilizing loose, unaggregated soil. Another sample was air-dried for moisture correction. Samples were initially weighed, then placed on a stack of four sieves (2mm, 1mm, 250 μm, and 53 μm), and soaked by capillary for 10 minutes (Kemper and Rosenau, 1986). Samples were then shaken using a mechanical sieving machine through a stroke length of 4 cm, 30 times per minute for 10 minutes. All sample material was collected from each sieve and oven dried above 100°C until a constant weight was achieved. Each fraction was then weighed and corrected for moisture and coarse fragments (Aksakal et al., 2020). Bulk Density At each sampling location, 7 cm soil cores were collected using an AMS Soil Core Sampler (AMS, USA). Two depths, 0 - 15 and 15 - 30 cm, were extracted at each location. The moist soil weight was measured. Samples were then dried at 110°C for at 51 least 24 hours, and dry soil weight was recorded. Using these measurements, soil water content was calculated: (weight of moist soil - weight of dry soil) / weight of dry soil (2) Bulk density was also calculated for each sample: weight of dry soil / volume of soil core (3) Textural Analysis To determine soil texture, Laser Particle Size Analysis (PSA) was conducted on a subset of samples (4 sampling locations at each site, at both depths). PSA was chosen as the preferred method due to its relative accuracy with smaller particle sizes (Miller and Schaetzl, 2012). Approximately 0.5 g of soil (dried at 50°C, ground, and passed through a < 2 mm sieve) was placed in a 30 mL bottle. Then, 5 mL of 5% sodium hexametaphosphate (dispersing agent) and 5 mL of bleach (to remove organic matter) were added to the sample. Samples were then brought to volume (approximately 25 mL) with deionized water, shaken for 24 hours at 30 rpm. Shaken samples were then analyzed using a Malvern Mastersizer 3000 (Malvern Panalytical, USA), determining fractions of particle size and soil texture. Potentially Mineralizable Carbon Potentially mineralizable carbon (PMC) measures part of the microbially available soil C. This assay indicates both the ability of the microbial community and the amount and quality of substrate available to that community (Cates et al., 2019; Franzluebbers et al., 2000). The PMC assay measures CO2 flush following optimal conditions for microbial communities. 3.0 g of ground soils were rewetted to 50% water- filled pore space and sealed in a 1 L canning jar. They were then incubated for 24 hours 52 at 25°C. After 24 hours, the concentration of CO2 in each jar was measured using a LiCor LI-830 and corrected using blanks (1 L canning jar without soil or water, incubated under the same conditions). Three replicates were measured and averaged for each sampling depth. Water-Extractable Organic Carbon/Nitrogen Water-extractable organic C (WEOC) and N (WEON), measures of microbially- available substrate, were estimated on 16 subsamples (Haney et al., 2012; method adapted from Haney et al., 2012). 4.0 g of ground, dried soil and 40 mL of deionized water were placed into centrifuge tubes. Samples were shaken for 10 minutes on a reciprocal shaker, then centrifuged for 5 minutes at 3500 rpm, separating the liquid and solid. Then, the separated liquid was filtered through Whatman 2V filter paper, placed into glass vials, and diluted with deionized water for analysis (samples were 1:4 diluted with deionized water) to create a 19.9 mL sample; 0.1 mL of 2 M hydrochloric acid was added to each sample. Samples were then mixed and measured in a Shimadzu TOC-L Laboratory Total Organic Carbon Analyzer (Shimadzu, Japan) for organic C and total nitrogen (TN). Samples were also evaluated at the University of Minnesota Research Analytical Laboratory for inorganic forms of N (nitrates, nitrites, and ammonium). These values were then subtracted from TN to calculate WEON. Statistical Analysis All statistical analyses were completed using the R programming language. Following checks for normality, mixed models were analyzed using analysis of variance (ANOVA). Terms of the model included: treatment (fixed), depth, and crop (random; interaction term between treatment and crop also used). Depth was a fixed effect, but if it 53 was found to be significant, it was removed from the model and measurements were modeled by depth. Before analysis, all data was evaluated for normality, and transformed if necessary (WEON and WEOC were log-transformed). All statistical thresholds were 10% level of significance. Although included in visual analysis, the undrained field was not included in statistical analysis due to lack of replication. 2.4. Results Climate and Sampling Conditions The summer of 2021 was a particularly dry (and hot) year in NW Minnesota (Table 2.2). While the mean annual precipitation of Crookston is 56.59 cm (1991-2020, Midwest Regional Climate Center), in 2021, Crookston (all sampling sites located within approximately 150 kilometers of Crookston) only received 43.40 cm of rain, with only 20.93 cm between the months of April and September. During the sampling period in July, Crookston (and nearby counties, including all sites from this study) received less than 0.25 cm of precipitation. Dry climatic conditions resulted in extremely dry soil conditions, rendering soil samples difficult to obtain (it was not possible to collect soil cores from 15 - 30 cm at Farm 2, sampled last chronologically, where soil moisture content ranged between 2.4 - 7.5 %). The lack of moisture generated homogeneous conditions on all sampled fields and resulted in no drainage at all sites. The samples in this study therefore represent long- term effects of subsurface drainage, not the immediate effect of drainage on soil water status. 54 Saturated Hydraulic Conductivity In the >15 fields, saturated hydraulic conductivity (Kfs) rates were 7.58 cm hr-1 faster than <5 fields (statistically significant, p < 0.05; Figure 2.2). There was no significant difference between fields with standing wheat and those with standing corn. In fields with known tile locations, there were no significant differences in Kfs between above and between subsurface drains. Aggregate Stability Overall, the <5 and >15 fields had statistically similar percentages of stable macroaggregates (Figure 2.3). Broadly speaking, fields with older drainage had slightly more small aggregates (0.25 – 2mm) and less large aggregates (>2 mm). Bulk Density At both the 0 – 15 cm and 15 – 30 cm depth increments, there was no significant difference between bulk density of the two treatments (Figure 2.4). The 15 – 30 cm increment demonstrated trends of decreased bulk density in > 15 fields compared to <5 fields. Additionally, there were no significant differences between crops at both depths. Potentially Mineralizable Carbon In the 0 - 15 cm depth increment, there was no significant difference in PMC difference between the two drainage treatments (Figure 2.5a). Additionally, there was no significant difference between the two crops. However, at the 15 – 30 cm depth increment, >15 years had marginally greater PMC than <5 fields (p < 0.1). Additionally, PMC was significantly greater in standing corn than wheat (Figure 2.6a). 55 Water-Extractable Organic Carbon/Nitrogen At the 0 - 15 cm depth increment, there is no significant difference between the two drainage treatments for WEOC (Figure 2.5b). Additionally, there is no difference between crops. However, at the 15 - 30 cm depth increment, >15 treatment had marginally greater WEOC than <5 (p < 0.1). Additionally, WEOC values in corn are significantly greater than in wheat (Figure 2.6b). WEON values were significantly greater in the >15 treatment than the <5 treatment at both the 0 - 15 and 15 - 30 cm depth increments (both p < 0.01; Figure 2.5c). At the 0 - 15 cm depth increment, there is no significant difference between wheat and corn, however, at the 15 - 30 cm depth increment, WEON values were greater in corn than wheat fields (Figure 2.6c). Summaries for all soil physical properties and biological indicators can be found in Table 2.3 2.5. Discussion While other studies have evaluated the impact of drainage age on discharge and nutrient loads, few studies have examined the soil properties mediating water quality outcomes (Kladivko and Bowling, 2021; Lundekvam, 1998; Randall and Mulla, 2001; Thorp et al., 2008). In this study, we observed associations between drainage age and selected soil physical and biological properties. Generally, fields with longer histories of drainage had greater Kfs, PMC, WEOC, and WEON. Further study of these metrics could provide more clarity on how drainage systems impact soil properties and ecosystems over time. It is important to note that this study occurred during drought conditions in NW Minnesota. This extremely low precipitation likely caused unique soil conditions. 56 Therefore, the results from this study should be viewed with this exceptional climatic pattern. Soil Physical Properties Although preferential flow pathways exist in soil profiles with or without subsurface drainage, the installation of drainage can increase the connectivity of macropores that enable preferential flow and have a significant effect on the rate of water movement and infiltration through soils (King et al., 2015; Pluer et al., 2020; Stamm et al., 1998; Stone and Wilson, 2006). Additionally, large abundances of earthworms are known to congregate near artificial drains in clay soils due to more frequent oxygenation rates (Jarvis, 2007; Nuutinen et al., 2001). The faster Kfs rates seen on the >15 fields compared to the <5 fields may point to development of prominent preferential flow paths to subsurface drains that occur over many years (decades) of subsurface drainage. While all fields were sampled at similar moisture levels (5-10% volumetric water content), it is likely that the smectite clay mineralogy seen in most of the clay-rich soils sampled in this study leads to shrink-swell condition and increased preferential flow paths (Beauchemin et al., 1998; Kladivko et al., 2001). Hydraulic conductivity rates have been demonstrated to vary with soil water content, with its highest values occurring when the soil has low antecedent water content (Jabro, 1996; Messing and Jarvis, 1990). It is possible that the Kfs values in this study are greater across all sites due to the drought conditions, but this was likely not the source of variation between treatments, as soil moisture was similar across most fields (the undrained field was the highest at 11% moisture, which is still quite low). Perhaps there were other factors occurring at depth that were not observed as a part of this study. 57 To our knowledge, no studies have specifically examined the relationship between distance to drainage line and hydraulic conductivity. We expected statistical differences in Kfs between measurements taken above subsurface drainage lines and those equidistant between lines, since studies in both Minnesota and Indiana have shown that smaller drain spacing leads to increased discharge and infiltration (Jin and Sands, 2003; Kladivko et al., 2004). Additionally, we expected hydraulic conductivity rates to be potentially impacted by historic crop (or standing crop) due to decomposed root material (Govaerts et al., 2007; Mitchell et al., 1995). Instead, it appears that long-term drainage treatment was the factor with the greatest influence on Kfs in this study. Because higher moisture regimes generally have less stable macroaggregates than lower moisture regimes, as well as microaggregates to a lesser extent (Cates et al., 2019; Lehrsch et al., 1991), we expected a relationship between drainage treatment and macroaggregate stability. Instead, it is likely that tillage is the main management factor impacting aggregate distribution in this study (Hajabbasi and Hemmat, 2000; Kasper et al., 2009; Tisdall and Oades, 1982). At Farm 2 (<5 drained), the only field with reduced tillage, 86.56% of all aggregates were > 2 mm, compared to the study average of 59.67%. The impact of reduced tillage on preserving macroaggregates is well studied (Six et al., 2000; Mikha and Rice, 2004; Kasper et al., 2009). For microaggregates, these results appear to be similar: Farm 2 has a much smaller percentage of microaggregates than the study average (6.46% compared to 11.75%). The differences are perhaps less extreme for microaggregates because they are more persistent through disturbance from tillage (Kasper et al., 2009; Six et al, 2000; Du et al., 2013). It is likely that this connection between aggregate size distribution and tillage is the major factor influencing these 58 results. While wet-dry cycles’ impact on aggregate stability has been studied (Cosentino et al., 2006; Lehmann et al., 2019), perhaps more work on the persistent hydraulic regime should be conducted to further provide a better insight to how moisture levels in the soil are altering aggregate size distribution. Due to the number of studies citing decreased bulk density with subsurface drainage installation (i.e., Baker et al., 2004, Jia et al., 2008), we expected bulk density to decrease in fields with longer-term subsurface drainage, as longer-term drainage would have a stronger effect. Additionally, bulk density is often a predictor of Kfs and used to approximate Kfs in models (Jabro, 1992; Rawls et al., 1997), and the increased Kfs would be a result of decreased bulk density. Our study found non-significant (also seen in Chow et al., 1993, Baker et al, 2004) but lower bulk density with increased drainage age, particularly at the 15 - 30 cm depth increment (Table 2.3). Hundal et al. (1976) found similar results when comparing drained and undrained soils, with statistically greater saturated hydraulic conductivity in drained soils, and smaller (but not statistically significant) bulk density in drained soils. It is possible that drainage decreases bulk density, but that effect may increase over time. The results from this study indicate that while functions (Kfs) related to soil physical properties changed due to drainage age, soil properties changed to a lesser extent (bulk density) or were controlled by other factors (aggregate stability). This may be at least partially explained by the dry condition, or perhaps we were not able to capture the variation in properties completely. 59 Biological Indicators Although soil physical properties in subsurface drained systems have been studied at length, little research has been conducted to understand the interaction of subsurface drainage with biological indicators. Some research has been conducted on soil moisture levels and mineralizable carbon, finding higher moisture content results in higher measurable mineralizable carbon (Das et al., 2019; Jauhiainen et al. 2016; Libohova et al., 2014). For these biological indicators, it appears that there are two factors on different timescales occurring: seasonally, standing crop does influence the biological communities, as seen by the higher PMC, WEOC, and WEON values in standing corn versus standing wheat, particularly at depth. On the decadal scale, it appears that drainage systems can increase these biological factors as well. The late July sampling, particularly with dry conditions, occurred during varying root activity in corn versus wheat cropping systems. At the time of sampling, the standing wheat was nearing physiological maturity, and therefore its photosynthetic rate was in decline (Del Blanco et al., 2000; Shah and Paulsen, 2003). In comparison, standing corn was at approximately the V10-V12 growth stages at sampling, when photosynthesis, exudate production, and microbial activity were likely significantly higher (Rochette and Flanagan, 1997), boosting levels of higher biological indicators. On a longer timescale, higher biological indicators in fields with older drainage may be due to biological activity along preferential flow paths (Franklin et al., 2021). Preferential flow paths have been identified as “hot spots” for biological activity, with larger microbial biomass, although these “hot spots” have proved difficult to quantify 60 (Baveye and Laba, 2015; Bundt et al., 2001a). Preferential flow paths include reduced anoxic conditions (Anthony et al., 2020), wet-dry cycles, and superior nutrient and substrate supply that can facilitate higher pockets of microbial activity compared to other parts of the soil profile (Bundt et al., 2001b; Chabbi et al., 2009; Franklin et al., 2021; Fuhrmann et al., 2019). A secondary explanation for these increases in biological activity may be the antecedent moisture conditions. It was long believed that C-based biological indicators had a linear relationship with moisture and were dependent only on current moisture content (Patel et al., 2021b). However, recent studies have demonstrated that antecedent moisture can impact microbial communities and activity (Hawkes et al., 2020; Patel et al., 2021a. Patel et al., 2021b; Smith et al., 2017). Soil may experience variable distribution of water based on their antecedent moisture conditions, therefore different sources of C may be activated based on antecedent moisture (Patel et al., 2021a). Soils with drier conditions have high levels of WEOC and C mineralization when re-wet, since they mobilize more C and provide greater access to substrates (Patel et al., 2021a, Patel et al., 2021b). In our high pH soils, where redox is not a main driver in microbial activity, it is possible that the longer-term wet-dry cycles and greater preferential flow after decades of drainage generated drier antecedent moisture conditions, leading to greater active C pools measured in this study. Another explanation for the larger biological metrics may be higher crop productivity due to drainage. While subsurface drainage may decrease carbon stocks in these soils over time (Fernández et al., 2017), carbon flows into and from the system are likely greater, as demonstrated by increases in yield over time (Janzen, 2006; Kladivko 61 and Bowling, 2021; Wood et al., 2016). Deeper sampling would provide more insight into this possibility, as carbon stocks would likely decrease deeper in the profile due to mineralization and aeration but remain elevated in the upper portion due to increased inputs (Mitchell, 2020). Management Implications Growers install drainage to increase productivity and field workability. The soil metrics measured here underlie the critical hydrological functions of drainage. Foremost, growers should expect that as their drainage systems age, water transport through the soil profile will be faster (Kladivko et al., 2021). In clay-rich soils, this will likely cause a higher percentage of nitrate and phosphorus transport to subsurface drains (Grant et al., 2019; Kelly and Pomes, 1998) and this potential for increased nutrient loads should be considered for on-farm nutrient management. Our results indicate that drainage systems may increase biological activity over time, a potential benefit alongside the well-known agronomic benefits of drainage, which could contribute to increased N use efficiency (Castellano et al., 2019; Paiao et al., 2021). 2.6. Conclusions In sum, we found soil properties in fields with subsurface drainage likely evolve over time. Significantly greater Kfs in longer drained fields suggests that preferential flow is a driver of this evolution. Interestingly, other soil physical properties changed with drainage age to a lesser extent, as tillage or other factors likely overpowered the effect of changing hydrology. But the changes in moisture and oxygen contents had an apparent effect on microbial activity, increasing active N and C, especially below 15 cm. 62 While many studies compare drained and undrained plots, our study clarifies that drainage is not a static treatment, so drainage status could represent a spectrum of conditions than a static dichotomy. Subsurface drainage is an anthropogenic force that can dynamically change soil properties, particularly as those drainage systems age. At minimum, studies should contextualize their drainage systems by considering their age. We anticipate that similar research on different soils under different crop management and climates will yield varying assessments of drainage effects on soil properties over time. While most long-term studies focus on water quality, more research should pair this idea with better understanding changing physical and biological properties, as these changes in water quality over time are a direct result of changing soil properties. 63 Table 2.1: Site descriptions for sampling locations. All sampling locations used conventional tillage and management practices unless noted. Field Year Drained Drain Spacing % Sand % Clay 2022 Crop Notes NWROC 1 2000 18 meters 33.20 30.04 Wheat NWROC 2 2000 18 meters 22.92 40.32 Corn NWROC 3 2019 15 meters 21.66 44.73 Wheat NWROC 4 Undrained N/A 13.71 50.97 Wheat Farm 1 2001 15 meters 21.92 40.71 Corn Organic Farm 2 2016 15 meters 39.83 18.90 Wheat Strip-Tillage Farm 3 2018 15 meters 31.13 30.69 Corn Manure Table 2.2: Precipitation and temperature for the past 30 years as well as in 2021 at each sampling location. Field 1991-2020 Average Daily Temp. (°C) 1991-2020 Average Annual Precip. (cm) 2021 Average Daily Temp. (°C) 2021 Precip. (cm) NWROC 1-4, Farm 1 4.43 56.59 6.11 43.40 Farm 2 4.39 58.67 5.72 43.03 Farm 3 6.71 64.59 8.89 69.46 64 Table 2.3: Summaries of soil physical properties and soil biological indicators. A star (*) indicates a significant difference (p < 0.1). Variable >15 Fields <5 Fields Kfs (cm hr-1) 19.23 ± 9.16 * 11.65 ± 9.71 Aggregate Fraction (2mm, %) 61.16 ± 20.78 65.38 ± 22.78 Aggregate Fraction (1 - 2 mm, %) 11.64 ± 7.78 6.79 ± 4.74 Aggregate Fraction (0.25 - 1 mm, %) 16.78 ± 9.44 15.08 ± 11.79 Bulk Density (g cm-3) 0 - 15 cm 1.10 ± 0.18 1.12 ± 0.28 Bulk Density (g cm-3) 15 - 30 cm 1.19 ± 0.17 1.28 ± 0.21 PMC (μg CO2-C g soil-1) 0 - 15 cm 90.08 ± 20.09 88.52 ± 34.97 PMC (μg CO2-C g soil-1) 15 - 30 cm 48.75 ± 14.26 * 39.82 ± 10.75 WEOC (mg kg soil-1) 0 - 15 cm 220.51 ± 46.17 200.68 ± 42.98 WEOC (mg kg soil-1) 15 - 30 cm 155.10 ± 31.08 * 130.91 ± 41.62 WEON (mg kg soil-1) 0 - 15 cm 16.73 ± 9.66 * 10.75 ± 4.26 WEON (mg kg soil-1) 15 - 30 cm 6.56 ± 2.85 * 3.67 ± 3.22 65 Figure 2.1: Site locations in the Red River of the North Basin (RRB). Approximate location of the RRB in Minnesota, North Dakota, and South Dakota highlighted. 66 Figure 2.2: Box plots of saturated hydraulic conductivity (Kfs) values of the three treatments. The >15 treatment has significantly higher (p < 0.1) Kfs than the <5 treatment. Figure 2.3: Percent water-stable aggregates by size class for the different treatments. 67 Figure 2.4: Bulk density of three treatments, split by depth. There is no significant difference between drainage treatments at either depth. Figure 2.5: A) Potentially mineralizable carbon (PMC), B) water-extractable organic carbon (WEOC), and water-extractable organic carbon (WEON) separated by depth. Starred (*) depths indicate significant difference between >15 years and <5 years (p < 0.1). 68 Figure 2.6: A) Potentially mineralizable carbon (PMC), B) water-extractable organic carbon (WEOC), and water-extractable organic carbon (WEON) separated by crop. Starred (*) depths indicate significant difference (p < 0.1). The undrained treatment has been removed from visual analysis to not skew data 69 REFERENCES Abid, M., and R. Lal. 2009. Tillage and drainage impact on soil quality: II. Tensile strength of aggregates, moisture retention and water infiltration. Soil Tillage Res. 103(2): 364–372. doi: 10.1016/j.still.2008.11.004. Ahmed, I., R. Rudra, K. McKague, B. Gharabaghi, and J. Ogilvie. 2007. Evaluation of the Root Zone Water Quality Model (RZWQM) for Southern Ontario: Part II. Simulating Long-Term Effects of Nitrogen Management Practices on Crop Yield and Subsurface Drainage Water Quality. Water Qual. Res. J. 42(3): 219–230. doi: 10.2166/wqrj.2007.025. Aksakal, E.L., I. Angin, and S. Sari. 2020. A new approach for calculating aggregate stability: Mean weight aggregate stability (MWAS). CATENA 194: 104708. doi: 10.1016/j.catena.2020.104708. Alakukku L., and E. Turtola. 2010. Surface Runoff and Soil Physical Properties as Affected by Subsurface Drainage Improvement of a Heavy Clay Soil. Am. Soc. Agric. Biol. Eng. Amézketa, E. 1999. Soil Aggregate Stability: A Review. J. Sustain. Agric. 14(2–3): 83– 151. doi: 10.1300/J064v14n02_08. Anthony, M.A., T.W. Crowther, D.S. Maynard, J. van den Hoogen, and C. Averill. 2020. Distinct Assembly Processes and Microbial Communities Constrain Soil Organic Carbon Formation. One Earth 2(4): 349–360. doi: 10.1016/j.oneear.2020.03.006. Averill, C., B. Waring, and C. Hawkes. 2016. Historical precipitation predictably alters the shape and magnitude of microbial functional response to soil moisture. Global change biology 22: 1957-1964. doi: 10.1111/gcb.13219. Baeumler, N.W., and S.C. Gupta. 2020. Precipitation as the Primary Driver of Variability in River Nitrogen Loads in the Midwest United States. JAWRA Journal of the American Water Resources Association 56(1): 113–133. doi: 10.1111/1752- 1688.12809. Baker, B.J., N.R. Fausey, and K.R. Islam. 2004. Comparison of Soil Physical Properties under Two Different Water Table Management Regimes. Soil Sci. Soc. Am. J. 68(6): 1973–1981. doi: 10.2136/sssaj2004.1973. Baker, J.L., and H.P. Johnson. 1981. Nitrate-Nitrogen in Tile Drainage as Affected by Fertilization. Journal of Environmental Quality 10(4): 519–522. doi: 10.2134/jeq1981.00472425001000040020x. Baker, J.L., K.L. Campbell, H.P. Johnson, and J.J. Hanway. 1975. Nitrate, Phosphorus, and Sulfate in Subsurface Drainage Water. Journal of Environmental Quality 4(3): 406–412. doi: 10.2134/jeq1975.00472425000400030027x. 70 Bandibas, J., A. Vermoesen, C. De Groot, and O. Van Cleemput. 1994. The effect of different moisture regimes and soil characteristics on nitrous oxide emission and consumption by different soils. Soil Science 158(2): 106–114. Banerjee, S., B. Helgason, L. Wang, T. Winsley, B.C. Ferrari, et al. 2016. Legacy effects of soil moisture on microbial community structure and N2O emissions. Soil Biology and Biochemistry 95: 40–50. doi: 10.1016/j.soilbio.2015.12.004. Barakat, M., B. Cheviron, and R. Angulo-Jaramillo. 2016. Influence of the irrigation technique and strategies on the nitrogen cycle and budget: A review. Agricultural Water Management 178: 225–238. doi: 10.1016/j.agwat.2016.09.027. Barry, D.A.J., D. Gooragoo, and M.J., Goss. 1993. Estimation of Nitrate Concentrations in Groundwater Using a Whole Farm Nitrogen Budget. J. Environ. Qual. 22: 767- 775. Baveye, P.C., and M. Laba. 2015. Moving away from the geostatistical lamppost: Why, where, and how does the spatial heterogeneity of soils matter? Ecol. Model. 298: 24–38. doi: 10.1016/j.ecolmodel.2014.03.018. Beauchamp, E.G., C. Gale, and J.C. Yeomans. 1980. Organic matter availability for denitrification in soils of different textures and drainage classes. Commun. Soil Sci. Plant Anal. 11(12): 1221–1233. doi: 10.1080/00103628009367119. Beauchemin, S., R.R. Simard, and D. Cluis. 1998. Forms and Concentration of Phosphorus in Drainage Water of Twenty-Seven Tile-Drained Soils. J. Environ. Qual. 27(3): 721–728. doi: 10.2134/jeq1998.00472425002700030033x. Benoy, G.A., R.W. Jenkinson, D.M. Robertson and D.A. Saad. 2016. Nutrient delivery to Lake Winnipeg from the Red—Assiniboine River Basin – A binational application of the SPARROW model. 41(3) 429-447, DOI: 10.1080/07011784.2016.1178601 Blann, K.L., J.L. Anderson, G.R. Sands, and B. Vondracek. 2009. Effects of Agricultural Drainage on Aquatic Ecosystems: A Review. Critical Reviews in Environmental Science and Technology 39(11): 909–1001. doi: 10.1080/10643380801977966. Borchers, J.G., and D.A. Perry. 1992. The influence of soil texture and aggregation on carbon and nitrogen dynamics in southwest Oregon forests and clearcuts. Canadian Journal of Forest Research 22(3): 298–305. Bouma, J. 1981. Soil morphology and preferential flow along macropores. Agric. Water Manag. 3(4): 235–250. Bouma, J., L.W. Dekker, and J.C.F.M. Haans. 1979. Drainability of some Dutch clay soils: A case study of soil survey interpretation. Geoderma 22(3): 193–203. doi: 10.1016/0016-7061(79)90018-1. 71 Bourne, R., N. Armstrong, G. Jones, R. Bourne, N. Armstrong. 2002. A Preliminary Estimate of Total Nitrogen and Total Phosphorus Loading to Streams in Manitoba, Canada. Manitoba Conservation Report 2002-2004. Bregliani, M.M., G.H. Ros, E.J.M. Temminghoff, and W.H. van. Nitrogen Mineralization in Soils Related to Initial Extractable Organic Nitrogen: Effect of Temperature. 41(11), 1383-1398, doi: https://doi- org.ezp2.lib.umn.edu/10.1080/00103621003759387 Bundt, M., F. Widmer, M. Pesaro, J. Zeyer, and P. Blaser. 2001a. Preferential Flow paths: biological `hot spots’ in soils. Soil Biol. 33(6) 729-738. Doi: https://doi.org/10.1016/S0038-0717(00)00218-2 Bundt, M., M. Jäggi, P. Blaser, R. Siegwolf, and F. Hagedorn. 2001b. Carbon and Nitrogen Dynamics in Preferential Flow Paths and Matrix of a Forest Soil. Soil Science Society of America Journal 65(5): 1529–1538. doi: 10.2136/sssaj2001.6551529x. Cassman, K.G., and D.N. Munns. 1980. Nitrogen Mineralization as Affected by Soil Moisture, Temperature, and Depth. Soil Science Society of America Journal 44(6): 1233–1237. doi: 10.2136/sssaj1980.03615995004400060020x. Castellano, M.J., S.V. Archontoulis, M.J. Helmers, H.J. Poffenbarger, and J. Six. 2019. Sustainable intensification of agricultural drainage. Nat Sustain 2(10): 914–921. doi: 10.1038/s41893-019-0393-0. Cates, A.M., M.D. Ruark, A.S. Grandy, and R.D. Jackson. 2019. Small soil C cycle responses to three years of cover crops in maize cropping systems. Agric. Ecosyst. Environ. 286: 106649. doi: 10.1016/j.agee.2019.106649. Chabbi, A., I. Kögel-Knabner, and C. Rumpel. 2009. Stabilised carbon in subsoil horizons is located in spatially distinct parts of the soil profile. Soil Biol. Biochem. 41(2): 256–261. doi: 10.1016/j.soilbio.2008.10.033. Cho, S.J., and B.A. McCarl. 2017. Climate change influences on crop mix shifts in the United States. Sci Rep 7(1): 40845. doi: 10.1038/srep40845. Chow, T., H. Rees, K. Webb, and D. Langille. 1993. Modification of subsoil characteristics resulting from drainage tile installation. Soil Sci. 156(5): 346–357. Ciarlo, E., M. Conti, N. Bartoloni, and G. Rubio. 2007. The effect of moisture on nitrous oxide emissions from soil and the N2O/(N2O+N2) ratio under laboratory conditions. Biol Fertil Soils 43(6): 675–681. doi: 10.1007/s00374-006-0147-9. Clark, J.D., K.S. Veum, F.G. Fernández, J.J. Camberato, P.R. Carter, et al. 2019. United States Midwest Soil and Weather Conditions Influence Anaerobic Potentially Mineralizable Nitrogen. Soil Sci. Soc. Am. j. 83(4): 1137–1147. doi: 72 10.2136/sssaj2019.02.0047. Colbourn, P., and R.J. Dowdell. Denitrification in field soils. 1984. Plant and Soil 76: 213-226. Collier, S.M., M.D. Ruark, L.G. Oates, W.E. Jokela, and C.J. Dell. 2014. Measurement of Greenhouse Gas Flux from Agricultural Soils Using Static Chambers. J Vis Exp (90): 52110. doi: 10.3791/52110. Cormier, S.N., J.L. Musetta-Lambert, K.J. Painter, A.G. Yates, R.B. Brua, et al. 2021. Sources of nitrogen to stream food webs in tributaries of the Red River Valley, Manitoba. Journal of Great Lakes Research 47(3): 751–760. doi: 10.1016/j.jglr.2020.08.007. Cosentino, D., C. Chenu, and Y. Le Bissonnais. 2006. Aggregate stability and microbial community dynamics under drying–wetting cycles in a silt loam soil. Soil Biol. Biochem. 38(8): 2053–2062. doi: 10.1016/j.soilbio.2005.12.022. Crookston, B.S., M.A. Yost, M. Bowman, K. Veum, G. Cardon, et al. 2021. Soil health spatial-temporal variation influence soil security on Midwestern, U.S. farms. Soil Security 3: 100005. doi: 10.1016/j.soisec.2021.100005. Das, S., B.K. Richards, K.L. Hanley, L. Krounbi, M.F. Walter, et al. 2019. Lower mineralizability of soil carbon with higher legacy soil moisture. Soil Biol. Biochem. 130: 94–104. doi: 10.1016/j.soilbio.2018.12.006. Davis, C.A., A.S. Ward, A.J. Burgin, T.D. Loecke, D.A. Riveros-Iregui, et al. 2014. Antecedent Moisture Controls on Stream Nitrate Flux in an Agricultural Watershed. Journal of Environmental Quality 43(4): 1494–1503. doi: 10.2134/jeq2013.11.0438. Del Blanco, I., S. Rajaram, W. Kronstad, and M. Reynolds. 2000. Physiological performance of synthetic hexaploid wheat‐derived populations. Crop Sci. 40(5): 1257–1263. Dobbie, K.E., and K.A. Smith. 2003. Nitrous oxide emission factors for agricultural soils in Great Britain: the impact of soil water-filled pore space and other controlling variables. Global Change Biology 9(2): 204–218. doi: 10.1046/j.1365- 2486.2003.00563.x. Dobbie, K.E., and K.A. Smith. 2006. The effect of water table depth on emissions of N2O from a grassland soil. Soil Use and Management 22(1): 22–28. doi: 10.1111/j.1475-2743.2006.00002.x. Drury, C.F., D.J. McKenney, W.I. Findlay, and J.D. Gaynor. 1993. Influence of Tillage on Nitrate Loss in Surface Runoff and Tile Drainage. Soil Science Society of America Journal 57(3): 797–802. doi: 10.2136/sssaj1993.03615995005700030028x. 73 Du, Z., T. Ren, C. Hu, Q. Zhang, and H. Blanco-Canqui. 2013. Soil Aggregate Stability and Aggregate-Associated Carbon Under Different Tillage Systems in the North China Plain. J. Integr. Agric. 12(11): 2114–2123. doi: 10.1016/S2095- 3119(13)60428-1. Eickenscheidt, N., and R. Brumme. 2013. Regulation of N2O and NOx emission patterns in six acid temperate beech forest soils by soil gas diffusivity, N turnover, and atmospheric NOx concentrations. Plant Soil 369(1–2): 515–529. doi: 10.1007/s11104-013-1602-7. Eickenscheidt, T., J. Heinichen, J. Augustin, A. Freibauer, and M. Drösler. 2014. Nitrogen mineralization and gaseous nitrogen losses from waterlogged and drained organic soils in a black alder Alnus glutinosa (L.) Gaertn.) forest. Biogeosciences 11(11): 2961–2976. doi: 10.5194/bg-11-2961-2014. Evans, C.A., E.K. Miller, and A.J. Friedland. 1998. Nitrogen mineralization associated with birch and fir under different soil moisture regimes. Can. J. For. Res. 28(9) 1890-1898. Evans, R.O., and N.R. Fausey. 2015. Effects of Inadequate Drainage on Crop Growth and Yield. In: Skaggs, R.W. and van Schilfgaarde, J., editors, Agronomy Monographs. American Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, WI, USA. p. 13–54 Evans, S.E., and M.D. Wallenstein. 2012. Soil microbial community response to drying and rewetting stress: does historical precipitation regime matter? Biogeochemistry 109(1–3): 101–116. doi: 10.1007/s10533-011-9638-3. Fernández, F.G., E.D. Nafziger, S.A. Ebelhar, and R.G. Hoeft. 2009. Managing nitrogen. Illinois agronomy handbook, 24th ed. Univ. of Illinois, Urbana-Champaign, IL: 113–132. Fernández, F.G., K.P. Fabrizzi, and S.L. Naeve. 2017. Corn and soybean’s season-long in-situ nitrogen mineralization in drained and undrained soils. Nutr. Cycl. Agroecosystems 107(1): 33–47. doi: 10.1007/s10705-016-9810-1. Fernández, F.G., R.E. Terry, and E.G. Coronel. 2015. Nitrous Oxide Emissions from Anhydrous Ammonia, Urea, and Polymer-Coated Urea in Illinois Cornfields. Journal of Environmental Quality 44(2): 415–422. doi: 10.2134/jeq2013.12.0496. Fernández, F.G., R.T. Venterea, and K.P. Fabrizzi. 2016. Corn Nitrogen Management Influences Nitrous Oxide Emissions in Drained and Undrained Soils. Journal of Environmental Quality 45(6): 1847–1855. doi: 10.2134/jeq2016.06.0237. Franklin, S.M., A.N. Kravchenko, R. Vargas, B. Vasilas, J.J. Fuhrmann, et al. 2021. The unexplored role of preferential flow in soil carbon dynamics. Soil Biol. Biochem. 161: 108398. doi: 10.1016/j.soilbio.2021.108398. 74 Franzluebbers, A.J., R.L. Haney, C.W. Honeycutt, H.H. Schomberg, and F.M. Hons. 2000. Flush of Carbon Dioxide Following Rewetting of Dried Soil Relates to Active Organic Pools. Soil Sci. Soc. Am. J. 64(2): 613–623. doi: 10.2136/sssaj2000.642613x. Fausey, N.R. 2015. Drainage, Surface and Subsurface. Enc. Soil. Env. 409-413. doi: 10.1016/b0-12-348530-4/00352-0 Fu-sheng, C., Z. De-hui, S. Anand Narain, and C. Guang-sheng. 2005. Effects of soil moisture and soil depth on nitrogen mineralization process under Mongolian pine plantations in Zhanggutai sandy land, P. R. China. Journal of Forestry Research 16(2): 101–104. doi: 10.1007/BF02857899. Fuhrmann, I., S. Maarastawi, J. Neumann, W. Amelung, K. Frindte, et al. 2019. Preferential flow pathways in paddy rice soils as hot spots for nutrient cycling. Geoderma 337: 594–606. doi: 10.1016/j.geoderma.2018.10.011. Gambrell, R.P., J.W. Gilliam, and S.B. Weed. 1975. Denitrification in Subsoils of the North Carolina Coastal Plain as Affected by Soil Drainage. Journal of Environmental Quality 4(3): 311–316. doi: 10.2134/jeq1975.00472425000400030005x. Gast, R.G., W.W. Nelson, and G.W. Randall. 1978. Nitrate Accumulation in Soils and Loss in Tile Drainage Following Nitrogen Applications to Continuous Corn. Journal of Environmental Quality 7(2): 258–261. doi: 10.2134/jeq1978.00472425000700020021x. Gentry, L.E., M.B. David, K.M. Smith, and D.A. Kovacic. 1998. Nitrogen cycling and tile drainage nitrate loss in a corn and soybean watershed. Ag. Eco. Env. 68: 85-97. doi: https://doi.org/10.1016/S0167-8809(97)00139-4 Gilliam, J.W., R.W. Skaggs, and S.B. Weed. 1979. Drainage Control to Diminish Nitrate Loss from Agricultural Fields. Journal of Environmental Quality 8(1): 137–142. doi: 10.2134/jeq1979.00472425000800010030x. Gobler, C.J., J.M. Burkholder, T.W. Davis, M.J. Harke, T. Johengen, et al. 2016. The dual role of nitrogen supply in controlling the growth and toxicity of cyanobacterial blooms. Harmful Algae 54: 87–97. doi: 10.1016/j.hal.2016.01.010. Gonçalves, J.L.M., and J.C. Carlyle. 1994. Modelling the influence of moisture and temperature on net nitrogen mineralization in a forested sandy soil. Soil Biology and Biochemistry 26(11): 1557–1564. doi: 10.1016/0038-0717(94)90098-1. Govaerts, B., M. Fuentes, M. Mezzalama, J.M. Nicol, J. Deckers, et al. 2007. Infiltration, soil moisture, root rot and nematode populations after 12 years of different tillage, residue and crop rotation managements. Soil Tillage Res. 94(1): 209–219. doi: 10.1016/j.still.2006.07.013. 75 Gramlich, A., S. Stoll, C. Stamm, T. Walter, and V. Prasuhn. 2018. Effects of artificial land drainage on hydrology, nutrient and pesticide fluxes from agricultural fields – A review. Agriculture, Ecosystems & Environment 266: 84–99. doi: 10.1016/j.agee.2018.04.005. Grant, K.N., M.L. Macrae, and G.A. Ali. 2019. Differences in preferential flow with antecedent moisture conditions and soil texture: Implications for subsurface P transport. Hydrol. Process. 33(15): 2068–2079. doi: 10.1002/hyp.13454. Grebliunas, B.D., S.D. Armstrong, and W.L. Perry. 2016. Changes in Water-Extractable Organic Carbon with Cover Crop Planting under Continuous Corn Silage Production. Air Soil Water Res. 9: ASWR.S30708. doi: 10.4137/ASWR.S30708. Groffman, P.M., and J.M. Tiedje. 1988. Denitrification Hysteresis During Wetting and Drying Cycles in Soil. Soil Science Society of America Journal 52(6): 1626–1629. doi: 10.2136/sssaj1988.03615995005200060022x. Guntiñas, M.E., M.C. Leirós, C. Trasar-Cepeda, and F. Gil-Sotres. 2012. Effects of moisture and temperature on net soil nitrogen mineralization: A laboratory study. Eur. Jour. Soil. Bio. 48: 73-80. doi: https://doi.org/10.1016/j.ejsobi.2011.07.015 Guo, X., C.F. Drury, X. Yang, W. Daniel Reynolds, and R. Fan. 2014. The Extent of Soil Drying and Rewetting Affects Nitrous Oxide Emissions, Denitrification, and Nitrogen Mineralization. Soil Science Society of America Journal 78(1): 194–204. doi: 10.2136/sssaj2013.06.0219. Hajabbasi, M.A., and A. Hemmat. 2000. Tillage impacts on aggregate stability and crop productivity in a clay-loam soil in central Iran. Soil Tillage Res. 56(3–4): 205–212. doi: 10.1016/S0167-1987(00)00140-9. Haney, R.L., A.J., Franzluebbers, V.L., Jin, M.V., Johnson, E.B., Haney. M.J., White, R.D., Harmel. 2012. Soil Organic C:N vs. Water-Extractable Organic C:N. OJSS 02(03): 269–274. doi: 10.4236/ojss.2012.23032. Hanrahan, B.R., K.W. King, M.L. Macrae, M.R. Williams, and J.H. Stinner. 2020. Among-site variability in environmental and management characteristics: Effect on nutrient loss in agricultural tile drainage. Journal of Great Lakes Research 46(3): 486–499. doi: 10.1016/j.jglr.2020.02.004. Hawkes, C.V., M. Shinada, and S.N. Kivlin. 2020. Historical climate legacies on soil respiration persist despite extreme changes in rainfall. Soil Biol. Biochem. 143: 107752. doi: 10.1016/j.soilbio.2020.107752. Helmers, M.J., L. Abendroth, B. Reinhart, G. Chighladze, L. Pease, et al. 2022. Impact of controlled drainage on subsurface drain flow and nitrate load: A synthesis of studies across the U.S. Midwest and Southeast. Ag. Wat. Mang. 259: 107265. doi: 10.1016/j.agwat.2021.107265. 76 Hundal, S.S., G.O. Schwab, and G.S. Taylor. 1976. Drainage System Effects on Physical Properties of a Lakebed Clay Soil. Soil Sci. Soc. Am. J. 40(2): 300–305. doi: 10.2136/sssaj1976.03615995004000020028x. Jabro, J. 1992. Estimation of Saturated Hydraulic Conductivity of Soils From Particle Size Distribution and Bulk Density Data. Trans. ASAE 35: 557–560. doi: 10.13031/2013.28633. Jabro, J. 1996. Variability of Field-saturated Hydraulic Conductivity in a Hagerstown Soil as Affected by Initial Water Content. Soil Sci. 161(11): 735–739. Jacinthe, P.A., R. Lal, and J.M. Kimble. 2001. Organic Carbon Storage and Dynamics in Croplands and Terrestrial Deposits as Influenced by Subsurface Tile Drainage: Soil Sci. 166(5): 322–335. doi: 10.1097/00010694-200105000-00003. Jacobs, T.C., and J.W. Gilliam. 1985. Riparian Losses of Nitrate from Agricultural Drainage Waters. Journal of Environmental Quality 14(4): 472–478. doi: 10.2134/jeq1985.00472425001400040004x. Janzen, H.H. 2006. The soil carbon dilemma: Shall we hoard it or use it? Soil Biol. Biochem. 38(3): 419–424. doi: 10.1016/j.soilbio.2005.10.008. Jarvis, N.J. 2007. A review of non-equilibrium water flow and solute transport in soil macropores: principles, controlling factors and consequences for water quality. Eur. J. Soil Sci. 58(3): 523–546. doi: 10.1111/j.1365-2389.2007.00915.x. Jauhiainen, J., H. Silvennoinen, M. Könönen, S. Limin, and H. Vasander. 2016. Management driven changes in carbon mineralization dynamics of tropical peat. Biogeochemistry 129(1): 115–132. doi: 10.1007/s10533-016-0222-8. Jaynes, D.B., T.S. Colvin, D.l. Karlen, C.A. Cambardella, and D.W. Meek. 2001. Nitrate Loss in Subsurface Drainage as Affected by Nitrogen Fertilizer Rate. Journal of Environmental Quality 30(4): 1305–1314. doi: 10.2134/jeq2001.3041305x. Jia, X., T. Scherer, T. Desutter, and D. Steele. 2008. Change of soil hardness and soil properties due to tile drainage in the Red River Valley of the North. Am. Soc. Agric. Biol. Eng. Annu. Int. Meet. 2008 ASABE 2008 8. doi: 10.13031/2013.24968. Jiang, Q., Z. Qi, L. Xue, M. Bukovsky, C.A. Madramootoo, et al. 2020. Assessing climate change impacts on greenhouse gas emissions, N losses in drainage and crop production in a subsurface drained field. Science of The Total Environment 705: 135969. doi: 10.1016/j.scitotenv.2019.135969. Jin, C.X., and G. Sands. 2003. The Long-term Field-scale Hydrology of Subsurface Drainage Systems in a Cold Climate. Trans. Am. Soc. Agric. Eng. 46: 1011–1021. doi: 10.13031/2013.13963. 77 Jin, C.X., G.R. Sands, H.J. Kandel, J.H. Wiersma, and B.J. Hansen. 2008. Influence of Subsurface Drainage on Soil Temperature in a Cold Climate. J. Irrig. Drain. Eng. 134(1): 83–88. doi: 10.1061/(ASCE)0733-9437(2008)134:1(83). Jungkunst, H.F., A. Freibauer, H. Neufeldt, and G. Bareth. 2006. Nitrous oxide emissions from agricultural land use in Germany— a synthesis of available annual field data. Journal of Plant Nutrition and Soil Science 169(3): 341–351. doi: 10.1002/jpln.200521954. Kalbitz, K., S. Solinger, J.H. Park, B. Michalzik, and E. Matzner. 2000. Controls on the dynamics of dissolved organic matter in soils; a review. Soil science 165(4): 277– 304. Kasper, M., G.D. Buchan, A. Mentler, and W.E.H. Blum. 2009. Influence of soil tillage systems on aggregate stability and the distribution of C and N in different aggregate fractions. Soil Tillage Res. 105(2): 192–199. doi: 10.1016/j.still.2009.08.002. Kaur, G., G. Singh, P.P. Motavalli, K.A. Nelson, J.M. Orlowski, et al. 2020. Impacts and management strategies for crop production in waterlogged or flooded soils: A review. Agronomy Journal 112(3): 1475–1501. doi: 10.1002/agj2.20093. Kelley, C.J., C.K. Keller, R.D. Evans, C.H. Orr, J.L. Smith, et al. 2013. Nitrate–nitrogen and oxygen isotope ratios for identification of nitrate sources and dominant nitrogen cycle processes in a tile-drained dryland agricultural field. Soil Biology and Biochemistry 57: 731–738. doi: 10.1016/j.soilbio.2012.10.017. Kelly, B.P., and M.L. Pomes. 1998. Preferential Flow and Transport of Nitrate and Bromide in Clay pan Soil. Groundwater 36(3): 484–494. doi: 10.1111/j.1745- 6584.1998.tb02820.x. Kemper, W.D., and R.C. Rosenau. 1986. Aggregate Stability and Size Dlstributlon. 9(2): 425-442. King, K.W., M.R. Williams, and N.R. Fausey. 2015. Contributions of Systematic Tile Drainage to Watershed-Scale Phosphorus Transport. J. Environ. Qual. 44(2): 486– 494. doi: 10.2134/jeq2014.04.0149. Kladivko, E.J., and L.C. Bowling. 2021. Long-term impacts of drain spacing, crop management, and weather on nitrate leaching to subsurface drains. J. Environ. Qual. 50(3): 627–638. doi: 10.1002/jeq2.20215. Kladivko, E.J., G.L. Willoughby, and J.B. Santini. 2005. Corn Growth and Yield Response to Subsurface Drain Spacing on Clermont Silt Loam Soil. Agronomy Journal 97(5): 1419–1428. doi: 10.2134/agronj2005.0090. Kladivko, E.J., J.R. Frankenberger, D.B. Jaynes, D.W. Meek, B.J. Jenkinson, et al. 2004. Nitrate Leaching to Subsurface Drains as Affected by Drain Spacing and Changes in 78 Crop Production System. J. Environ. Qual. 33(5): 1803–1813. doi: 10.2134/jeq2004.1803. Kladivko, E.J., L.C. Brown, and J.L. Baker. 2001. Pesticide Transport to Subsurface Tile Drains in Humid Regions of North America. Crit. Rev. Environ. Sci. Technol. 31(1): 1–62. doi: 10.1080/20016491089163. Kokulan, V., M.L. Macrae, D.A. Lobb, and G.A. Ali. 2019b. Contribution of Overland and Tile Flow to Runoff and Nutrient Losses from Vertisols in Manitoba, Canada. Journal of Environmental Quality 48(4): 959–965. doi: 10.2134/jeq2019.03.0103. Kokulan, V., M.L. Macrae, G.A. Ali, and D.A. Lobb. 2019a. Hydroclimatic controls on runoff activation in an artificially drained, near-level vertisolic clay landscape in a Prairie climate. Hydrological Processes 33(4): 602–615. doi: 10.1002/hyp.13347. Korsaeth, A., and R. Eltun. 2000. Nitrogen mass balances in conventional, integrated and ecological cropping systems and the relationship between balance calculations and nitrogen runoff in an 8-year field experiment in Norway. Agriculture, Ecosystems & Environment 79(2–3): 199–214. doi: 10.1016/S0167-8809(00)00129-8. Kovalev, I.V., V.M. Semenov, N.O. Kovaleva, T.N. Lebedeva, V.M. Yakovleva. 2021. Estimation of the Biogenicity and Bioactivity of Gleyed Agrogray Nondrained and Drained Soils. Eurasian Soil Sci. 54(7): 1059–1067. doi: 10.1134/S1064229321070073. Krichels, A., E.H. DeLucia, R. Sanford, J. Chee-Sanford, and W.H. Yang. 2019. Historical soil drainage mediates the response of soil greenhouse gas emissions to intense precipitation events. Biogeochemistry 142(3): 425–442. doi: 10.1007/s10533-019-00544-x. Kumar, S., T. Nakajima, A. Kadono, R. Lal, and N. Fausey. 2014. Long-term tillage and drainage influences on greenhouse gas fluxes from a poorly drained soil of central Ohio. Journal of Soil and Water Conservation 69(6): 553–563. doi: 10.2489/jswc.69.6.553. Kumar, S., T. Nakajima, E.G. Mbonimpa, S. Gautam, U.R. Somireddy, 2014. Long-term Tillage and Drainage Influences on Soil Organic Carbon Dynamics, Aggregate Stability and Corn Yield. 60(1): 108-118. doi: https://doi.org/10.1080/00380768.2013.878643 Kuzyakov, Y., J.K. Friedel, and K. Stahr. 2000. Review of mechanisms and quantification of priming effects. Soil Bio. 32(11-12): 1485-1498. doi: https://doi.org/10.1016/S0038-0717(00)00084-5 Lal, R., and N.R. Fausey. 1993. Drainage and tillage effects on a crosby-kokomo soil association in Ohio IV. Soil physical properties. Soil Technol. 6(2): 123–135. doi: 10.1016/0933-3630(93)90001-U. 79 Lehmann, A., K. Fitschen, and M. Rillig. 2019. Abiotic and Biotic Factors Influencing the Effect of Microplastic on Soil Aggregation. Soil Syst. 3(1): 21. doi: 10.3390/soilsystems3010021. Lehrsch, G.A., R.E. Sojka, D.L. Carter, and P.M. Jolley. 1991. Freezing effects on aggregate stability affected by Texture, Mineralogy, and Organic Matter. 55: 1401- 1406. Doi: https://doi- org.ezp2.lib.umn.edu/10.2136/sssaj1991.03615995005500050033x Libohova, Z., D.E. Stott, P.R. Owens, H.E. Winzeler, and S. Wills. 2014. Mineralizable Soil Organic Carbon Dynamics in Corn-Soybean Rotations in Glaciated Derived Landscapes of Northern Indiana. In: Hartemink, A.E. and McSweeney, K., editors, Soil Carbon. Springer International Publishing, Cham. p. 259–269 Lundekvam, H. 1998. P-losses from three soil types at different cultivation systems. K Skogs-Og Lantbr Akad Tidskr 137(7): 177–185. Marinho, E.V.A., R.D. DeLaune, and C.W. Lindau. 2004. Nitrous Oxide Flux from Soybeans Grown on Mississippi Alluvial Soil. Communications in Soil Science and Plant Analysis 35(1–2): 1–8. doi: 10.1081/CSS-120027630. Matejovic, I. 1995. Total nitrogen in plant material determinated by means of dry combustion: A possible alternative to determination by Kjeldahl digestion. Communications in Soil Science and Plant Analysis 26(13–14): 2217–2229. doi: 10.1080/00103629509369441. McIsaac, G.F., and X. Hu. 2004. Net N input and riverine N export from Illinois agricultural watersheds with and without extensive tile drainage. Biogeochemistry 70(2): 253–273. doi: 10.1023/B:BIOG.0000049342.08183.90. McTiernan, K.B., S.C. Jarvis, D. Scholefield, and M.H.B. Hayes. 2001. Dissolved organic carbon losses from grazed grasslands under different management regimes. Water Res. 35(10): 2565–2569. doi: 10.1016/S0043-1354(00)00528-5. Mehring, G.H., H. J. Kandel, J.K. Ransom, A. Schoch, and D.D. Steele. 2015. Spring Wheat Response to Disease Control and Subsurface Drainage Management in the Red River of the North Valley, USA. AS 06(10): 1220–1231. doi: 10.4236/as.2015.610117. Messing, I., and N.J. Jarvis. 1990. Seasonal variation in field-saturated hydraulic conductivity in two swelling clay soils in Sweden. J. Soil Sci. 41(2): 229–237. doi: 10.1111/j.1365-2389.1990.tb00059.x. Meter Group. 2017 http://library.metergroup.com/Manuals/20496_SATURO_Manual.pdf (accessed 25 August 2021). Midwest Regional Climate Center. 2021. https://mrcc.purdue.edu/ (accessed 25 April 80 2021). Mikha, M.M., and C.W. Rice. 2004. Tillage and Manure Effects on Soil and Aggregate- Associated Carbon and Nitrogen. Soil Sci. Soc. Am. J. 68(3): 809–816. doi: 10.2136/sssaj2004.8090. Miller, B.A., and R.J. Schaetzl. 2012. Precision of Soil Particle Size Analysis using Laser Diffractometry. Soil Sci. Soc. Am. J. 76(5): 1719–1727. doi: 10.2136/sssaj2011.0303. Mitchell, A.R., T.R. Ellsworth, and B.D. Meek. 1995. Effect of root systems on preferential flow in swelling soil. Commun. Soil Sci. Plant Anal. 26(15–16): 2655– 2666. doi: 10.1080/00103629509369475. Mitchell, K.S. 2020. Long-term Effects of Subsurface Drain Spacing on Soil Physical and Chemical Properties. https://doi.org/10.25394/PGS.12721790.v1 Morton, L.W., J. Hobbs, J.G. Arbuckle, and A. Loy. 2015. Upper Midwest Climate Variations: Farmer Responses to Excess Water Risks. Journal of Environmental Quality 44(3): 810–822. doi: 10.2134/jeq2014.08.0352. Motha, R.P., and W. Baier. 2005. Impacts of Present and Future Climate Change and Climate Variability on Agriculture in the Temperate Regions: North America. Clim. Chan. 70: 137-164. Mourtzinis, S., J.F. Andrade, P. Grassini, J.I.R. Edreira, H. Kandel, et al. 2021. Assessing benefits of artificial drainage on soybean yield in the North Central US region. Agricultural Water Management 243: 106425. doi: 10.1016/j.agwat.2020.106425. Mukhtar, S., J. L. Baker, and R. S. Kanwar. 1990. Corn Growth as Affected by Excess Soil Water. Transactions of the ASAE 33(2): 0437–0442. doi: 10.13031/2013.31348. Norris, C.E., G.M. Bean, S.B. Cappellazzi, M. Cope, K.L.H. Greub, et al. 2020. Introducing the North American project to evaluate soil health measurements. Agron. J. 112(4): 3195–3215. doi: 10.1002/agj2.20234. Nuutinen, V., S. Pöyhönen, E. Ketoja, and J. Pitkänen. 2001. Abundance of the earthworm Lumbricus terrestris in relation to subsurface drainage pattern on a sandy clay field. Eur. J. Soil Biol. 37(4): 301–304. doi: 10.1016/S1164-5563(01)01105-0. Paiao, G.D., F.G. Fernández, and S.L. Naeve. 2021. Drainage Conditions Influence Corn- Nitrogen Management in the US Upper Midwest. Agronomy 11(12): 2491. doi: 10.3390/agronomy11122491. Palmer, J., P.J. Thorburn, J.S. Biggs, E.J. Dominati, M.E. Probert, et al. 2017. Nitrogen Cycling from Increased Soil Organic Carbon Contributes Both Positively and 81 Negatively to Ecosystem Services in Wheat Agro-Ecosystems. Frontiers in Plant Science 8. https://www.frontiersin.org/article/10.3389/fpls.2017.00731 (accessed 26 January 2022). Patel, K.F., A. Myers-Pigg, B. Bond-Lamberty, S.J. Fansler, C.G. Norris, et al. 2021a. Soil carbon dynamics during drying vs. rewetting: Importance of antecedent moisture conditions. Soil Biol. Biochem. 156: 108165. doi: 10.1016/j.soilbio.2021.108165. Patel, K.F., S.J. Fansler, T.P. Campbell, B. Bond-Lamberty, A.P. Smith, et al. 2021b. Soil texture and environmental conditions influence the biogeochemical responses of soils to drought and flooding. Commun Earth Environ 2(1): 127. doi: 10.1038/s43247-021-00198-4. Patton, C.J., and J.R. Kryskalla. 2003. Methods of analysis by the US Geological Survey National Water Quality Laboratory: evaluation of alkaline persulfate digestion as an alternative to Kjeldahl digestion for determination of total and dissolved nitrogen and phosphorus in water. US Department of the Interior, US Geological Survey. Paul, E.A., and R.J.K. Myers. 1971. Effect of Soil Moisture Stress on Uptake and Recovery of Tagged Nitrogen by Wheat. Can. J. Soil. Sci. 51(1): 37–43. doi: 10.4141/cjss71-005. Pease, L., N.R. Fausey, J.F. Martin, and L. Brown. 2017. Projected climate change effects on subsurface drainage and the performance of controlled drainage in the Western Lake Erie Basin. Journal of Soil and Water Conservation 72: 240–250. doi: 10.2489/jswc.72.3.240. Pluer, W.T., M. Macrae, A. Buckley, and K. Reid. 2020. Contribution of preferential flow to tile drainage varies spatially and temporally. Vadose Zone Journal 19(1): e20043. doi: 10.1002/vzj2.20043. R Development Core Team. 2013. R: A language and environment for statistical computing. Rabalais, N.N., R.E. Turner, R.J. Díaz, and D. Justić. 2009. Global change and eutrophication of coastal waters. ICES Journal of Marine Science 66(7): 1528–1537. doi: 10.1093/icesjms/fsp047. Randall, G.W., and D.J. Mulla. 2001. Nitrate Nitrogen in Surface Waters as Influenced by Climatic Conditions and Agricultural Practices. J. Environ. Qual. 30(2): 337– 344. doi: 10.2134/jeq2001.302337x. Randall, G.W., and D.J. Mulla. 2001. Nitrate Nitrogen in Surface Waters as Influenced by Climatic Conditions and Agricultural Practices. Journal of Environmental Quality 30(2): 337–344. doi: 10.2134/jeq2001.302337x. 82 Randall, G.W., and M.J. Goss. Chapter 6. Nitrate Losses to Surface Water Through Subsurface, Tile Drainage. Nitrogen in the Environment: 31. Randall, G.W., and T.K. Iragavarapu. 1995. Impact of Long-Term Tillage Systems for Continuous Corn on Nitrate Leaching to Tile Drainage. Journal of Environmental Quality 24(2): 360–366. doi: 10.2134/jeq1995.00472425002400020020x. Rankinen, K., T. Salo, K. Granlund, and H. Rita. 2007. Simulated nitrogen leaching, nitrogen mass field balances and their correlation on four farms in south-western Finland during the period 2000–2005. Ag. Fd. Science 16: 387-406. Rattan, K.J., J.C. Corriveau, R.B. Brua, J.M. Culp, A.G. Yates, et al. 2017. Quantifying seasonal variation in total phosphorus and nitrogen from prairie streams in the Red River Basin, Manitoba Canada. Sci. Tot. Env. 575: 649-659. doi: https://doi.org/10.1016/j.scitotenv.2016.09.073 Rawls, W.J., D. Gimenez, and R. Grossman. Use of Soil Texture, Bulk Density, and Slope of the Water Retention Curve to Predict Saturated Hydraulic Conductivity. Am. Soc. Agric. Biol. Eng.41(4): 983-988. doi: 10.13031/2013.17270 Reddy, K.R., W.H. Patrick, and F.E. Broadbent. 1984. Nitrogen transformations and loss in flooded soils and sediments. C R C Critical Reviews in Environmental Control 13(4): 273–309. doi: 10.1080/10643388409381709. Reynolds, W.D., and D.E. Elrick. 1990. Ponded Infiltration From a Single Ring: I. Analysis of Steady Flow. Soil Sci. Soc. Am. J. 54(5): 1233–1241. doi: 10.2136/sssaj1990.03615995005400050006x. Rice, C.W., J.H. Grove, and M.S. Smith. 1987. Estimating Soil Net Nitrogen Mineralization as Affected by Tillage and Soil Drainage Due to Topographic Position. Can. J. Soil. Sci. 67(3): 513–520. doi: 10.4141/cjss87-048. Rochette, P., and L.B. Flanagan. 1997. Quantifying Rhizosphere Respiration in a Corn Crop under Field Conditions. Soil Sci. Soc. Am. J. 61(2): 466–474. doi: 10.2136/sssaj1997.03615995006100020014x. Ros, G.H., E. Hoffland, C. van Kessel, and E.J.M. Temminghoff. 2009. Extractable and dissolved soil organic nitrogen – A quantitative assessment. Soil Biology and Biochemistry 41(6): 1029–1039. doi: 10.1016/j.soilbio.2009.01.011. Ruark, M.D., S.M. Brouder, and R.F. Turco. 2009. Dissolved Organic Carbon Losses from Tile Drained Agroecosystems. J. Environ. Qual. 38(3): 1205–1215. doi: 10.2134/jeq2008.0121. Rubin, J.C., A.M. Struffert, F.G. Fernández, and J.A. Lamb. 2016. Maize Yield and Nitrogen Use Efficiency in Upper Midwest Irrigated Sandy Soils. Agronomy Journal 108(4): 1681–1691. doi: 10.2134/agronj2015.0552. 83 Rückauf, U. 2004. Nitrate removal from drained and reflooded fen soils affected by soil N transformation processes and plant uptake. Soil Biology and Biochemistry 36(1): 77–90. doi: 10.1016/j.soilbio.2003.08.021. Schindlbacher, A., S. Zechmeister-Boltenstern, and K. Butterbach-Bahl. 2004. Effects of soil moisture and temperature on NO, NO2, and N2O emissions from European forest soils. Journal of Geophysical Research: Atmospheres 109(D17). doi: 10.1029/2004JD004590. Schindler, D.W., R.E. Hecky, and G.K. McCullough. 2012. The rapid eutrophication of Lake Winnipeg: Greening under global change. Journal of Great Lakes Research 38: 6–13. doi: 10.1016/j.jglr.2012.04.003. Shah, N., and G. Paulsen. 2003. Interaction of drought and high temperature on photosynthesis and grain-filling of wheat. Plant Soil 257(1): 219–226. Shcherbak, I., N. Millar, and G.P. Robertson. 2014. Global metaanalysis of the nonlinear response of soil nitrous oxide (N 2 O) emissions to fertilizer nitrogen. Proc. Natl. Acad. Sci. U.S.A. 111(25): 9199–9204. doi: 10.1073/pnas.1322434111. Sierra, J. 1997. Temperature and soil moisture dependence of N mineralization in intact soil cores. Soil Biology and Biochemistry 29(9–10): 1557–1563. doi: 10.1016/S0038-0717(96)00288-X. Six, J., E.T. Elliott, and K. Paustian. 2000. Soil macroaggregate turnover and microaggregate formation: a mechanism for C sequestration under no-tillage agriculture. Soil Biol. Biochem. 32(14): 2099–2103. doi: 10.1016/S0038- 0717(00)00179-6. Skaggs, R.W., and J. W. Gilliam. 1981. Effect of Drainage System Design and Operation on Nitrate Transport. Transactions of the ASAE 24(4): 929–934. doi: 10.13031/2013.34366. Skaggs, R.W., M.A. Brevé, and J.W. Gilliam. 1994. Hydrologic and water quality impacts of agricultural drainage. Crit. Rev. Environ. Sci. Technol. 24(1): 1–32. doi: 10.1080/10643389409388459. Skaggs, R.W., N.R. Fausey, and R.O. Evans. 2012. Drainage water management. Journal of Soil and Water Conservation 67(6): 167A-172A. doi: 10.2489/jswc.67.6.167A. Smith, A.P., B. Bond-Lamberty, B.W. Benscoter, M.M. Tfaily, C.R. Hinkle, et al. 2017. Shifts in pore connectivity from precipitation versus groundwater rewetting increases soil carbon loss after drought. Nat. Commun. 8(1): 1335. doi: 10.1038/s41467-017-01320-x. Snyder, C.S., T.W. Bruulsema, T.L. Jensen, and P.E. Fixen. 2009. Review of greenhouse gas emissions from crop production systems and fertilizer management effects. 84 Agriculture, Ecosystems & Environment 133(3–4): 247–266. doi: 10.1016/j.agee.2009.04.021. Stamm, C., H. Flühler, R. Gächter, J. Leuenberger, and H. Wunderli. 1998. Preferential Transport of Phosphorus in Drained Grassland Soils. J. Environ. Qual. 27(3): 515– 522. doi: 10.2134/jeq1998.00472425002700030006x. Stewart, R.D., J. Jian, A.J. Gyawali, W.E. Thomason, B.D. Badgley, et al. 2018. What We Talk about When We Talk about Soil Health. Agric. Environ. Lett. 3(1): 180033. doi: 10.2134/ael2018.06.0033. Stone, W.W., and J.T. Wilson. 2006. Preferential Flow Estimates to an Agricultural Tile Drain with Implications for Glyphosate Transport. J. Environ. Qual. 35(5): 1825– 1835. doi: 10.2134/jeq2006.0068. Sugihara, S., S. Funakawa, and T. Kosaki. 2010. In situ short-term carbon and nitrogen dynamics in relation to microbial dynamics after a simulated rainfall in croplands of different soil texture in Thailand. Soil Science and Plant Nutrition 56(6): 813–823. doi: 10.1111/j.1747-0765.2010.00516.x. Sunohara, M.D., E. Craiovan, E. Topp, N. Gottschall, C.F. Drury, et al. 2014. Comprehensive Nitrogen Budgets for Controlled Tile Drainage Fields in Eastern Ontario, Canada. Journal of Environmental Quality 43(2): 617–630. doi: 10.2134/jeq2013.04.0117. Talukolaee, J.M., A. Darzi Naftchali, L. Zare Parvariji, and M.Z. Ahmadi. 2018. Investigating long-term effects of subsurface drainage on soil structure in paddy fields. Soil Tillage Res. 177: 155–160. doi: 10.1016/j.still.2017.12.012. Thorp, K. R., D. B. Jaynes, and R. W. Malone. 2008. Simulating the Long-Term Performance of Drainage Water Management Across the Midwestern United States. Trans. ASABE 51(3): 961–976. doi: 10.13031/2013.24534. Tisdall, J.M., and J.M. Oades. 1982. Organic matter and water-stable aggregates in soils. Journal of Soil Science 33(2): 141–163. doi: 10.1111/j.1365-2389.1982.tb01755.x. USDA. 2017. Census of Agriculture. https://www.nass.usda.gov/AgCensus/ (accessed April 1 2022). van der Werff, P.A., A. Baars, and G.J.M. Oomen. 1995. Nutrient Balances and Measurement of Nitrogen Losses on Mixed Ecological Farms on Sandy Soils in the Netherlands. Biological Agriculture & Horticulture 11(1–4): 41–50. doi: 10.1080/01448765.1995.9754692. Van Esbroeck, C.J., M.L. Macrae, R.I. Brunke, and K. McKague. 2016. Annual and seasonal phosphorus export in surface runoff and tile drainage from agricultural fields with cold temperate climates. Journal of Great Lakes Research 42(6): 1271– 85 1280. doi: 10.1016/j.jglr.2015.12.014. van Kessel, C., T. Clough, and J.W. van Groenigen. 2009. Dissolved Organic Nitrogen: An Overlooked Pathway of Nitrogen Loss from Agricultural Systems? Journal of Environmental Quality 38(2): 393–401. doi: 10.2134/jeq2008.0277. Venterea, R.T., A.D. Halvorson, N. Kitchen, M.A. Liebig, M.A. Cavigelli, et al. 2012. Challenges and opportunities for mitigating nitrous oxide emissions from fertilized cropping systems. Frontiers in Ecology and the Environment 10(10): 562–570. doi: 10.1890/120062. Wang, L., D.-C. Hao, S. Fan, H. Xie, X. Bao, et al. 2022. N2O Emission and Nitrification/Denitrification Bacterial Communities in Upland Black Soil under Combined Effects of Early and Immediate Moisture. Agriculture 12(3): 330. doi: 10.3390/agriculture12030330. Wassenaar, L.I., and Y.R. Rao. 2012. Lake Winnipeg: The forgotten great lake. Journal of Great Lakes Research 38: 1–5. doi: 10.1016/j.jglr.2012.04.004. Webster, E.A., and D.W. Hopkins. 1995. Contributions from different microbial processes to N20 emissionfrom soil under different moisture regimes. Bio. Fertil. Soils. 22: 331-335. Weitz, A.M., E. Linder, S. Frolking, P.M. Crill, and M. Keller. 2001. N 2 O emissions from humid tropical agricultural soils: effects of soil moisture, texture and nitrogen availability. Soil Biology and Biochemistry 33(7–8): 1077–1093. doi: 10.1016/S0038-0717(01)00013-X. Welage, D.T., 2020. Long-term Subsurface Drainage Effects on Soil Physical and Hydraulic Properties. https://doi.org/10.25394/PGS.12388502.v1 Wells, E.D., and B.L. Williams. 1996. Effects of drainage, tilling and PK-fertilization on bulk density, total N, P, K, Ca and Fe and net N-mineralization in two peatland forestry sites in Newfoundland, Canada. For. Ecol. Manag. 84(1–3): 97–108. doi: 10.1016/0378-1127(96)03741-3. Wiersma, J.J., G.R. Sands, H.J. Kandel, A.K. Rendahl, C.X. Jin, et al. 2010. Responses of Spring Wheat and Soybean to Subsurface Drainage in Northwest Minnesota. Agronomy Journal 102(5): 1399–1406. doi: 10.2134/agronj2010.0055. Williams, M.R., K.W. King, M.L. Macrae, W. Ford, C. Van Esbroeck, et al. 2015. Uncertainty in nutrient loads from tile-drained landscapes: Effect of sampling frequency, calculation algorithm, and compositing strategy. J. Hydrol. 530: 306– 316. doi: 10.1016/j.jhydrol.2015.09.060. Wolf, K.A., S.C. Gupta, and C.J. Rosen. 2020. Precipitation Drives Nitrogen Load Variability in Three Iowa Rivers. Journal of Hydrology: Regional Studies 30: 86 100705. doi: 10.1016/j.ejrh.2020.100705. Wood, S.A., N. Sokol, C.W. Bell, M.A. Bradford, S. Naeem, et al. 2016. Opposing effects of different soil organic matter fractions on crop yields. Ecol. Appl. 26(7): 2072–2085. doi: 10.1890/16-0024.1. Wurtsbaugh, W.A., H.W. Paerl, and W.K. Dodds. 2019. Nutrients, eutrophication and harmful algal blooms along the freshwater to marine continuum. WIREs Water 6(5): e1373. doi: 10.1002/wat2.1373. Yates, A.G., J.M. Culp, and P.A. Chambers. 2012. Estimating nutrient production from human activities in subcatchments of the Red River, Manitoba. Jr. Gr. Lak. Res. 38: 106-114. doi: doi:10.1016/j.jglr.2011.04.009 Zhu, Q., M.J. Castellano, and G. Yang. 2018. Coupling soil water processes and the nitrogen cycle across spatial scales: Potentials, bottlenecks and solutions. Earth- Science Reviews 187: 248–258. doi: 10.1016/j.earscirev.2018.10.005. Zumft, W.G. 1997. Cell biology and molecular basis of denitrification. Microbiol Mol Biol Rev 61(4): 533–616.