Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Forage crop nitrogen recovery and nitrogen field-losses determined on semi-virtual dairy farms under… Li, Yuchen 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_september_2019_li_yuchen.pdf [ 5.43MB ]
Metadata
JSON: 24-1.0378545.json
JSON-LD: 24-1.0378545-ld.json
RDF/XML (Pretty): 24-1.0378545-rdf.xml
RDF/JSON: 24-1.0378545-rdf.json
Turtle: 24-1.0378545-turtle.txt
N-Triples: 24-1.0378545-rdf-ntriples.txt
Original Record: 24-1.0378545-source.json
Full Text
24-1.0378545-fulltext.txt
Citation
24-1.0378545.ris

Full Text

FORAGE CROP NITROGEN RECOVERY AND NITROGEN FIELD-LOSSES DETERMINED ON SEMI-VIRTUAL DAIRY FARMS UNDER INTEGRATED NUTRIENT AND CROP MANAGEMENT SCENARIOS by  YUCHEN LI  B.Sc., Simon Fraser University, 2016  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Soil Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2019   © Yuchen Li, 2019  ii   The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled: Forage crop nitrogen recovery and nitrogen field-losses determined on semi-virtual dairy farms under integrated nutrient and crop management scenarios  submitted by Yuchen Li in partial fulfillment of the requirements for the degree of Master of Science in Soil Science  Examining Committee: Sean Smukler Co-supervisor Shabtai Bittman Co-supervisor  Nina von Keyserlingk Supervisory Committee Member Mark Johnson Additional Examiner  Additional Supervisory Committee Members: Maja Krzic Supervisory Committee Member  Supervisory Committee Member iii  Abstract In the Lower Fraser Valley region of BC, intense dairy production occupies a relatively small agricultural land base. Farm-generated manure is often applied to forage crops at a very high rate to deal with the large quantities produced by high animal density. Local farmers also purchase additional mineral fertilizer to maximize production and profitability, which together with manure has far exceeded the nutrient uptake capability of crops. Surplus nitrogen (N) will lead to N loss through nitrate leaching and nitrous oxide emission.   The objectives of this study are to quantify and compare (1) the crop N removal and apparent recovery of total N applied (TNR); and (2) nitrous oxide emission and nitrate leaching intensity of four dairy farm management scenarios that incrementally introduce beneficial management practices (BMPs) and advanced production techniques. I compared these four scenarios for annual crop yield, crop N removal, and TNR of silage corn and tall fescue. I quantified the seasonal and annual N field-losses through nitrous oxide emission and potential nitrate leaching. Finally, I also examined how adjusting cropland allocation would affect total crop yield, crop N removal, TNR, and N field-loss for the four scenarios.   Planting a relay crop removed more N from corn plots while producing a similar amount of feed as the conventional scenario and also reduced nitrate leaching and leaching intensity by 70 %. Reduced grass harvest frequency increased grass yield substantially. Nitrification inhibitor, DCD, and irrigation improved TNR of both corn and grass and also reduced nitrous oxide emission from grass plots in 2017 and corn plots in 2018. In the dual-crop forage production system of 50 % corn and 50 % grass, TNR increased significantly only when all available BMPs of this study were integrated. Total nitrate leaching intensity abated more than 50 % when the relay crop and reduced grass harvest frequency were implemented. Adding DCD and irrigation reduced annual total nitrous oxide emission intensity by 40 %. Allocating 10 % more land to grow corn increased total feed production, but had little impact on total nitrate leaching and nitrous oxide emission. iv  Lay Summary This study compares four dairy farm management scenarios in the Lower Fraser Valley that incrementally introduce beneficial management practices (BMPs) and advanced production techniques. The results  have shown that (1) the combination of manure separation and manure incorporation techniques can replace application of starter fertilizer, which can reduce both the cost of farm operations and nutrient surplus; (2) intercropping Italian ryegrass with silage corn as relay crop can be very effective in reducing potential groundwater contamination and also provide supplement to feedstuff; (3) although not for the region irrigation combined with a nitrification inhibitor can increase silage corn crop yield and recover more plant available nitrogen; and (4) increasing the area of silage corn can also improve nutrient recovery and increase feed production.  These finding provided an idea what the compounded benefits would be from integrating a suite of beneficial management practices in dairy forage crop production. v  Preface The work presented henceforth is the result of a collaboration between the Agassiz Research and Development Center (ARDC), and the Sustainable Agricultural Landscapes Lab, at the University of British Columbia. Design and initiation of the current experiment in 2016 were performed by Dr. Shabtai Bittman and his research crew at the ARDC. Data of chapter 2 and 3 were collected in collaboration with myself and the research team at the ARDC, including Dr. Shabtai Bittman, Derek Hunt, Frederic Bounaix, Anthony Friesen, and many other technicians and coop students. With the tremendous help of this team, we carried out soil sampling, crop sampling, air sampling, and soil water sampling. I conducted crop tissue N analysis, soil extractions, gas chromatography analysis on air samples. Anthony Friesen carried out flow injection analysis on soil extractions and soil water samples, and he also aided me with gas chromatography analysis on air samples. Frederic Bounaix carried out all field machinery works and manure N analysis. I am responsible for all data analysis and interpretation of chapter 2 and 3, including crop yield and N recovery, potential NO3-−N leaching; and N2O−N emission. I developed the soil water balance model used in chapter 3 based on the irrigation and drainage paper 56 by Food and Agriculture Organization (FAO) of the United Nations with help from Derek Hunt. Pierre Groenenboom aided me with the calculation of nitrous oxide fluxes. All data analysis and interpretation are my original work. vi  Table of Contents  Abstract ......................................................................................................................................... iii	Lay Summary ............................................................................................................................... iv	Preface .............................................................................................................................................v	Table of Contents ......................................................................................................................... vi	List of Tables ..................................................................................................................................x	List of Figures ............................................................................................................................. xiii	List of Abbreviations ...................................................................................................................xv	Acknowledgments ...................................................................................................................... xvi	Chapter 1: Introduction ................................................................................................................1	1.1	 Nitrogen dynamics in a forage production system ......................................................... 3	1.1.1	 Transformation of reactive nitrogen ........................................................................... 4	1.1.2	 Nitrogen field-loss pathways ...................................................................................... 4	1.1.3	 Dynamics of soil phosphorus ...................................................................................... 7	1.2	 Nutrient and crop management of the dairy farm in the Lower Fraser Valley, BC ....... 8	1.2.1	 Beneficial management practices for nutrients ........................................................... 9	1.2.2	 Beneficial management practices for cropping ......................................................... 11	1.3	 Research objectives and hypotheses ............................................................................. 13	1.3.1	 Study 1: Crop N removal and total N recovery of experimental farmlets under different nutrient and cropping management scenarios ........................................................ 14	vii  1.3.2	 Study 2: Evaluating the efficacy of integrated beneficial management practices on reducing potential nitrate leaching and nitrous oxide emission from dairy farm forage production systems ................................................................................................................ 15	Chapter 2: Crop N Removal and Total N Recovery of Experimental Farmlets under Different Nutrient and Cropping Management Scenarios .......................................................16	2.1	 Introduction ................................................................................................................... 16	2.2	 Material and Methods ................................................................................................... 18	2.2.1	 Site description and field trial setup .......................................................................... 18	2.2.2	 Experimental Design ................................................................................................. 19	2.2.3	 Forage production system management scenarios .................................................... 20	2.2.4	 Crop sampling and analysis ...................................................................................... 26	2.2.5	 Manure analysis ........................................................................................................ 29	2.2.6	 Total Nitrogen Recovery ........................................................................................... 30	2.2.7	 Farmlet Area-weight Total ........................................................................................ 31	2.2.8	 Statistical analysis ..................................................................................................... 32	2.3	 Results and Discussion ................................................................................................. 33	2.3.1	 Silage corn yield ....................................................................................................... 33	2.3.2	 Silage corn N removal and TN recovery .................................................................. 36	2.3.3	 Grass yield ................................................................................................................ 40	2.3.4	 Grass N removal and TN recovery ........................................................................... 42	2.3.5	 Farmlet area-weighted total dry-matter yield and N removal rate for different land allocation scenarios ............................................................................................................... 46	2.4	 Conclusions ................................................................................................................... 51	viii  Chapter 3: Evaluating the Effectiveness of Beneficial Management Practices on Reducing Potential Nitrate Leaching and Nitrous Oxide Emission from Dairy Farm Forage Production Systems ......................................................................................................................53	3.1	 Introduction ................................................................................................................... 53	3.2	 Material and Methods ................................................................................................... 56	3.2.1	 Site description and Experiment design .................................................................... 56	3.2.2	 Soil sampling and plant available N (PAN) analysis ................................................ 56	3.2.3	 Soil pore water sampling and quantification of potential NO3-–N leaching ............. 57	3.2.4	 Gas emission sampling and quantification of N2O–N emission ............................... 64	3.2.5	 Statistical analysis ..................................................................................................... 66	3.3	 Results and Discussion ................................................................................................. 68	3.3.1	 Potential nitrate leaching ........................................................................................... 68	3.3.2	 Nitrous oxide emissions ............................................................................................ 75	3.3.3	 Farmlet area-weighted total N losses ........................................................................ 85	3.4	 Conclusion .................................................................................................................... 88	Chapter 4: Conclusion .................................................................................................................90	4.1	 Research Conclusion ..................................................................................................... 90	4.1.1	 Crop N removal and total N recovery of experimental farmlets under different nutrient and cropping management scenarios ....................................................................... 90	4.1.2	 Efficacy of integrated beneficial management practices for reducing potential nitrate leaching and nitrous oxide emission from dairy farm forage production systems ............... 92	4.2	 Limitation and direction for future research ................................................................. 93	4.3	 The implication of this study for dairy farms in the Lower Fraser Valley ................... 94	ix  Bibliography .................................................................................................................................96	Appendices ..................................................................................................................................111	Appendix A Integration of beneficial management practices (BMPs). BMPs (in bold) were added to each scenario incrementally. .................................................................................... 111	Appendix B Soil sampling scheme for pre-sidedressing nitrate test (PSNT). Twenty cores of 0-15 cm deep soil samples were taken from each corn plot (6.1 m x 18.3 m) to form one composite sample. ................................................................................................................... 112	Appendix C Illustration of taking a soil water sample from ceramic-cup suction lysimeter. 113	Appendix D Basal crop coefficient (Kcb), plant height (h), and maximum rooting depth (Zr) of each crop type at different growing stage throughout two years. ........................................... 114	Appendix E List of parameters for calculation of evapotranspiration coefficient (ETc). ....... 115	 x  List of Tables Table 2.1 Timeline of field activities, including date of manure application, crop harvest, and N2O–N and NO3- –N measurement period. ................................................................................... 25	Table 2.2 Nitrogen and phosphorus input of two years, including commercial fertilizer and manure. .......................................................................................................................................... 26	Table 2.3 Characteristic of different manure types applied, including percentage of dry-matter content (DM%); total ammonium N (TAN), total N (TN); ammonium-N to total N ratio (TAN/TN); total phosphorus (TP); nitrogen to phosphorus ration (N/P), and pH. The range of variation is shown in parentheses for manure that was applied multiple times ............................ 30	Table 2.4 Analysis of variance of corn DM yield and crop N removal by corn. There are four levels of management scenario (df =3) and two levels of year (df = 1). Bolded values indicate statistical significance. .................................................................................................................. 33	Table 2.5 Crop DM yield of corn whole plant silage and sum of corn and relay crop. Numbers in parentheses are a standard error, n = 4 for each year, n = 8 for mean over two years. Relay crop DM yield in 2018-19 is not available. .......................................................................................... 34	Table 2.6 Corn grain DM yield and grain %. Numbers in parentheses are the standard error, n = 4 for each year, n = 8 for mean over two years. .............................................................................. 35	Table 2.7 Annual nitrogen (TAN and TN) application rate and nitrogen removal by silage corn. Numbers in parentheses are standard error, n = 4 for each year, n = 8 for mean over two years. Relay crop data is incomplete for 2018-19. .................................................................................. 38	Table 2.8. Analysis of variance of apparent recovery of TN. There are four levels of management scenario (df =3) and two levels of crop (df = 1). Bolded values indicate statistical significance. 39	xi  Table 2.9 Analysis of variance of tall fescue DM yield and tall fescue N removal (n=8) in 2017 and 2018, managed under four different scenarios. Significant results are highlighted in bold. .. 40	Table 2.10 Tall fescue DM yield in 2017 and 2018. Numbers in parentheses are the standard error. n = 4 for each year, n = 8 for mean over two years. ........................................................... 41	Table 2.11 Annual N (TAN and TN) application rate, N removal by tall fescue, and % TN recovery by tall fescue. Numbers in parentheses are the standard error. n = 4 for each year, n = 8 for mean over two years. ............................................................................................................... 45	Table 2.12 Analysis of variance of farmlet total yield and total N removal in 2017 and 2018, managed under four different scenarios (df =3) and two land allocation scenarios (df =1). Significant results are highlighted in bold. ................................................................................... 48	Table 2.13 Analysis of variance of total N recovery in 2017 and 2018, managed under four different scenarios (df =3) and two land allocation scenarios (df =1). Significant results are highlighted in bold. ....................................................................................................................... 48	Table 2.14 Annual crop yield, N removal, N (include commercial N and manure N) application rate, and P (including commercial P) application rate of four farmlets under two land allocation scenarios. n = 4 for mean of each farmlet and n=16 for annual mean of each crop area ratio. .... 50	Table 3.1 Soil hydraulic properties of each layer. Hydraulic parameters are explained in the text below. ............................................................................................................................................ 61	Table 3.2 Analysis of variance of non-production season potential NO3-N leaching, leaching factor, and leaching intensity (n=4) in the crop year 2017-18, managed under four different scenarios (df =3) and two types of crops (df =1). Significant results are highlighted in bold. ..... 71	Table 3.3 Potential nitrate leaching during the non-production season of the 2017-18 crop year. Values in parentheses are the standard error. n = 4 for each farmlet and n=16 for crop mean. xii  Differences between crop means were not discussed in the text when crop farmlet interaction is significant (P<0.05). ..................................................................................................................... 74	Table 3.4 Analysis of variance of cumulative N2O-N emission and emission factor (n=4) in two consecutive production seasons and annual cumulative emission of the crop year 2017-18, managed under four different scenarios (df =3) and two types of crops (df =1). Significant results are highlighted in bold .................................................................................................................. 80	Table 3.5 Cumulative nitrous oxide emission and emission factor of two consecutive production seasons. n = 4 for each farmlet and n=16 for crop mean. Differences between crop means were not discussed when crop farmlet interaction is significant (P < 0.05). ......................................... 81	Table 3.6 Annual cumulative nitrous oxide emission, emission factor, and emission intensity of the crop year 2017-2018. n = 4 for each farmlet and n=16 for crop mean. Difference between crop means are not discussed in the text when crop farmlet interaction is significant (P < 0.05). 83	Table 3.7 Analysis of variance of area-weighted total annual cumulative N2O–N emission, emission intensity, emission factor, annual cumulative NO3- –N leaching, and leaching intensity (n=4) in the crop year 2017-18, managed under four different management scenarios (df =3) and two land allocation scenarios (df =1). Significant results are highlighted in bold. ...................... 86	Table 3.8 Annual cumulative nitrous oxide emission, emission factor, and emission intensity of the crop year 2017-2018. n = 4 for each farmlet and n=16 for crop mean. Difference between crop means is not discussed when crop farmlet interaction is significant (P < 0.05). .................. 87	 xiii  List of Figures Figure 1.1 Nitrogen cycle of a dairy farm ....................................................................................... 7	Figure 2.1 Field trial setup. Randomized complete block design of 4 blocks with split-plot by type of crop and 4 management scenarios, i.e. treatments, were randomly assigned to 4 subplots of each crop. .................................................................................................................................. 20	Figure 2.2 Apparent recovery of TN applied for both crops in two consecutive years (2017-19). For each crop in one year, farmlets labeled with different lower-case letters are significantly different (P < 0.5). ......................................................................................................................... 39	Figure 2.3 Cumulative N removal for each farmlet by tall fescue throughout the growing season (March - October) in 2017-18 (left) and 2018-19 (right). Colors indicate different grass harvest frequency and arrows indicate the dates of manure applications. ................................................ 44	Figure 2.4 Area-weighted total recovery of TN applied for two land allocation scenarios in two consecutive years (2017-19). For each land allocation scenario in the same year, farmlets labeled with different lower-case letters are significantly different (P < 0.5). Between land allocation scenarios in the same year, same upper-case letters indicate lack of significant difference (P < 0.5). ............................................................................................................................................... 49	Figure 3.1 Daily precipitation and water available to leach during the non-production season of 2017-18. ........................................................................................................................................ 71	Figure 3.2 NO3- –N concentration in soil water (bottom) and cumulative potential NO3- –N leaching (top) of both crops during the non-production season of the 2017-18 crop year. Colors indicate different crop BMPs. Ribbon areas indicate standard errors (n=4). ................................ 72	xiv  Figure 3.3 2017-18 production season cumulative N2O–N emission (top) and daily N2O–N fluxes (bottom) from two crops of four farmlets. Ribbon areas indicate standard errors of each farmlet (n=4). ................................................................................................................................ 76	Figure 3.4 2018-19 production season cumulative N2O–N emission (top) and daily N2O–N fluxes (bottom) from two crops of four farmlets. Ribbon areas indicate standard errors of each farmlet (n=4). ................................................................................................................................ 77	Figure 3.5 Annual cumulative N2O–N emission (top) and daily N2O–N fluxes (bottom) from two crops of four farmlets. Ribbon areas indicate standard errors of each farmlet (n=4). .................. 84	 xv  List of Abbreviations   BMP Beneficial management practiceC CarbonCHU Corn heat unitDCD DicyandiamideDM Dry matterF FarmletLFV Lower Fraser ValleyN NitrogenN2 Nitrogen gasN2O Nitrous oxideNH3 AmmoniaNH4+ AmmoniumNO Nitric oxideNO3- NitrateP PhosphorusPAN Plant available nitrogenSWC Soil water contentSLF Separated liquid fractionTAN Total ammonium nitrogenTN Total nitrogenTNR Total nitrogen recoveryWFPS Water filled pore spaceWS Whole slurryxvi  Acknowledgments I would like to start with acknowledging my supervisors, Dr. Shabtai Bittman, and Dr. Sean Smukler, for granting me the opportunity to carry out this project and tirelessly guiding me through every step of the process. Without their consistent support and input, I would not be able to become the person I am today. Particular thanks to Dr. Shabtai Bittman for allowing me to work with him and the forage group for the past 5 years, which completely changed my life and will continue to inspire me in my career, and Dr. Sean Smukler for always keeping me on the right path with thoughtful feedback and incredible patience.   This research was supported by Agriculture and Agri-food Canada and conducted in Agassiz Research and Development Center. Without these resources, I would not be able to complete this study. I would like to thank the staff of Agassiz Research and Development Center, especially the forage group. Derek Hunt, Frederic Bounaix, Anthony Friesen, Xiao Wu, Pierre Groenenboom, Dean Babuin, and the field crew provided impeccable technical support and vital insight into my research. Many thanks to the amazing coop students as well, who helped with data collection and intense fieldwork, rain or shine.  I also want to thank my lab mates in the Sustainable Agriculture Landscape Lab and everyone at the UBC Dairy Education and Research Center. Their friendship balanced my life during these two years.   Special thanks are owed to my parents, who have supported me unconditionally throughout my years of education, and my girlfriend Carrie for her trust, patience, and willingness to endure my struggle. 1  Chapter 1:   Introduction Due to nutrient use inefficiencies, dairy farms, including their crop and animal husbandry components, have become a significant source of excess nitrogen (N) and phosphorous (P) in the environment. In particular, over-application of manure and fertilizer has been associated with nitrate (NO3-) contamination of groundwater, air pollution of ammonia (NH3) emission, and greenhouse gas (GHG) emission of nitrous oxide (N2O) and the eutrophication of water bodies in many countries including Canada (Sharpley et al. 2009; Powell et al. 2010; VandeHaar and St-Pierre 2010). The Lower Fraser Valley region of British Columbia is an important example of nutrient over application as a result of its density of animal production and relatively small agricultural land base (Bittman et al. 2017b). There is a clear need in the region to mitigate this nutrient surplus which represents both an economic loss given the fertilizer value of the nutrients and a threat to the regional ecosystem and drinking water. As one of the biggest importers of nutrients and economic drivers in the region, dairy farms, in particular, need improved measures for reducing surplus N and P that also ensure their productivity and economic viability.   Together, the import of nutrients in the forms of feed, fertilizers, bedding, and livestock to the dairy farm system likely has led to nutrient surpluses that can result in reduced nutrient use efficiency and profitability of dairy production (Powell et al. 2010; VandeHaar and St-Pierre 2010). Due to the limited land availability in the region, the two main crops on dairy farms, silage corn (Zea mays L) and perennial grass, e.g. tall fescue (Festuca arundinacea Schreb) and orchardgrass (Dactylis glomerata), are managed very intensively. Manure is typically applied to these crops at high rates to deal with the large quantities produced from an animal density that averages about 2.5 milking cows per ha across the region. These crops also receive additional 2  nutrients as mineral fertilizers to ensure high production rates; this includes starter P and N and an additional summertime application of N. The nutrient rates currently being used far exceed what the dairy crops are capable of taking up effectively and the excess is likely lost to the environment. There is an urgent need for dairy farm operations to address this inefficiency. To minimize dairy farm excess nutrient losses to improve nutrient use efficiency, improvements to forage production systems require a two-pronged approach. Dairy farmers must reduce nutrient losses or leakages while at the same time reduce nutrient surpluses by improving nutrient recovery. Together these strategies could lead to reducing the import of additional nutrients. Reducing dairy farm losses and leakages can be addressed through the adoption of beneficial management practices (BMPs) while reducing nutrient surpluses can be achieved by producing more feed on-site to efficiently to displace imported feedstuffs like grain and alfalfa. Researchers have developed BMPs to mitigate losses of N and P to air (e.g. low emission manure applicators) and to surface groundwater (e.g. shallow injection, cover crops). But the efficacy of these measures is constrained by the efficiency of nutrient recovery of the crop which is often determined by crop type, variety and a number of management factors.  Given that BMPs have not been optimized for both nutrient retention and recovery, farmer continues to use their current management practices to assure high levels of production per animal and per land unit which results in nutrient surpluses due to feed and fertilizer imports. Forage production involves a complex interaction between soil, nutrient, and plant. Research to mitigate nutrient losses is, however, often limited to a single nutrient and a single loss pathway and fails to consider the unintended consequence that conserving nutrients may actually add to surpluses. Furthermore, BMP research often does not account for the economic consequences of adopting practices that may require substantive investments in equipment or 3  require additional labor. Producing more feed requires challenging decisions such as the relative allocation of land between grass and corn. This allocation must consider the nutritional and economic value of the crops for milk production and the threats that each poses to the environment. For example, the leakier of the crops, corn, has no winter cover or roots but is an important source of energy that if otherwise supplied by grass would lead to excess protein in feed and excess N in manure. Thus, a holistic approach is needed (1) to evaluate the efficacy of integrated BMPs; (2) to address the potential tradeoffs between N loss pathways; (3) to explore the potential benefit of adjusting crop area allocation; and ultimately (4) to assess the nutrient use efficiency and economic performance of dairy farms at a whole-farm scale (Aarts et al. 2000). 1.1 Nitrogen dynamics in a forage production system  Among all major nutrients needed for agriculture production, N is the most abundant element in nature, but least readily available. Only 1 % of all N compounds in Earth’s atmosphere and the biosphere is reactive N (Nr) and it requires large quantities of energy to biologically or chemically convert non-reactive atmospheric N (N2) to Nr (Galloway et al. 2003). Reactive N compounds exist in many different forms and are constantly being transformed through biological and chemical reactions. Common Nr compounds involved in forage production are ammonia/ammonium (NH3/ NH4+), nitrate (NO3-), nitrogen oxides (NOX), nitrous oxide (N2O), and organic N (e.g. urea and proteins). 4  1.1.1 Transformation of reactive nitrogen Reactive N enters the forage production system in mineral form (i.e. mineral N fertilizer) and/or organic form (i.e. manure). Mineral N, as NH4+ and NO3-, are the primary forms of N for plant uptake and, when both available, NO3- is the predominant form of N absorbed by the plant (Figure 1.1). A wide range of soil organisms, mostly autotrophic bacteria, can convert NH4+ into NO3- via nitrification under aerobic conditions, which is a very important process to enhance the N availability for crop uptake. The rate of this process is governed by the supply of ammonia, e.g. mineral N fertilizers and manure, and aeration. However, when soil nitrate is not taken up by crop roots, NO3- can be reduced to N gases NO, N2O, and N2 by heterotrophic bacteria under anaerobic condition (Robertson and Groffman 2015). Organic N is generally not available for plant uptake until NH4+ is released through mineralization, which is also mediated by microbial activities. Ammonium is a volatile compound which can be easily lost to the atmosphere as NH3 gaseous emission.  1.1.2 Nitrogen field-loss pathways Volatilization of NH4+ is the largest pathway of N field-loss in dairy production of Canada (48 kt NH3−N yr-1), accounting more than for 60 % of total N field-loss from the dairy sector and 13 % of NH3 emission in Canada in 2005 (Bittman et al. 2017a).  Primary sources of NH3−N emission are the application of ammonium-based mineral fertilizers, land spreading of manure, and animal housing. The emission factor for NH3 varies based on the characteristics of manure (e.g. pH and water content), soil properties, weather, and application method. Minimizing the exposure of fertilizer and manure to air flow and warm temperatures is the principle way of mitigating NH4+ −N volatilization loss.  5  Nitrate leaching is another major N field-loss pathway of dairy forage production system. As NO3-−N is water soluble it can move downward easily through a saturated soil profile.  Draining water carries the NO3-−N until it reaches an impermeable layer or is conveyed to a water body causing contamination (Zebarth et al. 1998). Mitchell et al. (2003) identified manure fertilization as the regional source of elevated NO3- concentration in the Abbotsford-Sumas aquifer that sits across the international boundary between southern BC and northwestern Washington state. The potential of NO3-−N leaching is governed by soil physical characteristics, weather condition, nutrient loading, and crop uptake rate (Magdoff 1991). Many studies focused on forage crop production (Guillard et al. 1999; Sogbedji et al. 2000; Aparicio et al. 2008) have concluded that greater NO3-−N leaching losses occur when N application rates increase, because most forage crops recover less than 50 % of applied N.  Nitrogen unused by the crop, considered residual N, is extremely susceptible to leaching in regions with high wintertime rainfall and is the major source of leached NO3-−N. Soil with coarser texture is more vulnerable to NO3-−N leaching because of poor water holding capacity (Sogbedji et al. 2000; van Es et al. 2004). A number of studies (van Es et al., 2004; Shepherd and Newell-Price 2016) also found that the timing and frequency of manure application have a critical effect on NO3-−N leaching. Lack of synchronicity between mineralization from organic N fraction of manure and crop uptake would increase leaching risk of NO3-−N because accumulated mineralizable N continues to supply NH4+−N for nitrification when crops are removed during postharvest seasons (Paul and Zebarth 1996). For the same reason, Shepherd & Newell-price (2016) recommended avoiding annual manure application to the same field, which adds more organic matter to the pool of mineralizable N and increase the risk of NO3-−N leaching. In terms of NO3-−N leaching potential 6  under different forage crops, a number of studies (van Es et al. 2004; Eriksen et al. 2010; Hooijboer et al. 2010) have found that NO3-−N concentration in soil leachate were much higher from silage corn production than perennial grass.  Nitrate-N can also be lost as N2 gaseous emission into the atmosphere when denitrification occurs, with intermediary products as nitrite (NO2-), nitric oxide (NO) and N2O (Misselbrook et al. 2013). Nitrous oxide is a very potent greenhouse gas (GHG) and, according to a survey of 295 dairy farms in Canada (Alemu et al. 2017), nitrous oxide accounted for 24% (129.5 Mg CO2e) of the total GHG emission of a dairy farm, of which 29.7 Mg CO2e was emitted from the soil system and 99.8 Mg CO2e was from manure storage.  Unlike nitrification, denitrifying bacteria favor anaerobic condition to reduce NO3- in the soil while consuming organic carbon (C) as an energy source. Thus, the concentration of NO3- and abundance of C control the intensity of denitrification. To a lesser extent, aeration, temperature, and moisture of soil also have an impact on the magnitude denitrification (Shepherd et al. 1991; Skiba et al. 1993). Favorable conditions for denitrification are common in the forage production system of a dairy farm when intense rainfall causes temporary anaerobic conditions in saturated soils that were treated with N fertilizer. Given the same rate of N application to soils, using dairy slurry manure often causes greater denitrification loss and N2O emission than mineral fertilizer, because slurry manure contains additional supply of organic C and moisture, both are factors that sustain and intensify denitrifying microbial activity (Paul and Beauchamp 1989; Clayton et al. 1997). Although nitrification and denitrification both generate N2O, denitrification becomes the dominant source process for N2O production when water filled pores space (WFPS) exceeds 60 % (Linn and Doran 1984; Lemke et al. 1998b). If soil moisture further increases until WFPS reach 80 %, N2O emission starts to drop and N2 becomes the dominant end 7  product of denitrification (Veldkamp et al. 1998). Moreover, Lemke et al. (1998a) found that up to 70 % of total N2O emission occurs during snowmelt in response to freeze-thaw perturbation of soil. Several studies have suggested that the freeze-thaw release of N2O is produced by biological denitrification in thawing soil surface when freeze-killed microorganisms, frost damage of cover crop, and destruction of soil aggregates release decomposable organic C and mineral N (Christensen 1991; Ludwig et al. 2004; Mørkved et al. 2006; Tenuta and Sparling 2011).  Figure 1.1 Nitrogen cycle of a dairy farm 1.1.3 Dynamics of soil phosphorus  Soil amendments add both organic and inorganic forms of phosphorus (P) into the soil, but only soluble and labile P is available for crop uptake, i.e. biologically active, which is only a small portion soil P pool. Bioavailability of soil P in is largely controlled by the abundance of soil organic matter, soil pH, cation exchange capacity, and amount of amorphous and crystalline 8  complexes of Al, Fe, and Ca (Stewart and Tiessen 1987; Sharpley 1995). Transformation and release of plant available P, e.g. phosphate ions (H2PO4- and HPO42-), is very slow and ionic forms of P are susceptible to adsorption and subsequent precipitation as less soluble secondary P minerals either by Fe, Al, and Mn in acidic soil solution or by Ca compounds in alkaline soil solution (Stewart and Tiessen, 1987; Sharpley, 1995). Such fractionation processes will remove plant available P from soil P pool and reduce P availability. Erosion of P-carrying soil particle and runoff of water with dissolved P are two major field pathways of P loss from the soil. Natural weathering of primary P-rich mineral, stable soil pH, and abundant soil microbial activities, e.g. arbuscular mycorrhizae, are some well-studied factors that would enhance plant available P in soil. 1.2 Nutrient and crop management of the dairy farm in the Lower Fraser Valley, BC The Lower Fraser Valley (LFV) of British Columbia is one of the most intensively cultivated areas in Canada. Although only 1.6 % of the province’s agricultural land, the LFV generates more than 60 % of BC’s gross farm receipt and hosts 54 % of all the dairy operation in the province (Fraser Valley Regional District 2017). Two major types of forage crop in the LFV are silage corn and perennial grass, e.g. tall fescue and orchard grass. Silage corn is often harvested for grain or making whole plant silage, which contains higher dietary fiber and metabolizable energy comparing to grass.  Grass is typically harvested for hay and silage or sometimes used for grazing when land limitations are not an issue. The high cost of feed and fertilizer has been a challenge to a dairy operation in this area, which created an important economic incentive for farmers to produce feed with farm-grown forage (grass silage, corn silage, grains, etc) and to use farm-generated manure as a major nutrient source for crops. Compared to mineral fertilizer, manure, as a soil amendment, contains 9  more organic matter (OM) and organic N, which is beneficial for building up soil N stock and improving soil structure. However, the organic N in manure is not immediately available for plant uptake and the readily decomposable organic matter provides C and energy for heterotrophic denitrification, which can increase N2O emission (Petersen and Sommer 2011). Particulate OM and fibers can also increase the viscosity of slurry manure, which can retain NH4-N at the soil surface and stimulate NH3 emission (Petersen & Sommer, 2011). Thus, if not managed properly, the storage and application of slurry manure can cause significant N field-loss. In addition, dairy manure produced in LFV usually has a low N/P ratio, i.e. higher P content per unit of N than what plant growth requires. When manure is applied based on the crop demand for N, excessive mineralized P accumulates in the soil as insoluble compounds, which is not available for plant uptake but is prone to lose through surface runoff and soil erosion (Kleinman et al. 2002). Developing BMPs that enhance nutrient retention and recovery would likely not only reduce the environmental impacts but also increase net returns to the farmer. 1.2.1 Beneficial management practices for nutrients Most dairy manure in the Fraser Valley area is collected in slurry form and applied by broadcasting directly onto a crop field. Broadcasting slurry manure with a splash plate or similar equipment is the most widely used field application method in Canada (Webb et al. 2010) because of the simple operation and relatively low cost. However, studies have shown that surface broadcasting slurry is prone to substantial NH3 volatilization due to increased exposure to air and reduced infiltration to the soil, especially when slurry has high solid content (Sommer and Hutchings 2001). Loss of plant available N and uneven application in a windy condition often cause inconsistent response from crops (Bittman et al. 1999). Rapid incorporation of 10  manure into the soil and precise application near the root zone can potentially reduce volatilization loss and improve crop response (Bittman et al., 1999). Several widely used innovative manure application techniques include surface banding slurry with trailing shoe/hoes (TS/TH), open-slot injection (OSI), and rapid incorporation of the slurry by cultivation (Webb et al., 2009). TS and TH place slurry direct to soil surface beneath grass canopy, which minimizes the exposure to air and allows longer time for infiltration. OSI implement cuts trenches of at least 5 cm deep into the soil and injects separated manure. Webb et al. (2009) concluded that TS/TH and OSI are most effective in reducing NH3 emission when slurry manure is applied to grassland with an NH3 abatement rate between 45 and 90 %. When applying slurry to arable land, immediate incorporation of manure by the plow, disc or tine, is the most effective NH3 abatement technique (Webb et al., 2009).  Slurry manure contains 90-95 % water and most P and OM are concentrated in solids. Thus, removing solid fraction from a slurry can potentially increase N:P ratio of manure to avoid over-application of P. The dual-stream manure separation technique (Bittman et al. 2009) is often used to separate the whole manure slurry mechanically (e.g. screening and settling) or biologically (e.g. anaerobic digestion) into manure sludge and separated liquid fraction (SLF) to promote infiltration of liquid fraction. A number of studies have shown that application of SLF could mitigate NH3 volatilization and improve manure N uptake (Bhandral et al., 2009; Carter et al., 2010; Angers et al. 2007; and Zhou et al., 2009). Bittman et al. (2011) reported that separated liquid fraction of slurry (SLF) achieved higher grass yield and higher N recovery rate than the whole slurry (WS) at an equivalent moderate rate (300 kg ha-1) of total ammonium-N (TAN) while having greater application volumes and lower P loading. Kai et al. (2008) found that slurry acidification is also effective for mitigating NH3 emission from field application of liquid 11  manure. When injected close to corn roots, separated manure sludge can replace starter P fertilizer to boost early growth of corn (Bittman et al. 2012). Nitrification inhibitors (NI) are compounds, either naturally occurring or synthetic, that allow mineralization of organic N to produce NH4+ but temporarily inhibit the first stage of nitrification, the oxidation of NH4+ to NO2-, by interrupting the respiratory enzymes of Nitrosomonas bacteria (Amberger 1989). In this way, N is preserved in NH4+ form, which is less subject to loss from the soil. Most widely known synthetic NIs are nitrapyrin, dicyandiamide (DCD) and 3,4-Dimethylpyrazole phosphate (DMPP). They have been widely used in agricultural practice to control the transformation of plant available N (PAN) in soil and to minimize N loss by the leaching of NO3-−N (Erhardt et al. 2003; Di and Cameron 2005; Asing et al. 2008). Extensive studies by Cameron and Di (2002, 2003; 2005; Di et al. 2007) have concluded that DCD was highly effective in decreasing N2O−N emission factor and NO3-−N leaching loss from pasture applied with animal urine and repeated annual use of DCD does not alter its effectiveness (De Klein et al. 2011).  1.2.2 Beneficial management practices for cropping Hybrid of silage corn varies by the corn heat unit (CHU). Long-season corn requires higher CHU (2600 ~ 2800) to mature, i.e. late harvest, and has relatively taller and heavier stovers, which would yield higher dry-matter (DM) biomass. Short-season corn requires less CHU (2100 ~ 2300) to mature, resulting in earlier harvest, and has relatively thinner stoverS, but higher harvest index, i.e. greater grain content. Farmers generally choose corn hybrid varieties to meet their local climate and feed demand. After corn harvest in the fall, the soil surface is often left fallow or seeded with cover crops, such as winter wheat, clover, fall rye, or a mixture of these to prevent soil erosion from strong winter winds and nutrient leaching from heavy rainfall 12  (Langdale et al. 1991; Dabney et al. 2001, 2010). A number of studies have concluded that winter cover crops also have positive impacts on soil aggregate stability, soil organic C, and organic N (Hermawan and Bomke 1997; Liu et al. 2005). When used as green manure, cover crops have the potential to increase crop yield of the succeeding production season (Odhiambo and Bomke 2001). However, the cost of seeding cover crop and challenges of establishing cover crop in fall are known obstacles for their adoption. Franzluebbers (2007)  discussed the potential of integrating cover crop production into livestock systems, with emphasis on increasing yield of succeeding crop and reducing additional N fertilizer, which would subsequently create an economic incentive for farmers to implement cover crop. An alternative for local farmers in the LFV is relay cropping, which is the technique to inter-seed a resilient cover crop with high nutrient values, often Italian ryegrass, between young corn plants in the summer. The ryegrass will have a longer time to germinate and slowly grow under the canopy of corn plants during the production season. When corn is harvested in early fall, Italian ryegrass is already established and the growth resumes throughout the wet but temperate winter of southern coastal BC. In the spring, Italian ryegrass can yield 3 – 5 t DM ha-1 of biomass, capture 50 – 65 kg ha-1 residual N in the soil and also be used as an excellent feed for cows to offset the cost of seed and operation (Bittman et al. 2004). van Vliet and Zebarth (2004) found that using Italian ryegrass as relay crop between silage corns can effectively reduce nutrient loading, e.g. total N, P, K, and Cu, in runoff post corn harvest.  In contrast to silage corn, perennial forage grass contains higher protein as feed, has a longer production season, and covers the land all year long, but it requires a much higher N input. Typically, dairy farms in the LFV harvest grass 4 or 5 times annually with 6 weeks in between each cut, starts in April and ends in early October. Manure is applied first in March as 13  soon as the growth starts and then after each cut throughout the growing season. A number of studies have shown under various management practice and climatic conditions that harvest frequency has a great impact on productivity and nutritive value of forage grass (Van Man and Wiktorsson 2003; De Santis et al. 2004; Turner et al. 2006; Vinther 2006; Pontes et al. 2007). Despite the environmental condition and intrinsic property of different grass species, herbage production of forage grass is inversely related to harvest frequency due to the changes to plant morphological development (Binnie and Chestnutt 1991; Vinther 2006). As grass matures, leaf area increases, but the leaf:stem ratio (by weight) decreases. A longer regrowth interval allows grass to reach peak growth of the sigmoid growth curve through more photosynthesis and elongation of the stem, which increases DM biomass and metabolizable energy but decreases digestibility, and crude protein (CP) concentration. If the grass is harvested more frequently, i.e. short regrowth interval, leaf:stem ratio will increase as plants stay at a vegetative stage for a longer period. As a result, grass DM biomass will decrease, while digestibility and CP concentration increase. However, the increase of CP concentration is largely due to the decrease of DM accumulation rather than higher N uptake. Hence, reducing harvest frequency can produce more feed and potentially reduce the demand for imported feed and energy supplement.  1.3 Research objectives and hypotheses  The overall objective of my thesis is to examine progressive management options designed to mitigate nutrient losses, while avoiding unintended environmental consequences, reducing farm nutrient surpluses and maintaining the productivity of dairy farming systems. To study these combined effects, integrated systems are needed. Aarts et al (2000) implemented a prototype dairy farm system on an experimental farm, “De Marke”, which characterized the whole-farm approach of accounting for N budget of dairy farm systems. However, this prototype 14  system cannot be replicated effectively in a way that would enable the comparison to other theoretical systems.  To address this challenge, an alternative farmlet design without animals was implemented to represent the prototype system in a way that could be easily manipulated. In this project, I examined four dairy farm management scenarios in a miniaturized integrated farming system or famlet (F): Conventional farmlet (F1): representing a baseline farm with conventional nutrient and cropping practices; Innovative nutrient management farmlet (F2): the same production system as F1 but includes innovative BMPs designed to improve nutrient dynamics; Innovative cropping farmlet (F3): innovative crop production BMPs are integrated with those that optimize nutrients; and Advanced technique farmlet (F4): integrates both nutrient and crop BMPs with various advanced techniques to reduce loss pathways and improve feed production opportunities. To evaluate the crop N recovery and N field-losses (NO3-−N and N2O−N) of these four scenarios, I developed, in collaboration with Agriculture and Agri-Food Canada scientists, an experiment at the Agassiz Research and Development Centre.  In this experiment, we setup replicates of farmlets that represented each of these four scenarios in a completely randomized block design. This allowed me to compare the N recovery and field-losses across four systems and also between two types of forage crops. Using this field experiment I conducted two distinct studies that I present in chapters 2 and 3 of this thesis. 1.3.1 Study 1: Crop N removal and total N recovery of experimental farmlets under different nutrient and cropping management scenarios The objectives of the study I present in chapter 2 were to (1) compare the crop N removal and the total N recovery (TNR) of four management scenarios that incrementally introduce innovative nutrient and crop BMPs; and (2) assess how different amounts of land allocated to either corn or grass production would impact these metrics. I tested the following hypotheses:  15  H1: Crop N removal and TN recovery will increase incrementally as the management scenario becomes more advanced and complex.  H2: Reallocation of 10 % of farmland from grass to plant silage corn will improve total crop N removal and TN recovery of a farm when innovative BMPs are implemented. 1.3.2 Study 2: Evaluating the efficacy of integrated beneficial management practices on reducing potential nitrate leaching and nitrous oxide emission from dairy farm forage production systems In this study, presented in chapter 3, my objectives were to (1) to identify the scenario that achieved the lowest N2O−N emission and potential NO3-−N leaching intensity, i.e. g N kg-1 DM feed production; (2) to determine whether adjusting crop area ratios between silage corn and grass would affect N2O−N emission and potential NO3-−N leaching intensity of a farm. I tested the following hypotheses: H1: N2O−N emission and potential NO3-−N leaching intensity will decrease incrementally as the management scenario becomes more advanced and complex.  H2: Reallocation of 10 % of farmland from grass to plant silage corn will decrease N2O−N emission and potential NO3-−N leaching intensity of a farm when innovative BMPs are implemented.   16 Chapter 2: Crop N Removal and Total N Recovery of Experimental Farmlets under Different Nutrient and Cropping Management Scenarios 2.1 Introduction Dairy farms in the Lower Fraser Valley are highly motivated to recover nutrients, particularly nitrogen (N) and phosphorus (P), from farm-generated manure though forage production for both economic and environmental reasons. However, farmers are facing major challenges to maintaining a high level of production while minimizing the environmental impacts of manure application. Their primary challenge is accurately matching the nutrients supplied by manure with the need for crops. A number of beneficial management practices (BMPs), for both nutrient applications and cropping, have been developed to improve crop nutrient uptake and reduce nutrient loss to the surrounding environment. One example of a well-developed BMP is the separation of dairy manure solid and liquid fractions, i.e. the dual-stream manure separation technique. Most dairy manure in the Fraser Valley area is collected in slurry form, which contains 90 – 95 % water. Less soluble nutrient components, such as organic N and plant available P, are concentrated in the solid faction. In the liquid fraction of manure slurry, ammonium N (NH4+–N) is the major plant available N (PAN) and it is prone to lose through volatilization. Bittman et al. (2011) found that removing solids can improve N utilization in tall fescue production by promoting infiltration of separated liquid manure, which reduces ammonia (NH3) volatilization and improves crop N uptake (Angers et al. 2007; Bhandral et al. 2009; Carter et al. 2010). Removal of P-enriched manure solids can also prevent over-application of P on grass plots (Bittman et al., 2011; Bhandral et al., 2009). Manure incorporation techniques, such as banding with trailing shoe and manure injection, can potentially limit NH3 loss relative to broadcasting by reducing contact with moving air. Bhandral   17 et al. (2009) concluded that trailing shoes, by placing narrow bands of separated liquid manure directly on the soil beneath the grass canopy can reduce the surface area of manure applied to the soil and minimize contact of manure with plant material. In winter, prolonged heavy rainfall in the Fraser Valley can cause soil erosion, nutrient runoff and leaching of residual soil nitrate (NO3-) into groundwater or nearby water bodies. A number of studies have shown that adopting the BMP of Intercropping of Italian ryegrass on corn fields when corn plants are established could substantially increase N uptake comparing to corn monocrop systems (Zhou et al. 2000) and also significantly reduce N loading in soil of post-harvest corn field (van Vliet and Zebarth, 2004). However, the synergetic effect on crop N uptake by implementing both nutrient and crop BMPs together remains unknown. While these individual BMPs have been shown to be effective, there is little evidence that when used in combination their effect is additive. In fact, there is evidence that some combinations of BMPs may actually lead to adverse outcomes. If a combination of BMPs was incrementally improved we would expect that their beneficial outcomes together would be additive. This incremental improvement may also be dependent on the relative distribution of crop area. The objectives of this study are to (1) compare the crop N removal and the percentage of total N recovery (TNR) of four management scenarios that incrementally introduce innovative nutrient and crop BMPs; and (2) assess how different amounts of land allocated to either corn or grass production would impact these metrics. My hypothesis is that crop N removal and TNR will increase incrementally as the management scenario becomes more advanced and complex. My second hypothesis is that reallocation of 10% of farmland from grass to plant silage corn will improve total crop N removal and TNR of a farm when innovative BMPs are implemented.   18 2.2 Material and Methods 2.2.1 Site description and field trial setup Field trial of this study was conducted at Agassiz Research and Development Center (ARDC) in Agassiz, British Columbia, Canada (49°10′ N, 125°15′ W). The Agassiz area is a confined valley rural area located in the eastern portion of Lower Fraser Valley (about 130 km east of Vancouver), which is known for intense agricultural activities. Forage corn and grass, small fruits, and livestock (poultry and dairy) have occupied the majority of agricultural land in this area since the 1890s.  The experiment site occupies a topographically flat area (slope gradient < 2%, 20 meters above sea level). Gentle ridge-and-swale topography is evident at the site and is consistent with the alluvial parent material. A creek runs along the southeast side of the field. The site is underlain by a continuous unconfined aquifer that is comprised of Fraser River alluvial permeable sands and gravel. The Climate of the Agassiz area is a predominantly cool and humid inshore maritime climate, characterized by mild temperatures, 3.4 °C in January to 18.7 °C in August, and average monthly precipitation ranging from 58 mm in August to 285 mm in November based on 30 years (1981-2010) average from an Environment and Climate Change Canada (ECCC) weather station (Agassiz CDA) 2 km away (Environment and Climate Change Canada, 2018). About two-thirds of the yearly precipitation occurs during the winter months of October through March, which can lead to drainage problems. The region receives 67.4 cm of snow annually on average (Environment and Climate Change Canada, 2018). The water table rises seasonally under the influence of prolonged rainfall in Winter and snowmelt in late Spring. Conversely, soil moisture   19 deficiencies may occur during the summer months when rainfall is minimal (Luttmerding and Sprout 1967). The soil at the experiment site is classified as Melanic Brunisol by the Canadian System of Soil Classification (Soil Classification Working Group 1998). The silty to sandy loam soil texture was derived from stone-free, unconsolidated floodplain deposits of Fraser River and smaller streams (Luttmerding 1981). The soils are moderately previous, well to moderately well drained with moderate to high water and nutrient holding capacity. Root growth and water movement are unrestricted to at least 80 cm depth. Long-term cultivation has increased nutrient content in the surface mineral layers through the incorporation of soil amendment and decomposition of plant residues. 2.2.2 Experimental Design I designed this study with the forage research group at the ARDC to include four incrementally improved and more complex management scenarios for dairy forage production system that I term: conventional (F1), innovative nutrient (F2), innovative cropping (F3), and advanced techniques (F4). Each management scenario is a treatment that integrated different nutrient and cropping beneficial management practices (BMPs). Details of each treatment will be discussed in the next subsection and can also be found in Appendix A.   We developed the concept of “farmlet” to define the physical representation of a dual-crop dairy forage production system when livestock animals are absent. A farmlet is made of a pair of silage corn and grass plots that are implemented with the same management scenario, i.e. treatment. For example, Farmlet 1 consists of a pair of corn plot and grass plot implemented with conventional management scenario. The size of these plots was made to be large enough to be   20 compatible with operational farm-size machinery, such as tractors, manure applicators, and grass harvesters.  In Fall 2016, we established the field trial on a 90 m x 120 m field. This field was a 5 years old stand of tall fescue seeded in 2011 at the rate of 30 kg ha-1. Before trial setup, half of the grass field was plowed and sprayed to prepare for seeding silage corn. We first set up 4 blocks using a randomized complete block design where each block was divided into corn (Zea mays L.) and grass (Tall Fescue, Festuca arundinacea Schreb.) plot of the same size. Within each corn and grass plot, the field was further split into 4 subplots of 6.1 m x 18.3 m. Aforementioned four treatments were randomly allocated on 4 subplots of each crop, i.e. four corn plots and four grass plots in each block (Figure 2.1). On each corn subplot, we seeded eight rows of corn in May in 2017 and again in 2018 with 76.2 cm between each row to reach a plant population of 82,000 to 84,000 per hectare.  Figure 2.1 Field trial setup. Randomized complete block design of 4 blocks with split-plot by type of crop and 4 management scenarios, i.e. treatments, were randomly assigned to 4 subplots of each crop. 2.2.3 Forage production system management scenarios The conventional management scenario (F1) was designed to operate the same way as a typical conventional dairy farm in the Lower Fraser Valley area. Typically, nutrient   21 management practice of corn plots starts with manure application in May. Manure application rates were determined based on the target of 32 kg P ha-1. Whole manure slurry (WS) was agitated and applied on corn plots on May 4th, 2017 and May 2nd, 2018 (Table 2.1). Typical application method in the Lower Fraser Valley region is to broadcast whole manure slurry with double splash plates. Due to lack of a splash plate, whole manure was applied by lifting a 3.05 m wide trailing shoe implement 0.5 m above ground. We broadcasted 119 kg NH4+−N ha-1 on average over two years (Table 2.2).  After leaving the manure to infiltrate for one week, we seeded corn while applying 20 kg ha-1 of NH4+−N and 40 kg ha-1 P in the form of monoammonium phosphate (MAP, 11-52-0) (Table 2.2), i.e. the starter fertilizer, at 5-cm depth and 5-cm away from corn seeds. When corn plants reached the 6-leaf stage, we also side-dressed 55 kg NH4+−N ha-1 in the form of urea (46-0-0) in June of both years (Table 2.2) based on the result of Pre-sidedressing nitrate test (PSNT) (Magdoff, 1991). In order to perform PSNT, we randomly collected 20 cores of 0-15 cm deep soil samples from each corn plot (6.1 m x 18.3 m) to form one composite sample when corn plants reach 6-leaf phase (Appendix B). These soil samples were then sent to a commercial lab (Terralink®, Abbotsford, BC, Canada) to test for NO3-−N concentrations. We side-dressed urea with a customized 4-row corn sidedresser made by FABRO® Ltd. (Swift Current, SK, Canada). For grass plots, the first manure application took place in late March to early April. Annual manure rate was determined by the type of manure and the NH4+−N level of our manure supply, which varies from year to year. Based on the knowledge of previous trials, we set a target of 536 kg ha-1 yr-1 of total N for grass plots, when NH4+−N consists of 50% of total N of whole manure slurry. The target of 268 kg NH4+−N ha-1 yr-1were broadcasted on grass plots in five   22 applications throughout the growing season. The first application (30% of the annual total N) was applied in March and then followed with four more manure applications after each one of the first four grass harvests.  In terms of crop management, we also used typical management practices for the region. We seeded a commonly used late harvest variety of corn (CHU 2750), which was seeded on the first week of May and harvested on last week of September (Table 2.1). We harvested the grass 5 times each year. The first cut happened in May when flowers just started to emerge and then each following cut took place 5 weeks after the previous one for four more times throughout the growing season, from the first week of May to the second week of October (Table 2.1). In the innovative nutrient management scenario (F2), we kept the same conventional cropping practice as F1 but replaced conventional nutrient practice with innovative nutrient management BMPs. The core of these BMPs is the dual-manure system. With the assistance of a commercial dairy farm, we separated dairy manure slurry gravitationally into a thinner liquid portion and thicker sludge portion, i.e. separated liquid fraction (SLF) and manure sludge. Instead of broadcasting WS on the soil surface and incorporating starter fertilizer while seeding corn, we injected manure sludge into 6-inch deep trenches using tines trialing hoes and then seeded corn 5-cm beside the manure trenches with the intention of replacing imported P fertilizers with the P enriched manure sludge.  On grass plots, instead of broadcasting whole manure slurry, we applied SLF with the trailing shoe. For the first manure application, 30% of targeted annual total N, was applied in March and the rest of the separated manure was applied in four applications after each one of the first four grass harvests in equal portions. Regardless of the different type of manure and application method, the targeted total N application rate, the proportion applied in each manure   23 application, and date of manure application remained the same between the conventional (F1) and three advanced management scenarios (F2, F3, and F4) (Table 2.2).   For the innovative cropping management scenario (F3) we used the same nutrient management practices of F2 and then integrated several innovative cropping BMPs. The first BMP was seeding an early harvest variety of corn (2150 CHU) instead of seeding the 2750-CHU variety that was used in F1 and F2. The early season variety was intended to bring the harvest time approximately 2 weeks earlier to enable more effective relay cropping. The second BMP was to intercrop a relay crop of Italian ryegrass which was sown at 30 kg ha-1 between corn rows when corn plants reached the 6-leaf phase (Table 2.1). This BMP was designed so that when the corn is harvested in mid-September, removal of the competition will enhance ryegrass growth well into mid-November, which could potentially recover residual soil NO3- –N before prolonged winter precipitation starts. The ryegrass would then be harvested in following April for forage, before seeding corn, or even cut once in late-fall of the seeding year if there was enough growth before the soils were too saturated for heavy equipment (Table 2.1). With enough productivity, the relay crop could compensate for the potential corn yield penalty caused by competition (Bittman et al. 2004). Furthermore, the variety of Italian ryegrass was selected to stay vegetative for the year of seeding to ensure the production of high-quality dairy cow forage, e.g. high digestible energy, crude protein, and palatability. The third cropping BMP we incorporated was to harvest the grass three times throughout the growing season, instead of 5 times. Fewer grass cuts at a later maturity stage of production could potentially produce higher dry matter (DM) yield of feed with lower digestibility, which would potentially reduce the demand of imported feed and energy supplement (Bittman et al.   24 2013). The first harvest took place in late May when the grass fully flowered, and two more cuts took place after every 8 weeks (Table 2.1). We integrated several additional nutrients and cropping BMPs into F3 for the advanced technique scenario (F4), which is the most complex and improved scenario out of the four. The nutrient BMPs included adding a nitrification inhibitor in manure to sustain nutrient release and potentially reduce N2O emission (Cahalan et al., 2010; Di et al., 2007; and Ledgard et al., 2014); and releasing accumulated NH4+−N during dry season (early July to mid-September) through weekly irrigation.  For each manure application of F4, 20 L ha-1 of dicyandiamide (DCD) nitrification inhibitor was added into the manure tank and agitated before the manure was applied to both corn plots and grass plots. We irrigated all four blocks of grass plots of F4 weekly from July 19th to September 7th, in 2017. Due to the limitation of equipment and materials, we made two sets of 2.7 m x 18 .3 m, 9-line, dripping tubes to cover only the center sampling area of each plot. Each drip line has 60 holes each with a discharge rate of 2.3 L hr-1. Powered by a Honda WX10 pump, the 9-line dripping tubes could discharge a 1200 L water tank in 58 minutes at 20 psi, an equivalent of 24 mm of precipitation on the irrigated area. Weekly irrigation stopped when there was sufficient precipitation to compensate for daily moisture deficiency. In 2018, all four blocks of grass and corn plots were irrigated from July 17th to September 6th. New 3.05 m x 18.3 m, 8-line, dripping tubes were made for both grass plots and corn plots. We also dug a well near the field to supply water to the irrigation system and an equivalent of 35 mm precipitation was irrigated every week.  25 Table 2.1 Timeline of field activities, including date of manure application, crop harvest, and N2O–N and NO3- –N measurement period.  Activities 2017-18 2018-19GrassSpring	manure	application All	farmlets 21-Mar-17 13-Mar-18F2,	F3,	and	F4 4-Apr-17 3-Apr-18Spring	harvest F1	and	F2 9-May-17 7-May-18F3	and	F4 29-May-17 23-May-18Summer	manure	application F1	and	F2 15-May-17 14-May-18F3	and	F4 5-Jun-17 29-May-18F1	and	F2 27-Jun-17 20-Jun-18Summer	harvest F1	and	F2 22-Jun-17 15-Jun-18All	farmlets 1-Aug-17 23-Jul-18Fall	manure	application All	farmlets 8-Aug-17 25-Jul-18F1	and	F2 18-Sep-17 10-Sep-18Fall	harvest F1	and	F2 14-Sep-17 4-Sep-18All	farmlets 12-Oct-17 9-Oct-18N2O-N	measurement	period All	farmlets March	2017	to	March	2018 March	2018	to	March	2019NO3--N	measurement	period All	farmlets May	2017	to	July	2017 March	2018	to	July	2018October	2017	to	March	2018 October	2018	to	March	2019CornManure	application All	farmlets 4-May-17 1-May-18Starter	N	and	P	fertilizer F1 4-May-17 9-May-18Corn	seeding All	farmlets 15-May-17 9-May-18Side-dressed	Urea All	farmlets 21-Jun-17 19-Jun-18Corn	harvest F3	and	F4 11-Sep-17 12-Sep-18F1	and	F2 27-Sep-17 25-Sep-18N2O-N	measurement	period All	farmlets April	2017	to	April	2018 April	2018	to	April	2019NO3--N	measurement	period All	farmlets May	2017	to	July	2017 April	2018	to	July	2018October	2017	to	April	2018 October	2018	to	April	2019Relay	crop 	Italian	ryegrass	seeding F3	and	F4 21-Jun-17 19-Jun-18Manure	application F3	and	F4 12-Mar-18 20-Mar-19Harvest F3	and	F4 25-Apr-18 April	2019†Crop	YearManagement	scenario†	Relay	crop	harvest	date	has	not	determined.  26 Table 2.2 Nitrogen and phosphorus input of two years, including commercial fertilizer and manure.   2.2.4 Crop sampling and analysis I harvested and sampled silage corn twice a year to determine dry-matter (DM) yield on the second and fourth week of September depending on the variety of corn. In order to obtain a well representative sample of the plant population, the width of the sampling area was fixed at 1.52 m, the width of two corn rows, set on the center two rows of each plot where plant height and spacing were consistent, and the length of the sampling area was determined by the physiological condition of the plants. In 2017 this was 3.72 m2 (2.44 m x 1.52 m) and in 2018, 4.64 m2 (3.05 m x 1.52 m). I cut all corn plants at 10 cm in height. Plants were then counted and weighed fresh (Plot FW) before subsampling 10 plants. Subsampled plants were stripped for cobs and weighed fresh. The rest of the sampled corn plants were chopped whole with a shredder (shreddable diameter = 60 mm) (ELIET® Major 4S, Pittsburgh, PA, USA) and mixed on site to form a composite subsample, which was weighed before (SS FW) and after being oven-dried at 60 oC to a stable weight (SS DW) to determine percentage dry-matter (DM%). After sampling 2017 2018 2017 2018 2017 2018 2017 2018 2017 2018Silage	Corn kg	ha-1F1	 75 75 187 202 282 312 40 40 71 75F2 55 55 169 170 271 288 40 40 34 35F3 55 55 173 171 278 289 40 40 35 35F4 55 55 165 170 264 288 40 40 32 35Italian	RyegrassF3 - - 67 TBD¶ 124 TBD - - 15 TBDF4 - - 67 TBD 123 TBD - - 15 TBDTall	FescueF1	 - - 345 299 586 543 - - 98 85F2 - - 407 310 707 537 - - 96 64F3 - - 409 327 694 559 - - 93 67F4 - - 404 320 684 564 - - 91 68Management	scenarioTAN‡ TN§ TP§Fertilizer	N† Fertilizer	P††	Types	mineral	fertilizer	applied	were	11-52-0	(starter)	and	46-0-0	(sidedress)‡	TAN:	total	ammonium	nitrogen	from	both	mineral	fertilizer	and	manure.¶	Relay	crop	N	and	P	input	is	to	be	determined,	because	crop	year	2018-19	has	not	complete	yet.	§	TN:	total	nitrogen	from	both	mineral	fertilizer	and	manure.TP:	total	phosphorus	from	both	mineral	fertilizer	and	manure.  27 for yield, corn plants from the non-sampled area were harvested with standard commercial equipment. In the grass plots, I harvested two strips of tall fescue between the tire tracks from each to determine herbage DM yield. Tall fescue was harvested at 5- to 7-cm height using a plot forage harvester (Wintersteiger® Cibus F, Austria). Herbage from the entire sampling area was weighed fresh immediately by the scale equipped on the harvester (Plot FW). I took one subsample (~1000 g, in two perforated bags) from the harvested grass of each plot and weighed fresh (SS FW) and reweighed after oven-drying at 60 oC to a stable weight (SS DW) to determine DM%. After sampling for yield, grass from the non-sampled area was mowed with standard commercial equipment to the same height so that herbage was removed uniformly from the entire treatment areas. The herbage of relay crop, i.e. Italian ryegrass, was harvested and sampled the same way as tall fescue.  DM% of crop samples were calculated as, Equation 2.1 !"% = %%	!'	())%%	+'	()) ×-..% Crop DM yield calculated as, Equation 2.2 /012		34567	 819	!"	:;<- = =618	+'	(>))×!"%%;?2649)	;05;	(?@)	×-.A	?@-	:; × -	819-.B	>) Cob samples were weighed again after oven-dried at 60 oC to a stable weight and then shelled and weighed for grain dry weight (Grain DW). Corn grain DM was calculated as,    28 Equation 2.3 /109	)0;49	!"	C4567	 819	:;<- = D0;49	!'	(>))%;?2649)	;05;	(?@)	×-.A	?@-	:; × -	819-.B	>) Oven-dried crop subsamples were ground through a 1-mm mesh sieve and then analyzed at ARDC using automated dry combustion method (LECO® F-428 Nitrogen analyzer, Saint Joseph, MI, USA; VELP® NDA 701 Dumas Nitrogen Analyzer, Italy) to determine the concentration of crop N (g N per kg DM). Total N removed by crops was calculated by multiplying the N concentration with DM yield, Equation 2.4 /012	E	05?1F;6	 >)	:;<- = E	())!"	(>))	×/012	34567	(>)	!"):; 	× -	>)-.B	)	 In order to compare the seasonal DM yield and N removal of tall fescue between 3-cut and 5-cut scenarios throughout the year, I divided the entire grass growing season into three stages based on the seasons. Spring growth starts in mid-March when cumulative degree days is greater than 300 and ends on the day of first grass harvest of all farmlets in mid-May or late May. The summer growth ranges from the day after spring harvest to the end of July or first week of August, which included the second and third cut of 5-cut farmlets (F1 and F2) and the second cut of the 3-cut system (F3 and F4). The fall growth started from the day after summer harvest and ends on the second week of October after the final cut of all farmlets, which included the fourth and fifth cut of F1 and F2 and the third cut of F3 and F4. For example, N removal by tall fescue of F1 and F2 during summer harvest equals to the sum of N removal of the second and the third cut of 5 cuts. For 3-cut scenarios (F3 and F4), summer N removal by tall fescue equals to the N removed by the second cut of 3 cuts.    29 2.2.5 Manure analysis  Fresh dairy slurry manure used for the trial, both whole manure and separated manure, were collected in February 2017 and 2018 from typical commercial dairy farms using saw-dust bedding in free-stall barns with frequent scraping. Two types of slurry were stored undisturbed in two identical open pits on site at the ARDC. In spring, frequent rainfall would dilute the manure. Thus, before each application, we used Agro® Nitrogen meter to test NH4+–N concentration of manure and adjusted the volume of manure applied based on test results to meet the target N rate. Manure samples were taken before each application and stored in the freezer until future lab analysis. Manure samples were analyzed using steam distillation (Kjeltec™ 8400 Analyzer Unit, FOSS®, Denmark) for the concentration of NH4+–N and total Kjeldahl N (organic N and ammonium N). Results of Lab analysis were used to adjust the total NH4+–N applied for data analysis. Total N (TN) and total ammonium N (TAN) applied, both in kg per ha, were calculated using the equations below,  Equation 2.5 GE	;226457	(>)	E	:;<-) = HI567;:6	E	()	>)<-)×";9J05	0;85	 >)-... × -K226457	;05;	(?@)	× -.A	?@-	:;  GKE	;226457	(>)	E	:;<-) = K??194J?–E	()	>)<-)×";9J05	0;85	 >)-... × -K226457	;05;	(?@)	× -.A	?@-	:;  30 Table 2.3 Characteristic of different manure types applied, including percentage of dry-matter content (DM%); total ammonium N (TAN), total N (TN); ammonium-N to total N ratio (TAN/TN); total phosphorus (TP); nitrogen to phosphorus ration (N/P), and pH. The range of variation is shown in parentheses for manure that was applied multiple times  2.2.6 Total Nitrogen Recovery  Apparent recovery of total nitrogen applied (TNR) is an indicator of N use efficiency of the forage production system and it is defined as the fraction of applied total N that was recovered by total harvested above ground biomass after corrected for N recovered by non-treated crop from control plots. For each production season, Total N recovery of each crop species was determined by dividing the difference between N removal by crop from each farmlet YearWhole	Slurry DM % 4.9 (4.39-7.27) 4.5 (4.73-8.14) 4.7TAN g	kg-1 1.5 (1.27-1.71) 1.2 (0.93-1.61) 1.3TN g	kg-1 2.5 (2.28-3.09) 2.2 (1.88-2.97) 2.3TAN/TN - 0.6 (0.55-0.62) 0.6 (0.43-0.60) 0.6TP g	kg-1 0.4 (0.36-0.45) 0.3 (0.22-0.44) 0.4N/P - 6.1 (5.11-6.66) 6.7 (4.88-8.12) 6.4pH - 7.3 (6.9-7.6) 7.2 (6.9-7.3) 7.2Seperated	Liquid	Fraction DM % 3.6 (2.95-5.11) 2.5 (1.24-3.38) 3.1TAN g	kg-1 1.1 (0.83-1.32) 1.0 (0.64-1.13) 1.0TN g	kg-1 2.0 (1.29-2.39) 1.7 (1.23-2.08) 1.8TAN/TN - 0.6 (0.54-0.70) 0.6 (0.52-0.69) 0.6TP g	kg-1 0.3 (0.10-0.34) 0.2 (0.09-0.26) 0.2N/P - 7.7 (6.29-12.42) 8.7 (6.77-13.12) 8.2pH - 7.3 (6.9-7.7) 7.2 (7.0-7.5) 7.3Seperated	Sludge DM % 7.9 - 9.1 - 8.5TAN g	kg-1 1.7 - 1.5 - 1.6TN g	kg-1 3.2 - 2.9 - 3.1TAN/TN - 0.5 - 0.5 - 0.5TP g	kg-1 0.5 - 0.4 - 0.5N/P - 6.5 - 6.7 - 6.6pH - 7.0 - 6.8 - 6.9Manure	Type Parameters 2018 Mean2017Unit  31 and N removal by crop from non-treated control plots over the amount of applied TN (Equation 2.6). Equation 2.6 MNO% =	 PQRS	N	QTURVWX − PQRS	N	QTURVWX	Z[	\R\_^QTW^T_	`QRSMR^WX	N	WSSXaT_ ×100% 2.2.7 Farmlet Area-weight Total  Farmlet area-weighted total is the sum per unit area of a dual-crop forage production system based on the area occupied by each crop. It combines the N fluxes of two crops on a farmlet into comparable quantities across different management scenarios.  In this study, the default crop area ratio is 50:50, which means silage corn and tall fescue each occupies 50% of the available land of a farmlet. For example, total N removal (kg ha-1) by all the crop produced for F3 is calculated as,  Equation 2.7 PQRS	N	QTURVWX = 	0.5×(N	QTURVWX	fghi + 	N	QTURVWXklmno	phgq) 	+ 	0.5×(N	QTURVWXrhnss) where, N	QTURVWX	fghi is the amount total N removed (kg ha-1) by harvesting whole plant silage corn; and N	QTURVWX	klmno	phgq is the amount total N removed (kg ha-1) by harvesting Italian ryegrass that were sown into corn plots; and 0.5 is the area ratio of farm land occupied by corn (and later by Italian ryegrass after corn harvest). When land is limited, calculating area-weighted total is a helpful way to explore the nutrient use efficiency of the entire forage production system under different land allocation scenarios. I will assess the effects of crop allocations on farm productivity and N recovery (and N field-losses, discussed in chapter 3) by testing two scenarios of corn:grass area ratios, 50:50 (conventional) and 60:40 (experimental) using area-weighted totals.    32 2.2.8 Statistical analysis Effects of farmlet management scenario on crop yield, N removal by crops, and recovery of total N were analyzed using a linear mixed effects model, R package “lmerTest” (Kuznetsova et al. 2017), with management scenarios (four levels, F1, F2, F3, and F4) and crop year (2017-18 and 2018-19) as fixed effects and block as a random effect. Crop year of corn plots was from May to May based on the planting date and crop year of grass plots was from March to March based the date of when cumulative degree days reached 200. For example, the corn year 2017-18 started on April 26th, 2017 and ended on April 24th, 2018. The grass year 2017-18 started on March 24th, 2017 and ended on March 16th, 2018.  At the time of statistical analysis, the crop year 2018-19 has not ended, i.e. relay crop has not been harvested. Thus, I only present and discuss annual data from the crop year 2017-18 when relay crop was included in the analysis and crop year was not tested as a fixed effect. I used Type III ANOVA with Satterthwaite’s method (R package ‘lmerTest’) to determine significant treatment effects and where differences were found (P<0.05), estimated marginal means of each factor level were investigated using Fisher’s least squared difference test, R package ‘agricolae’ (de Mendiburu 2017) and ‘predictmeans’ (Luo et al. 2018). When farmlet year interactions were significant, the least squared means of each factor level were tested across the years. Data were tested for normality using the Shapiro-Wilk test, and when not normal was log-transformed to meet normality assumptions. All analyses were carried out in R (R Core Team 2017).   33 2.3 Results and Discussion 2.3.1 Silage corn yield Whole plant silage corn yield of F1 an F2 was significantly higher than F3 and F4 over two years. Management scenario has a significant effect on whole plant corn DM yield and grain yield (Table 2.4). The annual average yield of silage corn across four farmlets did not change substantially in 2018 compared to 2017. Corn grain yield of four farmlets was similar in 2017, but F1 and F2 increased by nearly 10% while F3 and F4 decreased by 10% and 17% in 2018, respectively (Table 2.5).  Table 2.4 Analysis of variance of corn DM yield and crop N removal by corn. There are four levels of management scenario (df =3) and two levels of year (df = 1). Bolded values indicate statistical significance.  Variable F P-value F p-value F P-valueCrop	DM	yieldSilage	corn	whole	plant 27.59 <	0.01 4.18 0.05 2.03 0.13Corn	grain	 18.70 <	0.01 0.37 0.55 14.92 <	0.01Corn	harvest	index 2.03 0.13 12.09 <	0.01 24.24 <	0.01Relay	crop†	 26.55 <	0.01 - - - -Corn	whole	plant	and	relay	crop 1.24 0.34 - - - -Harvested	crop	N	removal	Silage	corn	whole	plant 7.02 <	0.01 2.03 0.16 3.14 <	0.05Relay	crop	 8.06 <	0.05 - - - -Corn	whole	plant	and	relay	crop 59.87 <	0.01 - - - -†	At	the	time	of	completing	data	analysis,	Italian	ryegrass	of	crop	year	2018-19	was	not	harvested	yet.Management	scenario Year Management*Year  34 Table 2.5 Crop DM yield of corn whole plant silage and sum of corn and relay crop. Numbers in parentheses are a standard error, n = 4 for each year, n = 8 for mean over two years. Relay crop DM yield in 2018-19 is not available.  Corn yield on F1 (conventional scenario), in which long season (high CHU) corn received 297 kg ha-1yr-1 total N as broadcast slurry (222 kg N ha-1yr-1), starter mineral fertilizer (20 kg N ha-1yr-1 and 40 kg P ha-1yr-1), and side-dressing mineral fertilizer (55 kg N ha-1yr-1), was similar in 2017 and 2018, 20.6 and 19.7 t DM ha-1 respectively (Tables 2.4 and 2.5). Grain yield was greater in 2018 than 2017, but not significant, giving a higher harvest index (53% vs. 44%). The yield of F2 was similar to F1 in both years, but it should be noted that F2 received 271 kg N ha-1 in 2017 and 228 N kg ha-1 in 2018, of which 55 kg N ha-1yr-1 was received as mineral fertilizer.  Grain yield for F2 was similar to F1 in both years. The grain content of F2 in 2018 was greater than 2017 giving a significantly higher harvest index (53% vs 45%) (Table 2.6).  As expected, the lower CHU corn used in F3 yielded about 4.5 t ha-1 less DM than the higher heat unit corn in F2 in both years. However, grain yield was similar for these farmlets in 2017 suggesting a higher grain content (51% vs 44%) (Table 2.6). In contrast, grain yield was lower for F3 than F2 in 2018 and this can be attributed to delayed early growth due to poor seed Meant	DM	ha-1F1 20.80	(0.61) a† 19.43	(0.34) ab 20.12	(0.42) a - 20.80	(0.61) aF2 19.99	(0.52) a 20.53	(0.27) a 20.26	(0.29) a - 19.99	(0.52) aF3 16.76	(0.80) b 14.54	(0.89) c 15.65	(0.69) c 2.63	(0.09) b 19.39	(0.86) aF4 17.78	(1.10) b 17.41	(0.64) b 17.60	(0.59) b 2.90	(0.09) a 20.68	(1.14) aMean 18.84	(0.71) A‡ 17.98	(0.71) A - 2.76	(0.08) 20.22	(0.40)LSD0.05 2.13 2.11 1.24 0.16 1.88†Farmlet	values	in	each	column	followed	by	different	lower	case	letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).	‡Annual	mean	values	in	each	row	followed	by	different	upper	case		letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).	Corn	&	relay	cropManagement	scenariosSilage	corn	whole	plant2017-18 2017-18	Relay	crop2017-18 2018-19  35 vigor. Because of concern about a novel pest in the region (western corn rootworm, Diabrotica virgifera virgifera), a low heat unit BT hybrid was needed and only year-old seed were available. Famlet 4 yield was between F2 and F3 in each year and only significantly different from F3 in 2018, but the mean yield of F4 over two years was significantly lower than F2 (by 2.6 t ha-1) and significantly greater than F3 (2 t ha-1), likely due to addition of DCD to the injected manure in 2017 and to both DCD and irrigation in 2018 (P<0.05) (Table 2.5).   Table 2.6 Corn grain DM yield and grain %. Numbers in parentheses are the standard error, n = 4 for each year, n = 8 for mean over two years.  Italian ryegrass received 123 kg N ha-1 in March 2018 before being harvested a month later (Table 2.7). In the crop year 2017-18, corn plots of F3 and F4 produced 2.6 and 2.9 t ha-1 of Italian ryegrass, respectively. In total, corn plots of F3 and F4 produced 19.4 and 20.2 t ha-1 of biomass, including the relay crop, which were similar to the corn yield of F1 and F2 (Table 2.5). The deliberate decision of intercropping Italian ryegrass with short-season corn hybrid was mean to increase percentage grain content in corn silage while allowing relay crop to scavenge residual PAN in the soil before winter rainfall starts. Corn grain has high starch content, which is a Mean Meant	DM	ha-1 %F1 9.18	(0.30) a† 10.29	(0.23) a 9.74	(0.27) ¶ 44.2% b 52.9% a 48.6%F2 8.89	(0.23) a 10.75	(0.16) a 9.82	(0.38) 44.5% b 52.4% a 48.5%F3 8.66	(0.33) a 7.16	(0.52) c 7.91	(0.40) 51.7% a 49.2% b 50.5%F4 9.26	(0.49) a 8.30	(0.91) b 8.78	(0.36) 52.1% a 47.7% b 49.9%Mean 9.00	(0.18) A‡ 9.13	(0.18) A - 48.1% B 50.5% A -LSD0.05 0.87 0.87 0.61 2.9% 2.9% 2.0%†	Farmlet	values	in	each	column	followed	by	different	lower	case	letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).	‡	Annual	mean	values	in	each	row	followed	by	different	upper	case		letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).¶	Significant	difference	between	2-year	means	of	each	farmlet	are	not	shown	given	farmlet	*	year	interactions	are	significantManagement	scenariosCorn	grain	yield	 Corn	harvest	index2017-18 2018-19 2017-18 2018-19  36 critical indicator of feed quality for dairy cows. Greater grain yield of whole plant corn silage could potentially reduce the demand for purchasing energy supplement for cows. Although short season corn cannot produce as much whole plant silage as long season corn, with the addition of relay crop yield, farmers could potentially produce a similar amount of feed with a higher content of grain. 2.3.2 Silage corn N removal and TN recovery Silage corn of the F1 removed the lowest amount of N, 170 kg ha-1, in 2017, but received the highest input of N, 282 kg ha-1. Farmlet 2 removed 27 kg ha-1 more N than F1 over two years while receiving 35 kg N ha-1 less N (Table 2.7). Total N Recovery (TNR%) of F2 was 6% higher than F1 in 2017 and 9% higher in 2018 (P<0.05) (Figure 2.2). The improvement of N recovery could be attributed to replacing broadcasting whole manure slurry with an injection of manure sludge, which was expected to reduced volatilization of ammonium N. With the similar amount of TN input, F3 removed less N than F2 in both years, including a 42 kg ha-1 deficit in 2018. Total N recovery of F3 was similar to F2 in 2017, but 15% lower than F2 in 2018, which could also be the result of delayed early growth of low CHU corn. Farmlet 4 also received a similar amount of TN as F2 and F3, 264 kg ha-1 yr-1, through the same manure application method (Table 2.7) yet N removal in F4 was similar to F2 and significantly higher than F3 in both years. Total N recovery of F4 was significantly higher than all the other farmlets in 2017 and 10% higher than F3 in 2018 (Figure 2.2). Thus, despite using the old seed of low CHU corn, irrigation and DCD seemed effective in improving N recovery.  Considering N removed by relay crop, N removal of F3 and F4 were significantly higher than F1 and F2 in 2017-18 crop year. Italian ryegrass removed 61 and 65 kg N ha-1, respectively, which increased the annual N removal of F3 and F4 to 238 and 264 kg N ha-1 (Table 2.7). Local   37 farmers often spread manure in Fall, when production season end, to make room for manure storage during the wet winter months and spread again in mid-March when manure storage reaches full capacity over the winter. My data showed that compared to monocrop silage corn, intercropping Italian ryegrass with early mature corn can recover more PAN from manure applied during non-production months. Therefore, if farmers were to manure during the non-production season to meet storage objectives, this BMP was likely to prevent N losses and increase N recovery.  This was consistent with the analysis by Van Vliet and Zebarth (2004), which reported that using Italian rye-grass as relay crop on post-harvest corn field reduced NO3-–N (61%), NH4+–N (33%) and total N (56%) load in soil significantly.    38 Table 2.7 Annual nitrogen (TAN and TN) application rate and nitrogen removal by silage corn. Numbers in parentheses are standard error, n = 4 for each year, n = 8 for mean over two years. Relay crop data is incomplete for 2018-19. 2017-18 2018-19 2017-18 2018-19 2017-18 2018-19Corn kg	P	ha-1yr-1 kg	N	ha-1F1 40 71 75 75 187 202 282 312 170	(6.4) b† 189	(5.4) a 180 (9.1)¶F2 0 34 35 55 169 170 271 288 181	(4.1) ab 205	(5.5) a 193 (5.5)F3 0 35 35 55 173 171 278 289 178	(6.8) b 163	(13.4) b 170 (7.5)F4 0 32 35 55 165 170 264 288 199	(10.5) a 200	(9.1) a 200 (6.3)Mean 182	(4.3) A‡ 189	(5.8) A -LSD0.05 22 32 15Relay	cropF1 - - - - - - - - - - -F2 - - - - - - - - - - -F3 0 15 - 0 67 - 124 - 61	(1.3) a - -F4 0 15 - 0 67 - 123 - 65	(2.5) a - -Mean 63	(1.5)LSD0.05 5Corn	+	Relay	CropF1 40 73 - 75 187 - 282 - 170	(6.4) c - -F2 0 34 - 55 169 - 271 - 181	(4.1) c - -F3 0 50 - 55 240 - 401 - 238	(7.8) b - -F4 0 49 - 55 232 - 387 - 264	(21.0) a - -Mean 214	(2.6) - -LSD0.05 19 - -¶	Significant	difference	between	2-year	means	of	each	farmlet	are	not	shown	given	farmlet	*	year	interactions	are	significant‡	Annual	mean	values	in	each	row	followed	by	different	upper	case		letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).†	Farmlet	values	in	each	column	followed	by	different	lower	case	letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).	TN	applied	TAN	applied	 Annual	Crop	N	Removal2017-18 2018-19 MeanManagement	scenarioFertilizer	P	applied	Fertilizer	N	applied	TP	applied  39 Table 2.8. Analysis of variance of apparent recovery of TN. There are four levels of management scenario (df =3) and two levels of crop (df = 1). Bolded values indicate statistical significance.   Figure 2.2 Apparent recovery of TN applied for both crops in two consecutive years (2017-19). For each crop in one year, farmlets labeled with different lower-case letters are significantly different (P < 0.5).Variable F P-value F p-value F P-valueApparent	recovery	of	TN2017-18 16.41 <	0.01 24.31 <	0.01 2.28 0.102018-19 12.64 <	0.01 37.01 <	0.01 3.94 0.02Management	scenario Crop Management*Crop  40 2.3.3 Grass yield  The most advanced two farmlets, F3 and F4, produced significantly more grass than F1 and F2 in both 2017 and 2018.  In 2018, average grass yield of all four scenarios increased about 2.2 tons DM ha-1 comparing to the previous year and grass yield of each farmlet improved incrementally as the management scenario becomes more advanced and complex (Table 2.10). Grass yield of F3 was 2.6 t ha-1 higher than F2, which suggests less frequent grass cut would promote grass yield (Table 2.10). The most advanced farmlet, F4, which also received about 625 kg N ha-1 yr-1 through surface banding of SLF, produced 1.1 t ha-1 yr-1 more grass than F3 (Table 2.10), which could be attributed to the addition of DCD and irrigation. Total N applied to F4 was not substantially different from F3 in any stage of the growing season (Table 2.10), however F4 produced significantly higher grass yield than F3, 3.2 and 2.2 t DM ha-1 yr-1, during the Fall growth (from August to October) of both years, which could be attributed to the weekly irrigation.   Table 2.9 Analysis of variance of tall fescue DM yield and tall fescue N removal (n=8) in 2017 and 2018, managed under four different scenarios. Significant results are highlighted in bold. Variable F P-value F p-value F P-valueTall	fescue	DM	yieldSpring	growth 77.26 <	0.01 2.68 0.11 1.82 0.16Summer	growth 4.23 <	0.05 46.26 <	0.01 4.73 <	0.01Fall	growth 50.03 <	0.01 101.02 <	0.01 0.86 0.47Annual	total 121.45 <	0.01 137.03 <	0.01 0.83 0.48N	removal	by	tall	fescueSpring	growth 6.72 <	0.01 28.83 <	0.01 4.55 <	0.01Summer	growth 17.58 <	0.01 91.07 <	0.01 7.35 <	0.01Fall	growth 63.00 <	0.01 118.92 <	0.01 0.94 0.43Annual	total 36.86 <	0.01 197.79 <	0.01 2.14 0.12Management	scenario Year Management*Year  41 Table 2.10 Tall fescue DM yield in 2017 and 2018. Numbers in parentheses are the standard error. n = 4 for each year, n = 8 for mean over two years.  2017 2018Spring	HarvestF1 50 61 5.63	(0.10) b† 6.12	(0.17) b 5.87	(0.13) bF2 50 61 5.14	(0.07) b 5.83	(0.19) b 5.49	(0.16) bF3 70 77 7.88	(0.15) a 7.82	(0.32) a 7.85	(0.16) aF4 70 77 8.11	(0.51) a 7.98	(0.12) a 8.04	(0.25) aMean 6.69	(0.11) A‡ 6.93	(0.11) A -LSD0.05 0.89 0.60 0.44Summer	HarvestF1 85 77 3.25	(0.25) ab 3.47	(0.15) b 3.36	(0.14) ¶F2 85 77 3.67	(0.23) a 4.25	(0.15 a 3.96	(0.17)F3 65 61 3.08	(0.29) b 4.47	(0.20) a 3.77	(0.74)F4 65 61 3.12	(0.22) b 4.25	(0.11) a 3.68	(0.57)Mean 3.28	(0.10) B 4.11	(0.10) A -LSD0.05 0.51 0.51 0.17Fall	HarvestF1 73 78 0.92	(0.09) c 1.93	(0.13) c 1.42	(0.2) dF2 73 78 1.13	(0.12) c 2.36	(0.09) bc 1.75	(0.24) cF3 73 78 1.76	(0.26) b 2.64	(0.23) b 2.20	(0.23) bF4 73 78 2.55	(0.29) a 3.88	(0.10) a 3.22	(0.27) aMean 1.59	(0.08) B 2.7	(0.08) A -LSD0.05 0.51 0.46 0.32Annual	totalF1 208 216 9.80	(0.21) b 11.51	(0.31) d 10.66	(0.37) cF2 208 216 9.94	(0.29) b 12.45	(0.31) c 11.20	(0.51) cF3 208 216 12.72	(0.23) a 14.93	(0.38) b 13.83	(0.47) bF4 208 216 13.78	(0.43) a 16.10	(0.19) a 14.94	(0.49) aMean 11.56	(0.13) B 13.75	(0.11) A -LSD0.05 1.09 0.74 0.55¶	Significant	difference	between	2-year	means	of	each	farmlet	are	not	given	when	farmlet	*	year	interactions	are	significant‡	Annual	mean	values	in	each	row	followed	by	different	upper	case		letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).†	Farmlet	values	in	each	column	followed	by	different	lower	case	letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).	2017-18 2018-19 Annual	MeanManagement	scenarioTall	fescue	DM	yield	(ton	ha-1)Growth	days  42 2.3.4 Grass N removal and TN recovery Despite receiving less TN in 2018, tall fescue of all four farmlets removed an average of 72 kg ha-1 more N than in 2017 (Table 2.11). It is not surprising that the annual total N removal of F2 was 46 kg ha-1 higher than F1 over two years, giving F2 received 120 kg ha-1 more N in 2017 (Table 2.11). However, %TNR of F2 was only significantly higher than F1 in 2018 (55% vs. 44% when they both received about 540 kg ha-1 N. Annual total N recovery of F2 and F3 were not different in both years, as they removed similar amount of total N while received similar amount of TN (Table 2.11). Thus, reduced harvest frequency improved only annual grass yield of F3 but not annual N removal, which indicated relatively lower N content and nutrient value (e.g. crude protein content) in harvested grass. However, F3 removed more N than F2 during the fall growth over two years (Table 2.11), which indicates that giving tall fescue longer time to grow in late summer could recover more N than harvesting once every six weeks. Low soil moisture content could have been the constraint resulting in slow regrowth in late summer.  Farmlet 4 received a similar amount of TN as F3 in both years but removed 27 kg ha-1yr-1 more N (Table 2.11) on average over two years. Annual total N recovery of F4 was significantly higher than all the other farmlets in 2017 but only higher than the conventional farmlet in 2018 (55 % vs. 44 %) (Figure 2.2). By the end of July in both years, cumulative N removal of three more advanced management scenarios (F2, F3, and F4) were all evidently higher than the F1but not different from each other (Figure 2.3). Cumulative N removal of F4 surpassed F2 and F3 during the irrigation period from August to October (fall growth) in both years (Figure 2.3). Grass of F4 recovered about 28 kg ha-1 more N than F3 on average of two years (Table 2.11), but the TN recovery of F4 was only significantly higher in 2017 (Table 2.11). The differences observed between F3 and F4 are likely due to the combination of adding nitrification inhibitor   43 into manure SLF and weekly irrigation during summer. In a meta-study on effect of nitrification inhibitor on nitrous oxide emission and crop yield, Thapa et al (2016) found that a total of 21 studies have reported an average of 5.2% (CI: 2.1 – 8.1%) increase of crop yield from application of nitrification inhibitor in irrigated systems comparing to conventional N fertilizer. DCD added to the manure would prevent ammonium N from being nitrified all at once and sustained the release of plant available N. Ledgard et al. (2014) found that application of 10 kg ha-1 DCD delayed the declining of ammonium concentration in urine treated soil for 4-6 weeks. Subbarao et al. (2006) and Zaman and Blennerhassett (2010) concluded that DCD can potentially improve N recovery and crop yield by facilitating the crop uptake of N in NH4+ form with less cost of energy than NO3-, which is critical for regrowth of perennial grass.   44  Figure 2.3 Cumulative N removal for each farmlet by tall fescue throughout the growing season (March - October) in 2017-18 (left) and 2018-19 (right). Colors indicate different grass harvest frequency and arrows indicate the dates of manure applications.   45 Table 2.11 Annual N (TAN and TN) application rate, N removal by tall fescue, and % TN recovery by tall fescue. Numbers in parentheses are the standard error. n = 4 for each year, n = 8 for mean over two years.  2017 2018 2017 2018 2017 2018 2017 2018Spring kg	ha-1 kg	N	ha-1F1 50 61 28 20 100 85 174 152 123 (2.0)	b† 154 (11.9)	a 139 (10.0)¶F2 50 61 33 23 132 103 238 188 126 (5.1)	b 163 (5.4)	a 144 (7.7)F3 70 77 50 31 196 136 355 249 153 (6.1)	a 162 (9.1)	a 158 (15.1)F4 70 77 49 30 195 132 352 244 156 (7.3)	a 160 (5.5)	a 158 (4.3)Mean 140 (4.6)	B‡ 160 (3.1)	A -LSD0.05 15 15 11SummerF1 85 77 31 28 122 114 202 192 77 (6.1)	b 84 (3.2)	c 81 (3.4)F2 85 77 44 23 165 116 296 196 94 (5.9)	a 115 (3.4)	a 104 (5.1)F3 65 61 34 29 135 120 226 209 69 (5.2)	b 106 (4.3)	ab 87 (7.7)F4 65 61 33 27 131 115 220 199 73 (4.3)	b 101 (3.6)	b 87 (5.9)Mean 78 (3.5)	B 101 (3.3)	A -LSD0.05 10 10 7FallF1 73 78 40 37 123 100 210 199 28 (3.6)	c 52 (3.2)	c 40 (5.1)	dF2 73 78 19 18 110 91 173 153 35 (4.0)	c 71 (2.3)	b 53 (7.1)	cF3 73 78 9 8 79 70 113 101 49 (6.0)	b 76 (5.7)	b 63 (6.4)	bF4 73 78 9 11 78 73 112 120 75 (6.3)	a 106 (1.9)	a 91 (6.6)	aMean 47 (5.2)	B 76 (5.3)	A -LSD0.05 12 11 8Annual	totalF1 208 216 98 85 345 299 586 543 229 (7.0)	c 289 (9.2)	b 259 (15.2)	cF2 208 216 96 64 407 310 707 537 255 (10.1)	b 348 (19.3)	a 301 (18.9)	bF3 208 216 93 67 409 327 694 559 271 (7.1)	b 345 (25.8)	a 308 (15.5)	bF4 208 216 91 68 404 320 684 564 303 (10.0)	a 366 (17.1)	a 335 (12.8)	aMean 265 (7.7)	B 337 (8.7)	A -LSD0.05 25 31 15TP	applied¶	Significant	difference	between	2-year	means	of	each	farmlet	are	not	shown	given	farmlet	*	year	interactions	are	significant‡	Annual	mean	values	in	each	row	followed	by	different	upper	case		letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).†	Farmlet	values	in	each	column	followed	by	different	lower	case	letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).	Management	ScenarioN	removal	by	tall	fescueTN	appliedTAN	applied2017 2018 MeanGrowth	days  46 2.3.5 Farmlet area-weighted total dry-matter yield and N removal rate for different land allocation scenarios Comparing land allocation scenarios where N applied to and recovered by relay crop are excluded, F4 was superior for yield and N recover metrics for both the 50:50 and 60:40 scenarios in 2017, while the other three farmlets had little differences. F2 received 55 kg ha-1 more TN than F1 (Table 2.14). However, there was no significant difference between the crop yield and % TN recovery of these two farmlets during the crop year 2017, which indicated that advanced nutrient management BMPs did not improve crop production and N recovery at a system level (Figure 2.4). Increasing corn area to 60% would increase crop yield by 0.7 t ha-1 while recovering similar percentage of TN applied in 2017 (Table 2.14).  In the 2018-19 crop year, increasing corn area to 60% would increase the area-weighted total crop yield for 0.7 and 0.8 t DM ha-1 on F1 and F2, while F3 and F4 remained the same. Total N removal by crops decreased by an average of 15 kg ha-1 on every farmlet (Table 2.14). Same as 2017-18 crop year, increasing corn area to 60% would not change TN recovery of any farmlet (48.3% vs. 47.4%). Recall from the previous section, annual mean grass yield of all farmlets increased 2.2 t ha-1 (13.75 vs. 11.56 t DM ha-1, Table 2.8) and annual mean TN recovery by grass increased 22% (31% vs. 52%, Table 2.14) in 2018, while average yield and TN recovery of silage corn remained the same. Hence, land allocation had a great impact on crop production and N recovery of dual-crop forage systems.  When crop yield and N recovery metrics of relay crop were included for the crop year 2017-18, crop N removal increased incrementally as the management scenario became more improved. F3 received about 60 kg ha-1 more N than F2, while recovering 37 kg ha-1 more N through a 1.1 t ha-1 increase of yield. . Alternatively, F4 still produced the most total biomass   47 (15.8 t DM ha-1) and removed the highest amount of N (251 kg ha-1) (Table 2.14). The significantly higher crop yield and crop N removal of F4 over F3 indicated that irrigation and addition of the nitrification inhibitor effectively improved N recovery at the system level for dual-crop forage production.   If silage corn occupied 60% of the land, and grass 40%, the pattern of differences in crop DM yield and crop N removal between each management scenario remained the same as the land allocation scenario with a 50:50 distribution. Nonetheless, area-weighted total crop yield would increase by an average of 0.9 t DM ha-1, while total crop N removal decreased by 2.6% or 5 kg ha-1 (Table 2.14). Since the N content of tall fescue is approximately 3 times higher than whole plant silage corn, 10% less grass area would very likely to decrease total crop N removal of the whole farmlet.   48 Table 2.12 Analysis of variance of farmlet total yield and total N removal in 2017 and 2018, managed under four different scenarios (df =3) and two land allocation scenarios (df =1). Significant results are highlighted in bold.  Table 2.13 Analysis of variance of total N recovery in 2017 and 2018, managed under four different scenarios (df =3) and two land allocation scenarios (df =1). Significant results are highlighted in bold.  Variable F P-value F P-value F P-value2017-18	(Relay	crop	included)Area-weighted	total	crop	yield 17.16 <	0.01 15.36 <	0.01 0.25 0.86Area-weighted	total	crop	N	removal 313.02 <	0.01 5.53 <	0.05 0.18 0.912017-18	(Relay	crop	excluded)Area-weighted	total	crop	yield 4.52 0.01 11.04 <	0.01 0.74 0.54Area-weighted	total	crop	N	removal 75.32 <	0.01 12..26 <	0.01 0.19 0.902018-19	(Relay	crop	excluded)Area-weighted	total	crop	yield 30.86 <	0.01 5.33 <	0.05 1.45 0.25Area-weighted	total	crop	N	removal 97.43 <	0.01 55.16 <	0.01 0.79 0.51Management*Area	ratioManagement	scenario Crop	area	ratioVariable F P-value F p-value F P-valueApparent	recovery	of	TN2017-18 81.61 <	0.01 1.08 0.31 0.10 0.962018-19 146.50 <	0.01 4.20 0.05 0.44 0.73Management	scenario Crop	Area	Ratio Management*Area	ratio  49  Figure 2.4 Area-weighted total recovery of TN applied for two land allocation scenarios in two consecutive years (2017-19). For each land allocation scenario in the same year, farmlets labeled with different lower-case letters are significantly different (P < 0.5). Between land allocation scenarios in the same year, same upper-case letters indicate lack of significant difference (P < 0.5).  50 Table 2.14 Annual crop yield, N removal, N (include commercial N and manure N) application rate, and P (including commercial P) application rate of four farmlets under two land allocation scenarios. n = 4 for mean of each farmlet and n=16 for annual mean of each crop area ratio.    2017 2018 2017 2018 2017 2018kg	N	ha-1 %F1 85	(20)¶ 80	(20) 266 250 434	(38)¶ 428	(38) 15.3 a† 15.5 b 200 c 239 c 28.8 c 42.4 cF2 65 49 288 240 489	(23) 413	(23) 15.0 a 16.5 a 218 b 276 a 29.3 c 53.0 a50	:	50 F3 64 51 291 249 486	(23) 424	(23) 14.7 a 14.7 b 224 b 254 b 31.2 b 45.6 bF4 62 52 285 245 474	(23) 426	(23) 15.8 a 16.8 a 251 a 283 a 37.6 a 52.3 aLSD0.05 - - - - - - 1.1 0.9 13 10 1.9 1.8F1 82	(24) 79	(24) 250 241 404	(45) 404	(45) 16.4 a 16.2 a 194 c 229 b 29.0 c 41.9 cF2 59 46 264 226 445	(33) 388	(33) 16.0 a 17.3 a 211 b 262 a 30.0 bc 52.4 a60	:	40 F3 58 48 267 233 444	(33) 397	(33) 15.1 a 14.7 b 215 b 235 b 31.5 b 43.9 bF4 56 48 261 230 432	(33) 398	(33) 16.2 a 16.9 a 241 a 267 a 38.3 a 51.6 aLSD0.05 - - - - - - 1.3 1.1 13 12 1.9 1.850	:	50 Mean 69 58 282 246 471 423 15.2 B‡ 15.9 B 223 A 263 A 31.7 A 48.3 A60	:	40 Mean 64 55 261 232 431 397 15.9 A 16.3 A 215 B 248 B 32.2 A 47.4 ALSD0.05 - - - - - - 0.5 0.4 5 4 0.9 0.9F1 85	(20) - 266 - 434	(38) - 15.3 b - 200 d - - -F2 65 - 288 - 489	(23) - 15.0 b - 218 c - - -50	:	50 F3 71 - 325 - 548	(23) - 16.1 b - 255 b - - -F4 69 - 318 - 535	(23) - 17.2 a - 284 a - - -	LSD0.05 - - - - - - 1.2 - 12 - - -F1 82	(24) - 250 - 404	(45) - 16.4 b - 194 d - - -F2 59 - 264 - 445	(33) - 16.0 b - 211 c - - -60	:	40 F3 67 - 308 - 518	(33) - 16.7 ab - 251 b - - -F4 65 - 301 - 506	(33) - 17.9 a - 280 a - - -LSD0.05 - - - - - - 1.3 1250	:	50 Mean 73 - 299 - 502 - 15.9 B - 239 A - - -60	:	40 Mean 68 - 281 - 468 - 16.8 A - 234 B - - -LSD0.05 - - - - - - 0.5 5Crop	N	Removalkg	P	ha-1Total	N	Recovery2017 2018†	For	the	same	crop	area	ratio,	farmlet	values	in	each	column	followed	by	different	lower	case	letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).	‡	Farmlet	mean	values	of	each	crop	area	ratio	in	each	column	followed	by	different	upper	case		letters	are	significantly	different	by	Fisher’s	protected	least	square	difference	(LSD,	P<0.05).¶	Values	in	parenthses	are	amount	(kg	ha-1)	of	commercial	fertilizer	(N	or	P)	appliedkg	N	ha-1Relay	crop	excludedRelay	crop	includedManagement	scenarioCrop	area	ratio	(Corn	:	Grass) 2017 2018TN	AppliedTP	Applied TAN	Applied	 Crop	Yield	t	DM	ha-12017 2018  51 2.4 Conclusions Due to the nature of the short-season corn hybrid, Farmlet 3 and 4 produced less corn silage than F1 and F2 in both years of this study, however, the short-season corn could yield the similar amount of grain as the long-season corn if the supply of new seed is guaranteed. Intercropping Italian ryegrass with corn as a relay crop would likely compensate the corn silage yield penalty from planting short-season corn and also recover N from unavoidable manure application during the non-production season, e.g. F3 and F4. Reduced harvest frequency increased grass yield of F3 and F4 in both years, especially in early season and late season. When a similar amount of N was applied to F3 and F4, the combination of irrigation and DCD would improve the recovery efficiency of total N during the driest time of production season for both corn and grass. Hence, by implementing a relay crop, irrigation, and DCD recovery of TN applied was enhanced during both production and non-production seasons.  When combining corn and grass at a 50:50 crop area ratio to calculate area-weighted totals of a farmlet, F4 produced the similar amount of total feed while recovering the highest amount of N in two years. Increasing corn area to 60% of the farmlet area did not affect total N recovery but reduce total N removal, which could attribute to higher N content in grass. When relay crop was included to calculate area-weighted totals in 2017, annual feed production and crop N removal of F4 became significantly higher than all the other scenarios. Increasing corn area by 10% would increase crop yield but reduce TN removal significantly with less TN input. The efficacy of integrated management scenarios and reallocating crop area are largely affected by the annual performance of both crops, which means the multi-year trial is needed to validate the incremental improvement. Moreover, the implication of this study at a whole-farm scale is unclear until feed nutrient value and model simulated milk production are tested because   52 the ultimate goal of the future study is to identify management practices that reduce the import of feed and fertilizer while maintaining milk production.   53 Chapter 3: Evaluating the Effectiveness of Beneficial Management Practices on Reducing Potential Nitrate Leaching and Nitrous Oxide Emission from Dairy Farm Forage Production Systems 3.1 Introduction Although dairy farms produce substantial quantities of nutrients in cow manure that are valuable for crop production, such as nitrogen (N) and phosphorus (P), using these nutrients efficiently continues to be a challenge for farmers. In the Lower Fraser Valley (LFV) region, a typical manure slurry (wet weight) is about 0.3% N, of which, 50% is less soluble organic N, much of which will not be available to the crop in the immediate production season. Every year, an estimated 47% (24.4 kt N yr-1 lost as NH3 and 27.2 kt N yr-1 retained in soil) of the manure is lost through ammonia volatilization during land spreading of dairy manure in Canada (Bittman et al., 2017a). Because of losses and low availability of much of the remaining N, high rates of manure N are complimented with commercial mineral fertilizers to ensure high production of feed and crude protein. Applying such a high rate of manure TN can lead to the accumulation of labile C, organic N, and P in soil, which can result in both positive or negative outcomes. Additions of labile C and organic N into the soil provide an abundant substrate for microbial activities, such as mineralization of organic N into ammonium (NH4+) and nitrification to nitrate (NO3-), forms of plant available nitrogen (PAN) that are essential for crop productivity. However, high C content will also result in immobilization of ammonium, which reduces PAN in soil. During the production season, slow and sustained release of NO3- is beneficial for plant growth, but can also result in the gaseous by-product of nitrification, nitrous oxide (N2O), a greenhouse gas, with 298 times the global warming potential of carbon dioxide. During the non-production season, continued mineralization and nitrification will produce NO3- that, without crop uptake, is prone to lose to the surrounding environment through either runoff or leaching. In   54 LFV, prolonged precipitation in winter cause leaching of NO3- into the aquatic ecosystem. Denitrification of residual NO3- throughout the season can generate N2O emission. Accumulated excessive P in soil could be lost to water bodies through runoff and erosion and in some cases leaching. Dairy farmers are encouraged to implement beneficial management practices (BMPs) to improve the management of crops and nutrients to improve farm production and reduce field N-losses. Nutrient management BMPs were designed to reduce losses of nutrients from volatilization, inefficient transformations, leaching and runoff when manure is applied. Plot experiments have shown that methods to assist infiltration, e.g. remove manure solids, surface banding, and injection of manure slurry, greatly reduce ammonia volatilization (Bittman et al. 2005; Bhandral et al., 2009). However, BMPs that assist infiltration and reduce ammonia may increase N2O–N emission and NO3- –N leaching as more N enters the soil (Rubæk et al. 1996; Vallejo et al. 2005). Alternatively, nitrification inhibitors, such as dicyandiamide (DCD), have been shown to effectively inhibit NO3- production and N2O emission in the soil when applied with manure solid (Zhu et al. 2016); cow urine and dung (Ledgard et al. 2014). Cropping BMPs are designed to capture more nutrients from the soil to reduce the import of nutrient in forms of fertilizer or feed. Intercropping annual grass between corn rows, for example, has been shown to recover residual NO3- after corn harvest and reduce the potential of - NO3-−N leaching loss during the non-production season (Bittman et al. 2004). The growth of silage corn is very sensitive to water stress (Rhoads et al. 1990; Pandey et al. 2000), especially at a flowering stage in the summer (Farré and Faci 2009), which is the driest period of the LFV region. Irrigation, which is rarely used on silage corn in the LFV, has been shown to promote crop yield and N recovery of silage corn (Derby et al. 2005; Karasu et al. 2015). However,   55 studies have found a strong positive correlation between water-filled pore space (WFPS) of soil and N2O emission (Linn and Doran 1984; Dobbie and Smith 2003). The research that has been done on forage BMPs clearly illustrates the challenge of synchronizing transformations of soil organic N to PAN with crop demand while minimizing N losses through various pathways. Few studies, however, have examined how combining BMPs might address this challenge and improve N efficiency. Moreover, plot experiments on a single source of N flux fail to address the potential trade-offs between different N field-loss pathways and crop types. The synergistic effect, or possible trade-offs, of integrating a number of BMPs, especially the combination of both nutrient and crop BMPs, at the farm-scale is unknown and rarely examined before.  A more holistic approach is required to evaluate the efficacy of these BMPs, particularly if used together.  In this study, I quantified the annual (the crop year 2017-18) N filed-losses from N2O−N emission and potential NO3-−N leaching of a dual-crop forage system under four management scenarios incrementally improved with either nutrient or cropping BMPs or combinations of them. I also examined the effectiveness of adjusting cropland allocation on N field-loss. The objectives were to: (1) identify the scenario that achieved the lowest N2O−N emission and potential NO3-−N leaching intensity, i.e. g N loss per kg DM feed production; (2) determine whether adjusting crop area ratios between silage corn and grass would affect total N loss of a farm.   56 3.2 Material and Methods 3.2.1 Site description and Experiment design Complete block design with four blocks and split-plot of two types of forage crop, silage corn, and tall fescue. Four incrementally improved management scenarios of forage production system were randomly assigned to four sub-plots within each main plot of the crop. The study site and experimental design are described in detail in section 2.2.1.  3.2.2 Soil sampling and plant available N (PAN) analysis Soils cores were collected using a 2.03 cm soil probe from three depth ranges, 0-15 cm, 15-30 cm, and 30-60 cm. Soil samples from eight randomly selected locations within each plot were combined into composite samples by depth. Soils were sampled monthly from each plot throughout the year, particularly targeting key times of the year such as before manure application and after harvest. Gravimetric soil water content (SWC) was measured as the mass loss of a subsample (approximately 50 g) after oven-drying at 105°C until constant weight. The remainder of the fresh soil samples were either extracted with 2M KCl solution if time permitted or air-dried, crushed and sieved to < 2-mm. We extracted NO3- and NH4+ from 5 g of air-dried soil or 10 g of fresh soil with 50 ml 2M KCl solution by shaking for 1 h (1:10 soil/ solution ratio) and analyzed for NO3-−N and NH4+−N concentration using standard colorimetric methods on flow injection analyzer (FIA, FOSS®, Denmark) at ARDC.  Soil NO3-−N and NH4+−N content (kg ha-1) at each depth of soil profile was calculated using the equation below,   57 Equation 3.1 !	#$#%&'(	)$	*+)%	 ,-.#= !	#$#%&'(	0+$0($'1#')+$	+2	*+)%	(3'1#0'#$'	 ,-4 ×	62	×78 ,-9:×;+%<9(	+2	*+)%	%#&(1(9:.#)	where, concentration of NO3-−N and NH4+−N (kg L-1) were converted from the concentration given by FIA analysis (ppm); df is the dilution factor (L kg-1), which was calculated as 0.05 L KCl solution divided by the weight of extracted soil, 0.005 kg for air-dried soil or 0.01 kg fresh soil multiplied by the SWC; Db is the bulk density of soil at the depth soil samples were taken (kg m-3), volume of soil layer (m3 ha-1) was calculated by multiplying area of one hectare (104 m2) with the depth of the soil layer we took samples from, i.e. 0.15 m for soil samples taken from 0-15 cm and 15-30 cm layers and 0.3 m for soil samples taken from 30-60 cm layer. Bulk density samples were taken from three depth ranges of 8 random locations in the field (2 crop types x 4 blocks) using 2.03 cm diameter zero contamination probes with a predetermined volume. Soil samples are weighed before and after oven-dried at 105°C until constant weight. Bulk density is calculated by dividing the oven-dried sample weight by the volume of the core expressed as kilogram per cubic meter.  3.2.3 Soil pore water sampling and quantification of potential NO3-–N leaching Soil pore water samples were obtained through ceramic-cup suction lysimeters. Suction lysimeters were installed towards one end of each plot in May 2017, one in grass plots and two in corn plots (one in the corn row and one between corn rows), to collect soil pore water from 50-cm depth. The lysimeters were installed at an angle so that they were not disturbed by field operations. All lysimeters were evacuated to 70 kPa using a hand pump and left primed for at   58 least 16 hours. Subsamples were collected the next day from the lysimeters by pumping the soil extract into a washed Erlenmeyer flask and rubber tubing (Appendix C). Water samples were collected weekly from May 2017 to December 2018, whenever water was available. These samples were analyzed for concentrations of NO3- –N and NH4+–N using FIA as described previously in section 3.2.2.  Transport of NO3- and NH4+ in soil was simulated on daily steps with a soil water balance (SWB) model. SWB model was based on the soil water module of NLOS (NLEAP on Stella®, Shaffer et al. 1991; Bittman et al. 2001) developed on Microsoft® Excel assuming that water movement occurred only under saturated conditions and that transport of solutes in the soil profile was in a “Steady-state”. This means that each defined soil layer acts as a tipping bucket with homogenous soil hydraulic properties and water will percolate to a soil layer only when water influx exceeds the field capacity of the soil layer above (Shaffer et al. 1991; Bittman et al. 2001). I defined three soil layers, i.e. 0-150 mm (L1), 150-300 mm (L2), and 300-600 mm (L3) (Table 3.1) by convention and because these layers align with physical properties and the degree of impact from agricultural field management. Water draining below 600 mm was assumed to transport soluble NO3- and NH4+ below the root zone and to be lost form the soil. Soil bulk density (as described in section 3.2.2), texture (tested by a commercial lab), and soil organic matter content (determine by another study on the same site) of each layer were used to determine the field capacity and permanent wilting point of each soil layer. To initiate the simulation, the SWB model requires input parameters, including daily weather and solar radiation; a crop coefficient for transpiration; the starting date of each growing stage; and soil hydraulic properties. First of all, I calculated the daily reference evapotranspiration (ETo) from January 1st of 2017 to December 2018, using the Penman-  59 Monteith method formulated by the Food and Agriculture Organization of the United Nations (FAO) (Allen et al. 1998). Hourly weather data, air temperature (T), atmospheric pressure (P), wind speed at 2 meter above ground (u2), and relative humidity (RH) of local weather station (Environment Canada, 2018), and daily solar radiation data (obtained from the NASA Langley Research Center (LaRC) POWER Project funded through the NASA Earth Science/Applied Science Program) were used to calculate the input parameters of the FAO Penman-Monteith equation (Equation 3.2), Equation 3.2  ?@A = B. DBE	∆	 G$ − I + K( LBB@ + MN:)<M(;O7)∆ + K(P + B. :D<M)  where, T is air temperature at 2 m high (°C); VPD is vapor pressure deficit (kPa), i.e. difference between saturated vapor pressure and actual vapor pressure; u2 is wind speed at 2 m high (m s-1); Rn is the total daily net radiation at the crop surface (MJ m-2day-1); ∆ is the slope of vapor pressure curve (kPa °C-1) (Equation 3.3); γ is the psychometric constant (kPa °C-1) (Equation 3.4); and G is the soil heat flux density (MJ m-2day-1), but was ignored for daily calculation (Allen et al. 1998).  Equation 3.3  ∆= DBLE B. QPBE×R PN.MNSSTM:N.:(S + M:N. P:)M  Equation 3.4  K = B. QQU×PBV:×O    60 The next step was to determine potential crop evapotranspiration (ETc) of each crop species at each growing stage. Based on FAO Report-56 (Allen et al. 1998) and previous studies conducted at the same location (personal communication, Derek Hunt, 2018), I divided corn growing season into five stages: fallow (pre-planting), initial, development, mature, and end (post growing season) (Appendix D). FAO Report-56 (Allen et al. 1998) provided a basal crop coefficient (Kcb) of silage corn each growing stage and daily Kcb was calculated assuming the change of Kcb over time was linear (Appendix D). There is one corn growing cycle per year. Perennial tall fescue was a 5 years old stand harvested multiple times per year (3 and 5 in this study). Moreover, based on the assumption that grass crop would remain short after the last cut in October, and regrowth would not start until the following March when cumulative degree-days reach 300, I set the Kcb of tall fescue and Italian ryegrass at 0.85 during non-production season (personal communication Derek Hunt, 2018, Appendix D).  Height and rooting depth of plants were assumed to change during growing stages at the same rate as Kcb. Finally, Kcb was adjusted for local climate based on the plant height, daily relative humidity and daily wind speed at 2 m above ground (Equation 3.3) (Allen et al. 1998). Equation 3.5  W08 = W08	(@#8) + [B. BD <M − M − B. BBD(GY9)$ − DU)](.:)B.: where, Kcb (Tab) is the value for Kcb at different growing stage taken from Appendix D; u2 is the mean value for daily wind speed (m s-1) at 2 m height above ground; RHmin is the mean value for daily minimum relative humidity, and h is the mean plant height (m).   61 Table 3.1 Soil hydraulic properties of each layer. Hydraulic parameters are explained in the text below.   The soil evaporation coefficient (Ke), describes the evaporation component of ETc. Where the topsoil is wet, following rain or irrigation, Ke is maximal. Where the soil surface is dry, Ke is small and even zero when no water remains near the soil surface for evaporation. Ke was calculated as, Equation 3.6  W( = 9)$(W1 W0	9#3 − W08 , 2(\W0	9#3) where, Ke soil evaporation coefficient; Kcb basal crop coefficient; Kc max maximum value of Kc following rain or irrigation; Kr dimensionless evaporation reduction coefficient dependent on the cumulative depth of water depleted (evaporated) from the top layer; few is the fraction of the soil that is both exposed and wetted by rainfall or irrigation, i.e., the fraction of soil surface from which most evaporation occurs.  With Kcb and Ke known, daily ETc of a crop can be determined using the dual crop coefficient equation (Equation 3.7) for any given date. Evaporation component (E) and crop LayersL1 L2 L3Depth	(d) mm 150 150 300Bulk	density	(Db) g	cm-3 1.17 1.21 1.32Soil	organic	matter	content % 6 5 1%	Sand-Silt-Clay - 12-68-20 13-68-19 13-71-16Soil	water	content	at	field	capacity	(θFC		) m3	m-3 0.45 0.43 0.34Soil	water	content	at	permanent	wilting	point	(θPWP)	 m3	m-3 0.19 0.17 0.11Available	water	holding	capacity	(AWHC) mm 40 39 67Bounded	water	(BW) mm 28 26 34Field	capacity	(FC) mm 68 65 101Root	density	coefficient	of	corn	(Kd	corn) - 0.5 0.3 0.2Root	density	coefficient	of	grass	(Kd	grass) - 0.5 0.4 0.1Soil	hydraulic	properties Unit  62 transpiration (Tc) component can be further separated as KeETo and KcbETo, respectively, when necessary. Sources of all parameters involved in the calculation of ETc are listed in appendix E. Equation 3.7  ?@0 = ? + @] = W08?@+ + W(?@+ = W08 + W( ?@+ The third step of SWB simulation was acquiring soil hydraulic properties (Table 3.2). Given the depth of each soil layer (DL, mm), field capacity (^_` , m-3m-3), and permanent wilting point (^aba, m-3m-3), bounding water (BW, mm) was determined as, Equation 3.8  cd	 99 = 74×eOdO where, BW is the water content in a soil layer that is not accessible for plant root withdraw, gravitation movement, or capillary movement; and L (1, 2 or 3) indicates layer designation. Field capacity (FC) was calculated as, Equation 3.9  f]	 99 = 74×ef] where FC is the maximum amount of water that a layer of soil can hold after excess water percolated to the layer below. Available water holding capacity (AWHC, mm) was calculated as, Equation 3.10  gdY]	 99 = 74×(ef] − eOdO) where AWHC is the total amount of water in this soil layer that is available for plant roots to withdraw.   The last step of the SWB model is to simulate the downward movement of excess soil water from each layer and eventually determine the volume of potential leached water (L ha-1)   63 from the 60-cm depth of soil profile. At any timestep (daily), the water content in soil layer L1 (W1) was calculated as, Equation 3.11  dP = gdP + O + h − (? + @P + G<$+22) where, AW1 is the available water in L1 at timestep; P is precipitation (mm); I is irrigation (mm); E is the evaporation (mm), which is assumed to happen only in 0 to 10 cm depth of L1; T1 is transpiration of L1, which equals to TcKd1 (root density coefficient, Table 3.1); and runoff is assumed to be zero since the field has less than 2% of slope gradient. Water available to leach (WAL) from L1 to L2 was then calculated as,   Equation 3.12  dg4P = dP − f]P where, if WAL is less than zero, there will be no downward movement of water from L1 to L2 and on the next time step, i.e. day, AW1 will equal to W1 of the previous timestep. If WAL is greater than zero, the excess water will percolate into L2 and AW1 of the next day will equal to FC1. The water content in L2 (W2) was then calculated as, Equation 3.13  dM = gdM +dg4P − @M where AW2 is the available water in L2 at this timestep moment; WAL1 is the excess water percolated into L2 from L1; T2 is transpiration from L2, which equals to TcKd2 (root density coefficient, table 3.1). However, T2 equals to zero if crop rooting depth (Zr, Appendix D) has not reached L2 this time step, which means Tc would take 100% of its water demand from L1 instead of the root density coefficient (Kd) assigned to L1, e.g. 50% of corn plant transpiration would withdraw water form L1 layer (Kd1 = 0.5 for corn, Table 3.1). Same as how WAL1 was   64 determined, WAL2 would take place when W2 is greater than FC2. WAL3 is the amount of excess water that would percolate through the 60-cm deep soil profile and potentially leaching into groundwater. WAL3 was calculated as  Equation 3.14  dg4: = d: − f]: = 	 (gd: +dg4M − @:) −	f]: where AW3 is the available water in L3 at this time step; WAL2 is the excess water percolated into L3 from L3; T3 is the transpiration of L3, which equals to TcKd3 (root density coefficient, Table 3.1) only when crop rooting depth reached L3, otherwise, it was zero. Therefore, potential leaching events would only take place when the entire soil profile, 0-60 cm, was at field capacity. WAL3 was converted to L ha-1 for the convenience of further calculation.  Daily potential leached NO3- –N (kg ha-1) was the product of the simulated volume of leached water (L ha-1) from 60-cm depth times concentration of NO3- –N (mg L-1) in soil pore water samples collected by suction lysimeters from 50-cm soil depth. The concentration of NO3- –N between water samplings was calculated using linear interpolations from the samples taken before and after that day. Annual cumulative NO3- –N leaching was calculated by adding all daily leaching between March 2017 and March 2018.  Leaching proportion (unit-less) was calculated as annual cumulative NO3- –N leaching loss divided by total N applied. Leaching intensity (g N kg DM-1) for corn or grass was calculated as the annual cumulative NO3- –N leaching loss divided by annual total harvested biomass. 3.2.4 Gas emission sampling and quantification of N2O–N emission  Gas samples were obtained using vented dynamic chambers (Rochette et al. 2008). Often samples were taken on a weekly basis but more frequent sampling was done when high   65 emissions were anticipated such as after manure applications, major precipitation after drought, and freeze-and-thaw events. One square 60 cm x 60 cm aluminum collar was installed in each grass plot and three 15 cm x 30 cm collar were installed in corn plots, one between corn rows and two between plants across corn rows to include injection furrows. Vented and insulated lids (2.7 cm high) were placed in water channels on the collars which were filled with deionized water to create an airtight seal during sampling. At each sampling, four 20 ml air samples were taken from each chamber at four 15-minute intervals using a syringe and injecting the sampled air into pre-evacuated 10-ml vials. At the same time, 4 ambient air samples were taken. The chamber volume was measured multiple times during the study period; the average volume for each chamber was then used in the N2O flux calculation. Gas emission rates were adjusted for ambient air temperature and pressure. Soil temperature was measured at 5- and 7.5-cm depths from three random points adjacent to the chamber. Gas samples collected in the vials were analyzed for N2O concentration using a gas chromatograph (7890B GC System, Agilent Technologies®) at ARDC. Nitrous oxide flux (g N2O m-2 min-1) was calculated by the change of N2O molar concentration (mole L-1) in the chamber as sampling time progressed (Equation 3.15) (Rochette et al. 2008), Equation 3.15  f%<3!MA = 6$6' ×i9!MA×	$!MAg ×O − (OO  where, jkjl  is the rate of change of N2O molar concentration (mol mol-1min-1), i.e. slope of the fitted curve; mnopq is the molar weight of N2O (44 g mol-1);	ropq is mole of N2O at time = 0; P is atmospheric pressure at time = 0; and eP is vapor pressure at time = 0; and A is the surface area   66 inside the collar (m-2). Moles of N2O in the chamber were calculated by multiplying N2O molar concentration analyzed by GC with the moles of air in the chamber, which was calculated using ideal gas law (Equation 3.16), Equation 3.16  $ = O;G@ Where, n = number of moles of air, P = atmospheric pressure (kPa); T = temperature in Kelvin; R = molar gas constant, 8.3144598 m2 kg s-2 K-1 mol-1; and V = volume of chamber (L). Measurements of change in N2O concentration in the chamber over the sampling period were fitted into a linear or non-linear model, choosing curve with greater R2. The slope of the non-linear curve was calculated as the first derivative of the equation at time = 0. Flux values were then converted to g N2O–N ha-1 day-1 and cumulative N2O–N emission over time (kg ha-1) was approximated by linearly interpolation between sampling dates.  We assumed that the measurements typically made between 0900 and 1200 h were good estimates of average daily N2O–N emission (Rochette et al. 2008). Emission factor was calculated as annual cumulative N2O–N, minus N2O–N emission loss from control plot divided by TN applied times 100. Annual cumulative N2O–N emission per crop yield (intensity, mg N kg DM-1) was also calculated. 3.2.5 Statistical analysis Effects of farmlet management scenario and crop species on annual cumulative NO3 –N leaching and N2O–N emission were analyzed using a linear mixed effects model, with management scenarios (four levels, F1, F2, F3, and F4) and crop (silage corn and tall fescue) as fixed effects, and block as a random effect. The experimental design was split plot with the crop   67 as the main plot and farmlet as subplot treatment with two error terms. For analysis of within crop effects, a randomized complete block design was used. Significance was set at a priori P<0.05, and farmlets were compared using the Fisher Protected LSD. I used Type III ANOVA with Satterthwaite’s method (R package ‘lmerTest’) (Kuznetsova et al. 2017) to determine significant treatment effects and where differences were found (P<0.05), estimated marginal means of each factor level were investigated using Fisher’s least squared difference test (R package ‘agricolae’ and ‘predictmeans’) (de Mendiburu 2017, Luo et al. 2018). When farmlet crop interactions were significant, the least squared means of each factor level were tested for each crop separately. I also conducted an orthogonal contrast analysis to determine the effect of Farmlet 1 and 2 vs 3 and 4 (R package ‘emmeans’ and ‘car’) (Fox and Weisberg 2011; Lenth 2018). All data were tested for normality using the Shapiro-Wilk test, and where not normal was transformed either by logarithm or square root to meet normality assumptions. All analyses were carried out in R (R Core Team 2017) For corn plots, the crop year of was from May to May based on the manure application and planting date, so that the relay crop was included with the previous corn crop when it was established. For grass plots, crop year was from March to March based the start of grass growth assumed to be when cumulative degree days reached 300 (Kowlenko 1987??). The actual dates of nutrient application and harvesting are shown on a timeline chart. For corn, the year was from 26 April 2017 to 24 April 2018 and 25 April 2018 to 25 April 2019. For grass the years ran from 24 March 2017 to 16 March 2018 and 16 March 2018 to 25 March 2019.  At the time of statistical analysis, the crop year 2018-19 has not ended, i.e. relay crop has not been harvested. Thus, I only present and discuss annual N losses from the crop year 2017-18. I did, however, analyze N2O–N emission during production season for two consecutive years. For these   68 production seasons, I conducted statistical analysis on cumulative emission between March and October for grass soil and April to September for corn soil on both years.   3.3 Results and Discussion 3.3.1 Potential nitrate leaching  Based on the simulation of soil water balance model, 0-60 cm soil profile reached field capacity in mid-October 2017. Excessive water from heavy precipitation started to percolate through the three soil layers defined in the SWB and became available to leach into groundwater (Figure 3.1).  On corn plots, NO3- –N concentration in soil water started to increase in Mid-October as a result of the increased soil moisture content, enabling microbial activity helping to mineralize organic N in the soil, manure and microbes, and nitrification of NH4 (Fig 3.2). The NO3- –N concentration of F3 and F4 peaked at 10 and 13 ppm respectively by the end of October, while NO3- –N concentration of F1 and F2 kept increasing until the third week of November and maintained a concentration two times higher than F3 and F4 for nearly two months (Figure 3.3).  The decline in soil NO3- –N was due to leaching and probably a declining amount of new NO3 due to depleting substrate and lower temperatures. Cumulative NO3- –N leaching of F1 and F2 sharply increased in November due to high NO3- –N concentration (Fig 3.2) and water percolation (not shown). Management scenario had a significant impact on the cumulative NO3- –N leaching (Table 3.2, 3.3) and when cumulative NO3- –N leaching of all farmlets plateaued in February, corn plots with no relay crop (F1 and F2) lost nearly 4 times more NO3- –N than corn plots covered with a relay crop (F3 and F4) (Table 3.3). Recall from section 2.3.2, Italian grass, when intercropped with corn, captured an average of 63 kg N ha-1 from corn plots of F3 and F4 during the non-production season and the relay crop may have supported more soil microbes   69 through root exudates which could help tie up NO3. Kuo et al. (2005) reported that planting ryegrass as winter cover crops in Washington State could reduce NO3- –N concentration in leachate from 18.7 mg L-1 to 8.1 mg L-1 with total N fertilizer input of 201 kg ha-1. Perego et al. (2012) also found that silage corn-Italian ryegrass double cropping system in Italy reduced up to 90 % of NO3- –N leaching and annual leaching loss is as low as 14 kg ha-1 with mean annual NO3- –N fertilizer input of 314 kg ha-1. Eriksen et al. (2010) found that introducing annual ryegrass as a winter catch crop for grazing crop rotation also significantly decreased nitrate leaching. There was no significant difference between F1 and F2, indicating the manure sludge injection was did not affect potential NO3- –N leaching, despite the concentration of N in a band.  This could be due to the high C:N ration in the manure band sludge. Cumulative NO3- –N leaching from F4 was also not significantly different from F3 (Table 3.3), despite higher N uptake (Table 2.5) and suggesting that the DCD had decomposed over the summer (Amberger, 1989). Soil water NO3- –N concentration of grass plots dropped more quickly than corn plots. For F1 and F2, NO3-–N concentration in soil water peaked a month sooner than corn plots (Figure 3.4), which can be attributed in part to the manure application in September after the fourth harvest. As with corn, the amount of NO3- –N leaching from F1 and F2 were very similar throughout the winter, cumulatively 80 and 82 kg ha-1 respectively after 12 months (Table 3.3). In contrast for F1 and F2, rather than reaching a peak, NO3- –N concentration of F3 steadily increased throughout the non-production season. I also noticed that NO3- –N concentration of F4 started lower than F3, but stayed at around 10 ppm for the entire November and remained slightly higher than F3 for the rest of the non-production season (Figure 3.3). This slow response of NO3- –N concentration in soil water could be attributed to the addition of DCD to manure,   70 which likely preserved ammonium N and delayed nitrification when the soil moisture content is ideal for microbial transformation. Cumulatively, however, F3 and F4 lost the same amount of NO3- –N, 53 kg ha-1, during the non-production season of the crop year 2017-18 (Table 3.3). Several lysimeter studies by Di and Cameron (2003, 2005); Zaman and Blennerhassett (2010); and De Klein et al. (2011) have found that DCD could reduce NO3- –N leaching by 10 – 63 % when applied with urine to graze pasture systems. Similar to corn plots, cumulative NO3- –N leaching of grass plots also separated into two groups based on whether innovative cropping BMPs were implemented, in this case reducing the number of grass cuts. Although there was no significant difference across cumulative NO3- –N leaching of all four farmlets (Table 3.3), by calculating mean cumulative NO3- –N leaching of the 3-cut scenarios (F3 and F4) and 5-cut scenarios (F1 and F2), an orthogonal contrast analysis indicated significant decrease of cumulative NO3- –N leaching loss (53.8 vs 80.8, P < 0.05) when harvest frequency was reduced to 3 cuts per year.  71  Figure 3.1 Daily precipitation and water available to leach during the non-production season of 2017-18. Table 3.2 Analysis of variance of non-production season potential NO3-N leaching, leaching factor, and leaching intensity (n=4) in the crop year 2017-18, managed under four different scenarios (df =3) and two types of crops (df =1). Significant results are highlighted in bold. Variable F P-value F P-value F P-valueNO3-N	LossAnnual	leaching 17.49 <	0.01 0.02 0.88 4.09 <0.05Leaching	factor 26.61 <	0.01 50.56 <	0.01 11.57 <	0.01Leaching	intensity 17.25 <	0.01 23.74 <	0.01 0.12 0.95Management	scenario Crop	 Management*Crop  72  Figure 3.2 NO3- –N concentration in soil water (bottom) and cumulative potential NO3- –N leaching (top) of both crops during the non-production season of the 2017-18 crop year. Colors indicate different crop BMPs. Ribbon areas indicate standard errors (n=4).  73 Since they received about the same amount of total N, corn plots of F1 and F2 lost 106 and 110 kg ha-1 NO3- –N to groundwater, respectively, which accounted for 37% and 41% of TN applied in 2017. Potential NO3-N leaching loss from corn plots of F3 and F4 accounted for only 12% and 9% of annual TN applied (33 and 23 kg N ha-1), which were significantly lower than F1 and F2 (Table 3.3). Excluding yield of Italian ryegrass, for every 1 kg of corn silage produced on F3 and F4, 2 and 0.0013 kg NO3- –N could potentially leach into groundwater, which were significantly lower than the leaching intensity of F1 and F2 (5.1 and 5.5 g NO3- –N per kg of DM) (Table 3.3). When yield of Italian ryegrass was included, corn plots of F3 and F4 would produce comparable amount feed as F1 and F2 at even lower potential NO3- –N leaching intensities (1.7 and 1.1 g N per kg DM) despite the extra input of TN (117 kg ha-1) to the relay crop in April, though this leaching would also show up in the following year (Table 3.3). For grass crop, F1 and F2 lost 8 g of NO3- –N for every kg DM produced, while F3 and F4 only lost 4g of NO3- –N (Table 3.3). These data suggest that crop management more than nutrient management reduced no3 leaching. Also, unexpectedly, corn caused less leaching per unit of dry matter, due to its lower N requirement. Intercropping Italian ryegrass with silage corn and reducing the harvesting frequency of tall fescue both could greatly decrease potential NO3- –N leaching intensity from forage production. Nonetheless, despite the lower cumulative potential NO3- –N leaching from grass plots, a farmlet would actually lose have higher leaching intensity from grass plots than silage corn (6.1 vs. 3.5 g N per kg DM) (Table 3.3). The combined use of DCD and irrigation did not increase or decrease leaching form grass.   74 Table 3.3 Potential nitrate leaching during the non-production season of the 2017-18 crop year. Values in parentheses are the standard error. n = 4 for each farmlet and n=16 for crop mean. Differences between crop means were not discussed in the text when crop farmlet interaction is significant (P<0.05). Crop Management	scenarioTotal	N	appliedCrop	N	yield Crop	yieldkg	N	ha-1 kg	N	ha-1 t	DM	ha-1Corn F1 282 170 20.8 106	(17.5) a† 37% a 5.1	(0.7) aF2 271 181 20.0 110	(19.7) a 41% a 5.5	(0.9) a(relay	crop	excluded) F3 278 178 16.8 33	(9.1) b 12% b 2.0	(0.5) b(relay	crop	excluded) F4 264 199 17.8 23	(2.2) b 9% b 1.3	(0.1) b(relay	crop	included) F3 401 238 19.4 33	(9.1) b 8% b 1.7	(0.9) b(relay	crop	included) F4 387 264 20.7 23	(2.2) b 6% b 1.1	(0.2) bCtrl 0 85 10.0 38	(10.4) - 3.8	(1.1)LSD0.05 - - - 34 11% 1.6Grass F1 586 229 9.8 80	(13.4) a 14% a 8.2	(1.4) aF2 707 255 9.9 82	(8.8) a 12% ab 8.4	(1.0) aF3 694 271 12.7 53	(16.5) a 8% b 4.2	(1.3) bF4 684 303 13.8 53	(4.3) a 8% ab 3.9	(0.4) bCtrl 0 62 4.7 0.4	(0.1) - 0.1	(0.03)LSD0.05 - - - 39 6% 3.6	5-cut 647 242 10 81 a 13% a 8.3 a3-cut 689 287 13 54 b 8% b 4.1 b5-cut	vs.	3-cut	§	 P	<	0.01 P	<	0.01 P	<	0.01 P	=	0.04 P	=	0.02 P	<	0.01Corn Mean 274 182 19 68 25% 3.5	(0.6) B‡Grass	 Mean 668 265 12 67 10% 6.1	(0.7) ALSD0.05 - - - 16 4% 1.1§	An	orthogonal	contrast	analysis	between	the	least	square	means	of	3-cut		and	5-cut	scenarios.Leaching	intensityg	N	(kg	DM)-1P	<	0.01 P	<	0.01Leaching	factor%Annual	NO3-N	leachingkg	N	ha-1†	Different	lower	case	letters	behind	values	in	each	coloum	indicate	significant	difference	between	farmlets	of	the	same	crop	(LSD,	P<0.05).	‡	Different	upper	case	letters	behind	values	in	each	coloum	indicate	significant	difference	between	means	of	each	crop	¶	NS:		interaction	between	crop	and	farmlet	is	not	significant.	NS¶Crop	x	Farmlet  75 3.3.2 Nitrous oxide emissions In contrast to the NO3- –N leaching, most of the annual N2O–N emission took place in production season, i.e. from March to end of October (Fig 3.4 and 3.5). N2O–N fluxes from corn plots of all farmlets increased soon after manure application and peaked at the highest daily flux of 2017 and 2018, respectively, 200 and 602 g N ha-1 day-1 (Figure 3.4 & 3.5). Daily fluxes fell back to a pre-application level within 6 weeks and the cumulative N2O–N emission during this 6-week period accounted for more than 50% of the cumulative annual N2O–N emissions. Another spike of N2O–N flux occurred soon after urea side-dressing application in mid-June of both years (Figure 3.4). An increase of daily N2O–N fluxes from grass plots also corresponded to manure application following each harvest event (Figure 3.4 & 3.5). Grass plots received 668 kg ha-1 total N (TN) and 391 kg ha-1 total ammonium N (TAN) (2.3 times more than corn plots) in the production season of 2017 through 4 to 6 separated applications depending on the harvesting frequency. Multiple peak N2O–N fluxes were observed with the highest being 890 g N2O–N ha-1day-1 from F3, which happened on June 9th, a week after the third manure application with high rates also on the date prior and after this peak (Figure 3.4). In the 2018 production season, grass plots received 551 kg ha-1 TN and 314 kg ha-1 TAN (1.8 times more than corn plots). Farmlet 2 had the highest peak flux, 334 g N ha-1 day-1, on June 25th, after the third manure application (Figure 3.5).   76  Figure 3.3 2017-18 production season cumulative N2O–N emission (top) and daily N2O–N fluxes (bottom) from two crops of four farmlets. Ribbon areas indicate standard errors of each farmlet (n=4).   77  Figure 3.4 2018-19 production season cumulative N2O–N emission (top) and daily N2O–N fluxes (bottom) from two crops of four farmlets. Ribbon areas indicate standard errors of each farmlet (n=4).   78 Management scenario had a significant impact on cumulative N2O–N emission and the interaction between the management scenario and type of crop was also significant (Table 3.4). Cumulative N2O–N emission during production season of 2017 was the highest in the corn plot of F2, 3.5 kg N ha-1 (P<0.05) (Table 3.5). Cumulative N2O–N emission of F2 was significantly higher than F1, which could be attributed to the injection of sludge with a high C content. Sludge injection was designed to increase P accessibility by placing manure closer to juvenile roots of corn plants. However, incorporating manure into the soil would conserve more NH4+–N from volatilization and promote the nitrification process, which subsequently generated higher N2O–N fluxes. In a laboratory study (Flessa and Beese 2000) found that when cattle slurry injected into moist soil, the concentrated slurry will cause temporary anaerobic condition at the injection sites and subsequently promote N2O production. Elevated N2O–N emission was also observed in field study when liquid dairy slurry was injected before planting corn (Cambareri et al. 2017). N2O–N emission of F3 and F4 accounted for 0.5% and 0.2% of the TN applied, which were significantly less than F2 (P<0.05). This was likely due to the enhanced removal of N by relay crop (F3) and adding DCD to manure (F4).  Among all grass plots, F3 emitted the highest amount of N2O–N, 6.6 kg N ha-1 (P<0.05), during production season of 2017, while F4 emitted the lowest amount of N2O–N, 1.6 kg N ha-1 (P<0.05).  Although the farmlets received similar amounts of both TN and TAN, the emission factor of F3, 0.95%, was 4 times higher than F4 (0.23%). Comparing these two 3-cut farmlets (F3 and F4), the combination of adding DCD plus irrigation reduced N2O–N emission production emission factor substantially. The emission factor of F3 was almost twice as high as F2 (0.50%) (P<0.05, Table 3.5). However, nearly 50% of the cumulative production-season emission on F3 was generated during the highest recorded peak flux, 890 g N2O–N ha-1 day-1, of   79 this two-year study. This concentrated peak flux could be the result of manure application happened a week before 9.2 mm rainfall which occurred a day before the peak flux. Extensive studies have established the critical impact of soil WFPS, temperature, and soil NO3- –N content (Linn and Doran, 1984; Dobbie and Smith, 2003; Bateman and Baggs 2005).   Grass plots of all farmlets generated a similar amount of N2O–N emission during the 2018 production season with the emission factor range from 0.5% to 0.8% which was similar to 2017-18. Cumulative N2O–N emission from corn plots was highest on F3 (7.7 kg N ha-1, P<0.05) but there was little difference among the other three farmlets. Cumulative N2O–N emission of F3 increased substantially compared to 2017 (7.7 vs. 2.2 kg N ha-1, Table 3.5), which was very likely due to nitrification of residual ammonium N from manure applied on relay crop (why is this likely?). Since the crop year 2017-18 was the first year of the study, there was no manure application on relay crop from the previous year. However, during a 6-week period between March 12th and May 1st, F3 and F4 received 123 kg N ha-1 on Italian ryegrass at the end of the crop year 2017-18, and 289 kg N ha-1 before corn seeding at the beginning of the crop year 2018-19. Peak N2O–N flux from corn plots of F3 was the highest of all four farmlets in 2018 and also 6 times higher than its peak flux at the same time of previous year (702 vs. 118 g N2O–N ha-1 day-1, Table 3.5). Peak N2O–N flux of similar magnitude was not observed from F4 (81 g N2O-N ha-1 day-1) despite receiving the same amount the TN during this period, which could be attributed to the DCD added to manure. Nonetheless, cumulative N2O–N emission from F4 in 2018 production season increased substantially compared to 2017 (3.9 vs. 1.5 kg N ha-1) which could be the combined result of additional N input during the non-production season and irrigation in summer (Table 3.5). In summer 2017, corn plots of F4 were not irrigated due to a lack of equipment. Comparing the irrigation period, July 19th to September 6th, of both years,   80 average daily N2O–N flux in summer 2017, 2.4 g ha-1 day-1, was tripled in summer 2018, 7.3 g ha-1 day-1, which indicates that irrigation would stimulate nitrification when the soil moisture content is low in the summer. Thus, combining DCD with irrigation could reduce the magnitude of peak N2O–N flux immediately after manure application, but potentially increase daily N2O–N flux later in the season when elevated soil moisture content enhances nitrification and denitrification (Davidson 2010). A number of studies in New Zealand (de Klein et al., 2011; Luo et al. 2015) have also found that 10 kg ha-1 of DCD mixed with dairy urine reduced peak N2O–N flux by 120-180 g N ha-1 day-1 in the first 4-6 weeks after urine application and also decrease emission factor by 60%.  Table 3.4 Analysis of variance of cumulative N2O-N emission and emission factor (n=4) in two consecutive production seasons and annual cumulative emission of the crop year 2017-18, managed under four different scenarios (df =3) and two types of crops (df =1). Significant results are highlighted in bold Variable F P-value F P-value F P-value2017	Production	seasonCumulative	emission 28.64 <	0.01 28.66 <	0.01 19.75 <	0.01Adjusted	emission 31.50 <	0.01 63.38 <	0.01 21.73 <	0.01Emission	factor 19.40 <	0.01 0.62 0.43 11.70 <	0.012018	Production	seasonCumulative	emission 11.53 <	0.01 3.96 0.05 8.59 <	0.01Adjusted	emission 11.86 <	0.01 0.43 0.51 8.80 <	0.01Emission	factor 19.50 <	0.01 51.90 <	0.01 17.05 <	0.012017-18	AnnualCumulative	emission 25.05 <	0.01 4.62 0.04 14.00 <	0.01Adjusted	emission 25.05 <	0.01 33.78 <	0.01 14.00 <	0.01Emission	factor 16.08 <	0.01 2.78 0.11 8.61 <	0.01Emission	intensity 26.79 <	0.01 68.17 <	0.01 13.79 <	0.01Management	scenario Crop	 Management*Crop  81 Table 3.5 Cumulative nitrous oxide emission and emission factor of two consecutive production seasons. n = 4 for each farmlet and n=16 for crop mean. Differences between crop means were not discussed when crop farmlet interaction is significant (P < 0.05).  2017 2018 2017 2018Corn F1 282 312 187 202 2.1	(0.17) b† 3.0	(0.75) b 1.3	(0.07) b 2.2	(0.66) b 0.45% b 0.71% bF2 271 288 169 170 3.5	(0.50) a 3.5	(0.39) b 2.6	(0.50) a 2.7	(0.50) b 0.99% a 0.94% bF3 278 289 173 171 2.2	(0.39) b 7.7	(0.44) a 1.4	(0.20) b 6.9	(0.50) a 0.49% b 2.40% aF4 264 288 165 170 1.5	(0.07) b 3.9	(0.36) b 0.6	(0.18) b 3.2	(0.29) b 0.23% b 1.10% bCtrl 0 0 0 0 0.9	(0.12) 0.8	(0.22) - - - -LSD0.05 - - - - 0.9 1.7 0.9 1.7 0.37% 0.58%Grass F1 586 543 345 299 2.1	(0.60) c 3.1	(0.22) a 1.8	(0.60) c 2.8	(0.21) a 0.31% bc 0.52% aF2 707 537 407 310 3.8	(0.21) b 4.9	(1.21) a 3.5	(0.21) b 4.6	(1.18) a 0.50% b 0.85% aF3 694 559 409 327 6.9	(0.83) a 4.1	(0.40) a 6.6	(0.83) a 3.8	(0.43) a 0.95% a 0.67% aF4 684 564 404 320 1.9	(0.23) c 3.2	(0.28) a 1.6	(0.24) c 2.9	(0.29) a 0.23% c 0.52% aCtrl 0 0 0 0 0.3	(0.01) 0.3	(0.04) - - - -LSD0.05 - - - - 1.6 2.3 1.6 2.3 0.25% 0.42%Corn Mean 274 294 174 178 2.3	(0.23) 4.5	(0.58) 1.5	(0.23) 3.7	(0.58) 0.54% 1.29%Grass Mean 668 551 391 314 3.7	(0.57) 3.8	(0.36) 3.4	(0.57) 3.5	(0.35) 0.49% 0.64%LSD0.05 - - - - 0.6 0.9 0.6 0.9 0.14% 0.22%- - - - P	<	0.01 P	<	0.01 P	<	0.01 P	<	0.01 P	<	0.01 P	<	0.01†	Different	Lower	case	letters	behind	values	in	each	coloum	indicate	significant	difference	between	farmlet	under	the	same	crop	(LSD,	P<0.05).	Production	season	N2O-N	emission	factor2017 2018%kg	N	ha-1 kg	N	ha-1Production	season	N2O-N	emissionEmission	corrected	for	control2017 20182017 2018Crop	x	FarmletTAN	appliedTN	appliedkg	N	ha-1 kg	N	ha-1CropManagement	Scenario  82 In terms of annual N2O–N emission of 2017-18, when relay crop N yield and TN input were included, corn plot of F2 and F4 generated the highest and lowest amount of annual cumulative N2O–N emission, 4.4 and 2.6 kg N ha-1, respectively, while emissions from F1 and F3 fell in-between (Table 3.7). Despite the additional 117 kg ha-1 TN applied on the relay crop before seeding corn for the next year, F3 and F4 lost only 0.5% and 0.3% of its annual TN applied as N2O–N emission, which was significantly lower than the emission factor of F1 (1.1%, Table 3.7). Also, corn plots of F3 and F4 generated 65% and 57% of their annual N2O–N emission during the production season, which was much lower than F1 and F2, 75% and 78%, respectively (Table 3.7). Moreover, there were multiple spikes of daily N2O–N flux during the non-production season of 2017-18, which might be caused by various reasons, such as increasing of water-filled soil pore space, freeze-and-thaw events, and manure application on the relay crop (Figure 3.6). Comparing the N2O–N emission intensity across four farmlets, F4 generated the least amount of N2O–N, 127 mg, for every kilogram of feed produced from corn plot, which was similar to F1 (139 mg N per kg DM), but much less than F3 (181 mg N per kg DM) and significantly less than F2 (220 mg N per kg DM, P<0.05) (Table 3.7).  Annual cumulative N2O–N emission from grass plots was the highest on F3 and lowest on F4, 7.3 and 2.1 kg N ha-1 (P<0.05), respectively, and an average of 92% of annual emission took place during production season (Table 3.6). Taking crop yield into consideration, F3 had the lowest emission intensity, 571 mg N2O–N emitted for every kilogram DM of tall fescue production, which was significantly higher than F2 (415 mg N per kg DM) and F4 (151 N per kg DM) (Table 3.6). Also, grass plots have substantially higher emission intensity than corn plots of all farmlets.  83 Table 3.6 Annual cumulative nitrous oxide emission, emission factor, and emission intensity of the crop year 2017-2018. n = 4 for each farmlet and n=16 for crop mean. Difference between crop means are not discussed in the text when crop farmlet interaction is significant (P < 0.05).  Total	N	appliedCrop	N	yield Crop	yieldkg	N	ha-1 kg	N	ha-1 t	DM	ha-1Corn F1 282 170 20.8 2.7	(0.25) b 81% a 1.5	(0.25) b 0.53% b 128(12.6) bcF2 271 181 20.0 4.1	(0.51) a 84% a 2.9	(0.51) a 1.10% a 205(26.3) a(relay	crop	excluded) F3 278 178 16.8 2.9	(0.32) b 78% a 1.7	(0.32) b 0.61% b 174(22.9) ab(relay	crop	excluded) F4 264 199 17.8 2.0	(0.09) b 75% a 0.8	(0.09) b 0.31% b 112(3.6) cCtrl 0 85 10.0 1.2	(0.12) 73% - - 118(15.9)LSD0.05 - - - 1.0 11% 1.0 0.42% 55Corn F1 282 170 20.8 2.9	(0.27) b† 75% ab 1.5	(0.27) b 0.52% b 139(13.3) bF2 271 181 20.0 4.4	(1.04) a 78% a 3.0	(0.52) a 1.11% a 220(27.3) a(relay	crop	included) F3 401 238 19.4 3.5	(0.36) ab 65% bc 2.0	(0.36) ab 0.51% b 181(24.0) ab(relay	crop	included) F4 387 264 20.7 2.6	(0.11) b 57% c 1.2	(0.11) b 0.30% b 127(6.4) bCtrl 0 85 10.0 1.4	(0.19) 60% - - -LSD0.05 - - - 1.1 10% 1.1 0.39% 56Grass F1 586 229 9.8 2.3	(0.61) c 90% b 2.0	(0.61) c 0.33% bc 235(57.7) cF2 707 255 9.9 4.1	(0.21) b 93% ab 3.7	(0.21) b 0.53% b 415(25.2) bF3 694 271 12.7 7.3	(0.96) a 95% a 6.9	(0.96) a 1.00% a 571(65.0) aF4 684 303 13.8 2.1	(0.22) c 90% b 1.7	(0.22) c 0.25% c 151(21.2) cCtrl 0 62 4.7 0.4	(0.003) 79% - - -LSD0.05 - - - 1.7 4% 1.7 0.27% 135Corn Mean - 182 19 2.9	(0.25) 69% 1.9	(0.25) 0.61% 167(12.6)Grass Mean - 265 12 4.0	(0.60) 92% 3.6	(0.60) 0.53% 343(46.8)LSD0.05 - - - 0.60 3% 0.60 0.14% 44P	<	0.01 P	<	0.01 P	<	0.01 P	<	0.01 P	<	0.01†	Different	Lower	case	letters	behind	values	in	each	coloum	indicate	significant	difference	between	farmlet	under	the	same	crop	(LSD,	P<0.05).	mg	N	(kg	DM)-1Crop Management	scenarioCrop	x	Farmletkg	N	ha-1 % kg	N	ha-1 %Annual	N2O-N	emissionProduction	season	emission	as	%	of	annual	emissionAnnual	emission	corrected	for	controlEmission	factorEmission	intensity  84  Figure 3.5 Annual cumulative N2O–N emission (top) and daily N2O–N fluxes (bottom) from two crops of four farmlets. Ribbon areas indicate standard errors of each farmlet (n=4).   85 3.3.3 Farmlet area-weighted total N losses When crop area ratio equals to 50:50, F2 lost 0.82% of total N applied as N2O –N emission, which was twice as much as the emission factor of F1 (P<0.05, Table 3.8). At a similar level of annual crop production and TN recovery efficiency, N2O –N emission intensity of F2 was significantly higher than F1 (285 vs. 170 mg N per kg DM) (Table 3.8). As I expected, implementing innovative nutrient BMPs elevated N2O –N emission, because dual-stream manure separation and assisting infiltration manure application methods would reduce NH3 volatilization and promote NH4+–N entering the soil. Comparing to F2, implementing relay crop and reducing grass harvest frequency on F3 did not affect the N2O –N emission factor and N2O –N emission intensity substantially. However, F3 produced a similar amount of feed as F2 (16.1 vs 15.0, LSD0.05 = 1.15 tonnes) in 2017-18 while losing less than half of the potential NO3- –N leaching from F2 (43 vs. 99 kg NO3- –N ha-1, P<0.05) (Table 3.8), which indicated that integrated innovative cropping BMPs, such as planting a relay crop and reducing grass harvest frequency reduced potential total NO3- –N leaching significantly but did not affect farmlet total N2O –N emission. Moreover, received similar amount of TN as F3 in 2017-18 crop year, F4 lost similar amount of NO3- –N through leaching, but achieved substantially lower N2O –N emission factor (0.26 % vs. 0.82 %) and N2O –N emission intensity than F3 (135 vs. 334 mg N2O –N per kg DM) (Table 3.8). This significant improvement implies that advanced techniques, e.g. irrigation and DCD, were very effective in reducing N2O –N emission but inconsequential on reducing NO3- –N leaching loss. Taking conventional scenario (F1) as the baseline, advanced technique scenario (F4) produced 2 t DM ha-1 more feed at a 7% higher N recovery efficiency while generating 35 mg less N2O –N (LSD0.05 = 65 mg) and losing 3.9 g less NO3- –N for one kilogram of feed   86 produced (Table 3.8). Hence, integrating all the available BMPs in this study substantially improved almost all the aspects of a dual-crop forage production system I measured in this study.   As I reported in the previous chapter (section 2.3.5), by adjusting the crop area ratio to 60:40, crop yield of all farmlet would increase 900 kg ha-1 with a 2.3% higher N recovery rate. However, crop area ratio did not have a significant impact on area-weighted total N2O–N emission and NO3- –N leaching (Table 3.7). Annual N2O–N emission and potential NO3- –N leaching of 2017-18 did not change when the silage corn area was adjusted to 60 %. Table 3.7 Analysis of variance of area-weighted total annual cumulative N2O–N emission, emission intensity, emission factor, annual cumulative NO3- –N leaching, and leaching intensity (n=4) in the crop year 2017-18, managed under four different management scenarios (df =3) and two land allocation scenarios (df =1). Significant results are highlighted in bold. Variable F P-value F P-value F P-valueFarmlet	area-weighted	totals	Annual	N2O-N	emission 47.15 <	0.01 0.10 0.76 0.30 0.83Annual	N2O-N	emission	intensity 61.89 <	0.01 1.94 0.18 0.44 0.73Annual	N2O-N	emission	factor 36.18 <	0.01 0.02 0.88 0.26 0.85NO3-N	Leaching 99.35 <	0.01 0.01 0.94 0.24 0.87NO3-N	Leaching	intensity 113.30 <	0.01 1.25 0.27 0.00 1.00Management*Area	ratioManagement	scenario Crop	area	ratio  87 Table 3.8 Annual cumulative nitrous oxide emission, emission factor, and emission intensity of the crop year 2017-2018. n = 4 for each farmlet and n=16 for crop mean. Difference between crop means is not discussed when crop farmlet interaction is significant (P < 0.05). Crop	area	ratio	(Corn	:	Grass) FarmletTN	Applied%F1 434 15.3	(0.41) b† 200	(5.9) d 46.0% b 2.6	(0.39) c 1.7	(0.39) c 0.38% b 170	(22) b 95	(11.5) a 6.2	(0.7) aF2 489 15.0	(0.36) b 218	(5.1) c 44.5% b 4.2	(0.26) b 3.3	(0.26) b 0.67% a 285	(22) a 99	(5.7) a 6.6	(0.3) a50:50 F3 548 16.1	(0.54) b 255	(5.1) b 46.5% b 5.4	(0.55) a 4.5	(0.61) a 0.82% a 334	(27) a 43	(8.4) b 2.7	(0.5) bF4 535 17.2	(0.41) a 284	(3.5) a 53.0% a 2.3	(0.16) c 1.4	(0.16) c 0.26% b 135	(7) b 39	(1.5) b 2.3	(0.1) b	LSD0.05 NA 1.15 12.2 2.1% 1.1 1.2 0.23% 65 19 1.2F1 404 16.4	(0.45) b 194	(5.8) d 48.0% b 2.7	(0.36) b 1.6	(0.34) b 0.40% b 162	(19) b 98	(12.2) a 6.0	(0.7) aF2 445 16.0	(0.39) b 211	(4.3) c 47.3% b 4.3	(0.30) a 3.2	(0.30) a 0.72% a 269	(23) a 103	(8.5) a 6.4	(0.4) a60:40 F3 518 16.7	(0.61) ab 251	(5.3) b 48.5% b 5.0	(0.48) a 4.0	(0.53) a 0.77% a 299	(24) a 41	(7.6) b 2.5	(0.5) bF4 506 17.9	(0.55) a 280	(4.8) a 55.3% a 2.4	(0.14) b 1.4	(0.15) b 0.27% b 133	(6) b 36	(1.4) b 2.0	(0.1) bLSD0.05 NA 1.32 12.2 2.3% 1.1 1.1 0.23% 59 19 1.250:50 Mean 502 15.9	(0.30) B 239	(8.7) A 47.5% B 3.6	(0.36) A 2.7	(0.36) A 0.54% A 231 A 69	(8.0) A 4.5 A60:40 Mean 468 16.8	(0.30) A‡ 234	(9.5) B 49.8% A 3.6	(0.32) A 2.5	(0.32) A 0.53% A 216 A 69	(8.8) A 4.2 ALSD0.05 NA 0.46 4.5 0.8% 0.4 0.4 0.09% 23 7 0.4NS¶ NS NS NS NS†	Different	Lower	case	letters	behind	values	in	each	coloum	indicate	significant	difference	between	farmlet	under	the	same	crop	area	ratio	(LSD,	P<0.05).	‡	Different	upper	case	letters	behind	values	in	each	coloum	indicate	significant	difference	between	means	of	each	crop	area	ratio	(LSD,	P<0.05).	¶	NS:		interaction	between	crop	area	ratio	and	farmlet	is	not	significant.	Leaching	intensitykg	N	ha-1 kg	N	ha-1 % mg	N	(kg	DM)	-1 kg	N	ha-1 g	N	(kg	DM)	-1Annual	N2O-N	emissionAnnual	N2O-N	emission	corrected	for	controlEmission	factorEmission	intensityAnnual	NO3-N	leachingCrop	area	ratio	x	FarmletCrop	Yield Crop	N	removalTotal	N	Recoverykg	N	ha-1t	DM	ha-1  88 3.4 Conclusion There were great differences between the N field-losses through N2O-N emission and potential NO3- –N leaching in terms of seasonality, the magnitude of flux, and response to nutrient and cropping BMPs. The majority of annual N2O-N emission happened during the production season immediately after soil received N influx, either manure or mineral fertilizers. N2O-N flux and cumulative emission were highly affected by the nutrient BMPs rather than cropping BMPs. Sludge injection and surface banding of separated liquid fraction, which were intended to assist in the incorporation of manure and reduce ammonium volatilization, significantly enhanced the production season cumulative N2O-N emission. Alternatively, this study did not show definitive impacts of implementing a relay crop and reducing grass harvest frequency on cumulative N2O-N emission. However, N from manure applied to the relay crop to enhance yield in spring would add to the emission of N2O-N in the following crop year. The combination of irrigation and nitrification inhibitor, DCD, was shown to mitigate N2O emissions compared to injected sludge but increased the cumulative emission potentially by delaying the nitrification process later in the growing season when the soil moisture content was increased by irrigation.  Potential NO3- –N leaching events often coincide with heavy rainfall in early fall, when the concentration of residual nitrate is high after the summer drought. Potential NO3- –N leaching was more subjected to the impact of cropping BMPs rather than nutrient BMPs. Based on data from one non-production season, intercropping Italian ryegrass with silage corn decreased the potential leaching of NO3- –N substantially. When the yield of relay crop was included, farmlets with relay crop would lose 30% or less NO3-N through leaching per DM production as farmlets without relay crop. Reduced grass harvest frequency also reduced potential leaching of NO3- –N   89 by 27 kg ha-1 during the non-production season of 2017-18. When the yield improvement of reduced harvest frequency was accounted for, 3-cut farmlets only lost about half as much NO3- –N through leaching (i.e. leaching intensity) as 5-cut farmlets for every kilogram of feed produced.  When N loss fluxes from both crops were combined at 50:50 area ratio as a dual-crop forage production system, potential leaching of NO3- –N was a bigger concern than N2O-N emission in terms of loss of PAN, since it was always one or two orders of magnitude greater. However, both N loss pathways have a negative impact on the environment and it is very difficult to compare their environmental impact. Implementing innovative nutrient management scenario doubled the cumulative N2O-N emission and emission intensity, which was a trade-off of reducing NH3 volatilization loss. The two N field-loss pathways I monitored in this study, i.e. NO3- –N leaching and N2O-N emission, responded to the integration of BMPs very differently. Integration of all the BMPs available in this study, i.e. advanced technique scenario (F4), achieved the highest N loss cutback.  Moreover, reallocating 10% of land space to grow more silage corn did not change total NO3- –N leaching and N2O-N emission substantially. In conclusion, nutrient and cropping BMPs were effective on different N loss pathway and growing larger area of silage corn would not reduce N losses.     90 Chapter 4: Conclusion 4.1 Research Conclusion In this thesis, I compared four management scenarios for dual-crop forage system “farmlets” that were incrementally improved with either nutrient or cropping BMPs or a combination of both over two consecutive crop years, 2017-18 and 2018-19. In study one (chapter 2), I quantified the annual crop dry-matter (DM) yield, crop N removal, and percentage recovery of TN applied of silage corn, Italian ryegrass, and tall fescue. I also examined how adjusting cropland allocation impacted the total crop dry-matter (DM) yield; crop N removal; recovery of TN. In study two (chapter 3), I quantified the seasonal and annual (the crop year 2017-18) N filed-losses through N2O−N emission and potential NO3-−N leaching. I also examined how adjusting cropland allocation would affect N field-loss for the four scenarios. 4.1.1 Crop N removal and total N recovery of experimental farmlets under different nutrient and cropping management scenarios Due to the intrinsic nature of two hybrids of corn used in this study, the long-season corn of farmlets (F) 1 and F2 produced significantly more whole plant silage in both years than the short-season corn of F3 and F4. However, when Italian ryegrass was implemented as a relay crop (F3 and 4), total harvested biomass from corn was similar across the four farmlets in the 2017-18 crop year. I found the combination of DCD and irrigation (F4) were effective in increasing DM yield of both silage corn whole plants and Italian ryegrass. As expected, silage corn of the most complex and advanced management scenario (F4) removed the highest amount of N and achieved the highest TN recovery efficiency over two years. Although the hypothesized incremental improvement was not observed, the lack of distinction between yield and N recovery of F1 and F2 indicated that the injection of manure sludge near the root zone can replace the   91 starter fertilizer of N and P. When the relay crop was included in 2017-18 crop year, crop N removal did not increase incrementally across all four scenarios, but annual N removal was improved by implementing a relay crop and was further improved by adding DCD and irrigation. Also, implementing relay crop with short-season corn (F3) did not increase the combined TN recovery from cornfield unless DCD and irrigation were added (F4).  Average annual forage grass production over two years did not increase incrementally across all four farmlets, because surface banding of the separated liquid fraction of manure slurry did not increase grass yield significantly. However, reducing harvest frequency to 3 cuts improved annual yield of tall fescue substantially and adding DCD and irrigation further increased annual yield of tall fescue. Average annual N removal by tall fescue over two years increased incrementally over two years when innovative nutrient BMPs and advanced techniques were added to the management scenario, but reducing annual harvest frequency to 3 times was inconsequential. Irrigation and DCD (F4) significantly improved TN recovery only in 2017. Integration of BMPs did not increase the percentage TN recovery of tall fescue in 2018 (F2 vs. F3 vs. F4). In 2017-18 crop year, farmlet area-weighted total N removal by all crops (including Italian ryegrass as relay crop) improved incrementally at default 50:50 crop area ratio and the effects of all management scenario were statistically significant. Farmlet area-weighted TN recovery was substantially higher in F4. The integration of nutrient and cropping BMPs increase TN recovery by 1.9 % (F3 vs. F2). These results suggested that by implementing the advanced technique scenario (F4), a dual-crop forage production dairy farm can produce 1.9 t ha-1 more feed and recover 84 kg ha-1 more N than the conventional scenario. If corn area is increased to   92 60% of the total farmland, average feed production of all farmlets, on average across all four scenarios, would increase by 0.9 t DM ha-1 yr-1.  4.1.2 Efficacy of integrated beneficial management practices for reducing potential nitrate leaching and nitrous oxide emission from dairy farm forage production systems Instead of incremental improvement as more BMPs were integrated into each of the four scenarios I compared, the N2O−N flux and cumulative emission were affected by the nutrient BMPs only and not the cropping BMPs. Sludge injection and surface banding of separated liquid fraction, which were intended to assist the incorporation of manure and reduce ammonium volatilization, enhanced production season cumulative N2O−N emissions. Implementing relay crop and reducing grass harvest frequency did not reduce cumulative N2O−N emission. However, residual N from manure application on the relay crop during the non-production season of 2017 exacerbated the peak flux and cumulative emission of N2O−N in the following year. Combination of irrigation and nitrification inhibitor, DCD, mitigated the exacerbated peak flux but increased the cumulative emission by delaying the nitrification process later in the growing season when the soil moisture content was increased by irrigation. Alternatively, potential NO3- –N leaching was more  affected by cropping BMPs than nutrient BMPs. Intercropping Italian ryegrass with silage corn decreased the potential leaching of NO3- –N substantially. When comparing leaching intensity with a yield of the relay crop included, farmlets with relay crop only lost < 30 % of NO3- –N through leaching while producing a similar amount of feed as farmlets without relay crop. Reducing the grass harvest frequency to 3 cuts annually also reduced potential leaching of NO3- –N by 23 kg ha-1 during the non-production season of 2017-18. Considering the yield improvement caused by reducing harvest   93 frequency, 3-cut farmlets lost only half as much NO3- –N through leaching as 5-cut farmlets for every kilogram of feed produced. Comparing the quantity of N losses from the two pathways that were monitored in this study, NO3- –N leaching was a bigger concern than N2O−N emission when a dual-crop forage production system consists 50 % silage corn and 50 % tall fescue. However, both NO3- –N leaching and N2O−N emission have a negative environmental impact and they are both tightly controlled by the local climate pattern. Dual-stream manure separation and assisting manure incorporation doubled the cumulative N2O−N emission and emission intensity as a tradeoff of NH3 abatement. Although NH3 loss was not quantified in this study, NH3−N loss reduced by manure separation and assisted infiltration are generally two orders of magnitude greater than the increase of N2O−N emission observed in this study (Webb et al., 2009; Bhandral et al., 2009). Implementing relay crop and 3-cut grass harvest frequency reduced potential NO3- –N leaching by 60 % but remained the same level of N2O−N emissions as F2. Adding irrigation and DCD reduced total N2O−N emissions and emission intensity effectively but did not further reduce NO3- –N leaching. Moreover, reallocating 10 % of land space to grow more silage corn did not affect NO3- –N leaching and N2O−N emission significantly. In conclusion, as more BMPs were integrated, total abatement of N field-loss was increasing, however, none of the BMPs can reduce N loss from all pathways due to tradeoffs. 4.2 Limitation and direction for future research Unfortunately, I was not able to include relay crop data, N2O–N measurement, and N2O–N measurement of the second non-production season (from November 2018 to April 2019) into my thesis. These results could have allowed me to complete the analysis of farmlet annual totals   94 for the second crop year, which would provide some insight into whether the effect of each management scenario is consistent over two years.  Secondly, it would have been very beneficial if I have included the nutrient recovery metrics for P and compared the effect of each management scenario on improving P use efficiency. According to Bittman et al. (2017), the LFV region and surrounding waterbodies is a sink of 7854 tonnes surplus P. Feed and fertilizer imports of dairy farms account for 1600 tonnes of net P influx, 1210 tonnes as feed and 390 tonnes as fertilizer for forage production, and a large portion of the fertilizer P would just accumulate in soil and become unavailable to crop uptake (Bittman et al. 2017b). Previous studies (Bittman et al. 2012; Schröder et al. 2015) have shown that dual-stream manure separation and precision placement of manure sludge can avoid over-application of P in forage grass and replace starter P fertilizer. Evidence also suggests cropping BMPs could potentially improve P recovery as well. Finally, nutrient quality analysis of my crop samples was not completed, which could have been used in a feed model to predict milk production, manure excretion rate, and determine nutrient use efficiency of the animal component of my farmlets. This information, together with the forage production data, could be utilized to evaluate the effect of management scenarios at whole-farm scale using whole-farm system models, e.g. Integrated Farm System Model (IFSM) (Rotz et al. 2011). This model, for example, would be capable to predict if integrated BMPs can actually reduce feed and fertilizer import while maintaining the same level of milk production and whether it would be economically feasible to do so (Rotz et al. 2010, 2013)  4.3 The implication of this study for dairy farms in the Lower Fraser Valley The available results from this thesis have shown that (1) injection of separated manure sludge near the root zone can replace application of starter fertilizer of N and P, which can   95 reduce both the cost of farm operations and nutrient surplus; (2) intercropping Italian ryegrass with silage corn as relay crop can be very effective on reducing NO3- –N leaching loss and also provide supplement to feedstuff; (3) although irrigation is not a typical practice for silage corn in the LFV, the combination of irrigation and DCD can increase crop yield and N recovery substantially; and (4) increasing area of silage corn can improve nitrogen recovery and increase feed production. Together these results provide an idea of the potential benefits of integrating a suite of BMPs for local dairy farmers to consider.    96 Bibliography Aarts, H.F.M., Habekotté, B., and Van Keulen, H. 2000. Nitrogen (N) management in the ‘De Marke’dairy farming system. Nutr. Cycl. Agroecosystems 56: 231–240. Springer. Alemu, A.W., Amiro, B.D., Bittman, S., MacDonald, D., and Ominski, K.H. 2017. Greenhouse gas emission of Canadian cow-calf operations: A whole-farm assessment of 295 farms. Agric. Syst. 151: 73–83. doi:https://doi.org/10.1016/j.agsy.2016.11.013. Allen, R.G., S, P.L., Raes, D., and Martin, S. 1998. Crop evapotranspiration : Guidelines for computing crop water requirements / by Richard G. Allen ... [et al.]. FAO Irrig. Drain. Pap. 56: 1–15. doi:10.1016/j.eja.2010.12.001. Amberger, A. 1989. Research on dicyandiamide as a nitrification inhibitor and future outlook. Commun. Soil Sci. Plant Anal. 20: 1933–1955. doi:10.1080/00103628909368195. Angers, D.A., Rochette, P., Côté, D., Bélanger, G., Massé, D., and Chantigny, M.H. 2007. Gaseous Nitrogen Emissions and Forage Nitrogen Uptake on Soils Fertilized with Raw and Treated Swine Manure. J. Environ. Qual. 36: 1864. doi:10.2134/jeq2007.0083. Aparicio, V., Costa, J.L., and Zamora, M. 2008. Nitrate leaching assessment in a long-term experiment under supplementary irrigation in humid Argentina. Agric. Water Manag. 95: 1361–1372. doi:10.1016/j.agwat.2008.06.003. Asing, J., Saggar, S., Singh, J., and Bolan, N.S. 2008. Assessment of nitrogen losses from urea and an organic manure with and without nitrification inhibitor, dicyandiamide, applied to lettuce under glasshouse conditions. Aust. J. Soil Res. 46: 535–541. doi:10.1071/SR07206. Bateman, E.J., and Baggs, E.M. 2005. Contributions of nitrification and denitrification to N2O emissions from soils at different water-filled pore space. Biol. Fertil. Soils 41: 379–388. doi:10.1007/s00374-005-0858-3.   97 Bhandral, R., Bittman, S., Kowalenko, G., Buckley, K., Chantigny, M.H., Hunt, D.E., Bounaix, F., and Friesen, A. 2009. Enhancing Soil Infiltration Reduces Gaseous Emissions and Improves N Uptake from Applied Dairy Slurry. J. Environ. Qual. 38: 1372. doi:10.2134/jeq2008.0287. Binnie, R.C., and Chestnutt, D.M.B. 1991. Effect of regrowth interval on the productivity of swards defoliated by cutting and grazing. Grass Forage Sci. 46: 343–350. doi:10.1111/j.1365-2494.1991.tb02393.x. Bittman, S., Hunt, D., and Kowalenko, G. 2004. Cover crops and relay crops. Pages 89–93 in S. Bittman and G. Kowalenko, eds. Advanced silage corn management. Pacific Field Corn Association, Agassiz, BC. Bittman, S., Hunt, D., Swift, M. lL., Casler, M., Undersander, D., and Papadopoulos, Y. 2013. How to improve nutrient efficiency of whole dairy farms. Pages 104–107 in S. Bittman and D. Hunt, eds. Cool forages: Advanced management of temperate forages. Pacific Field Corn Association, Agassiz, BC. Bittman, S., Hunt, D.E., Kowalenko, C.G., Chantigny, M., Buckley, K., and Bounaix, F. 2011. Removing Solids Improves Response of Grass to Surface-Banded Dairy Manure Slurry: A Multiyear Study. J. Environ. Qual. 40: 393. doi:10.2134/jeq2010.0177. Bittman, S., Hunt, D.E., and Shaffer, M.J. 2001. NLOS (NLEAP On Stella). A nitrogen cycling model with a graphical interface: Implications for model developers and users. Pages 383–402 in Modeling carbon and nitrogen dynamics for soil management. CRC Press LLC Boca Raton, FL. Bittman, S., Kowalenko, C.G., Hunt, D.E., and Schmidt, O. 1999. Surface-banded and broadcast dairy manure effects on tall fescue yield and nitrogen uptake. Agron. J. 91: 826–833.   98 Bittman, S., Liu, A., Bhandral, R., Hunt, D., Kowalenko, G., Buckley, K., and Chantigny, M. 2009. Sustainable use of dairy slurry on crops: Dual manure stream concept. Pages 37–38 in Farming Systems Design 2009: Methodologies for Integrated Analysis of Farm Production Systems. Monterey, CA, USA. Bittman, S., Liu, A., Hunt, D.E., Forge, T.A., Kowalenko, C.G., Chantigny, M.H., and Buckley, K. 2012. Precision Placement of Separated Dairy Sludge Improves Early Phosphorus Nutrition and Growth in Corn (Zea mays L.). J. Environ. Qual. 41: 582–591. The American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Inc., Madison, WI. doi:10.2134/jeq2011.0284. Bittman, S., Sheppard, S.C., and Hunt, D. 2017a. Potential for mitigating atmospheric ammonia in Canada. Soil Use Manag. 33: 263–275. doi:10.1111/sum.12336. Bittman, S., Sheppard, S.C., Poon, D., and Hunt, D.E. 2017b. Phosphorus flows in a peri-urban region with intensive food production: A case study. J. Environ. Manage. 187: 286–297. Elsevier Ltd. doi:10.1016/j.jenvman.2016.11.040. Bittman, S., Van Vliet, L.J.P., Kowalenko, C.G., McGinn, S., Hunt, D.E., and Bounaix, F. 2005. Surface-banding liquid manure over aeration slots: A new low-disturbance method for reducing ammonia emissions and improving yield of perennial grasses. Agron. J. 97: 1304–1313. doi:10.2134/agronj2004.0277. Cambareri, G., Drury, C., Lauzon, J., Salas, W., and Wagner-Riddle, C. 2017. Year-round nitrous oxide emissions as affected by timing and method of dairy manure application to corn. Soil Sci. Soc. Am. J. 81: 166–178. The Soil Science Society of America, Inc. Cameron, K.C., and Di, H.J. 2003. The use of a nitrification inhibitor, dicyandiamide (DCD), to decrease nitrate leaching and nitrous oxide emissions in a simulated grazed and irrigated   99 grassland. Soil Use Manag. 18: 395–403. doi:10.1079/sum2002151. Carter, J.E., Jokela, W.E., and Bosworth, S.C. 2010. Grass forage response to broadcast or surface-banded liquid dairy manure and nitrogen fertilizer. Agron. J. 102: 1123–1131. doi:10.2134/agronj2009.0382. Christensen, S. 1991. Organic matter available for denitrification in different soil fractions: effect of freeze/thaw cycles and straw disposal. J. Soil Sci. 19: 703–647. doi:10.1111/j.1365-2389.1991.tb00110.x. Clayton, H., McTaggart, I.P., Parker, J., Swan, L., and Smith, K.A. 1997. Nitrous oxide emissions from fertilised grassland: A 2-year study of the effects of N fertiliser form and environmental conditions. Biol. Fertil. Soils 25: 252–260. doi:10.1007/s003740050311. Dabney, S.M., Delgado, J. a, Collins, F., Meisinger, J.J., Schomberg, H.H., Liebig, M. a, Kaspar, T., and Mitchell, J. 2010. Chapter 9 Using Cover Crops and Cropping Systems for Nitrogen Management. Adv. nitrogen Manag. water Qual.: 230–281. Dabney, S.M., Delgado, J. a, and Reeves, D.W. 2001. Communications in Soil Science and Plant Analysis USING WINTER COVER CROPS TO IMPROVE SOIL AND WATER QUALITY IMPROVE SOIL AND WATER QUALITY. Commun. Soil Sci. Plant Anal. 3624: 37–41. doi:10.1081/CSS-100104110. Davidson, E.A. 2010. Sources of Nitric Oxide and Nitrous Oxide following Wetting of Dry Soil. Soil Sci. Soc. Am. J. 56: 95. doi:10.2136/sssaj1992.03615995005600010015x. Derby, N.E., Steele, D.D., Terpstra, J., Knighton, R.E., and Casey, F.X.M. 2005. Interactions of nitrogen, weather, soil, and irrigation on corn yield. Agron. J. 97: 1342–1351. doi:10.2134/agronj2005.0051. Di, H.J., and Cameron, K.C. 2005. Reducing environmental impacts of agriculture by using a   100 fine particle suspension nitrification inhibitor to decrease nitrate leaching from grazed pastures. Agric. Ecosyst. Environ. 109: 202–212. doi:10.1016/j.agee.2005.03.006. Di, H.J., Cameron, K.C., and Sherlock, R.R. 2007. Comparison of the effectiveness of a nitrification inhibitor, dicyandiamide, in reducing nitrous oxide emissions in four different soils under different climatic and management conditions. Soil Use Manag. 23: 1–9. doi:10.1111/j.1475-2743.2006.00057.x. Dobbie, K.E., and Smith, K.A. 2003. Nitrous oxide emission factors for agricultural soils in Great Britain: the impact of soil water-filled pore space and other controlling variables. Glob. Chang. Biol. 9: 204–218. [Online] Available: http://www.blackwell-synergy.com/doi/abs/10.1046/j.1365-2486.2003.00563.x. Environment and Climate Change Canada 2018. Historical Weather Data: Agassiz CDA. [Online] Available: http://climate.weather.gc.ca/historical_data/search_historic_data_e.html. Erhardt, K., Barth, T., Pasda, G., Zerulla, W., Horchler von Locquenghien, K., Wissemeier, A., Dressel, J., and Rädle, M. 2003. 3,4-Dimethylpyrazole phosphate (DMPP) - a new nitrification inhibitor for agriculture and horticulture. Biol. Fertil. Soils 34: 79–84. doi:10.1007/s003740100380. Eriksen, J., Askegaard, M., and Kristensen, K. 2010. Nitrate leaching from an organic dairy crop rotation: the effects of manure type, nitrogen input and improved crop rotation. Soil Use Manag. 20: 48–54. doi:10.1111/j.1475-2743.2004.tb00336.x. van Es, H.M., Sogbedji, J.M., and Schindelbeck, R.R. 2004. Effect of manure application timing, crop, and soil type on nitrate leaching. J. Environ. Qual. 35: 670–679. doi:10.2134/jeq2005.0143.   101 Farré, I., and Faci, J.-M. 2009. Deficit irrigation in maize for reducing agricultural water use in a Mediterranean environment. Agric. Water Manag. 96: 383–394. Elsevier. doi:10.1016/J.AGWAT.2008.07.002. Flessa, H., and Beese, F. 2000. Laboratory estimates of trace gas emissions following surface application and injection of cattle slurry. J. Environ. Qual. 29: 262–268. American Society of Agronomy, Crop Science Society of America, and Soil~…. Fox, J., and Weisberg, S. 2011. An {R} Companion to Applied Regression. Second. Sage, Thousand Oaks {CA}. [Online] Available: http://socserv.socsci.mcmaster.ca/jfox/Books/Companion. Franzluebbers, A.J. 2007. Integrated crop-livestock systems in the southeastern USA. Agron. J. 99: 361–372. doi:10.2134/agronj2006.0076. Fraser Valley Regional District 2017. Regional Snapshot Series: Agriculture Agricultural Economy in the Fraser Valley Regional District. : 1–20. [Online] Available: http://www.fvrd.ca/assets/Government/Documents/AgricultureSnapshot.pdf. GALLOWAY, J.N., ABER, J.D., ERISMAN, J.W., SEITZINGER, S.P., HOWARTH, R.W., COWLING, E.B., and COSBY, B.J. 2003. The Nitrogen Cascade. Bioscience 53: 341. doi:10.1641/0006-3568(2003)053[0341:TNC]2.0.CO;2. Guillard, K., Morris, T.F., and Kopp, K.L. 1999. The pre-sidedress soil nitrate test and nitrate leaching from corn. J. Environ. Qual. 28: 1845–1852. doi:10.2134/jeq1999.00472425002800060022x. Hermawan, B., and Bomke, A.A. 1997. Effects of winter cover crops and successive spring tillage on soil aggregation. Soil Tillage Res. 44: 109–120. doi:10.1016/S0167-1987(97)00043-3.   102 Hooijboer, A., Oenema, J., Burgers, S., Verloop, K., Boumans, L., and Berge, H. ten 2010. Multiscale Effects of Management, Environmental Conditions, and Land Use on Nitrate Leaching in Dairy Farms. J. Environ. Qual. 39: 2016. doi:10.2134/jeq2010.0035. Kai, P., Pedersen, P., Jensen, J.E., Hansen, M.N., and Sommer, S.G. 2008. A whole-farm assessment of the efficacy of slurry acidification in reducing ammonia emissions. Eur. J. Agron. 28: 148–154. Elsevier. doi:10.1016/J.EJA.2007.06.004. Karasu, A., Kuşcu, H., Öz, M., and Bayram, G. 2015. The effect of different irrigation water levels on grain yield, yield components and some quality parameters of silage maize (Zea mays indentata Sturt.) in marmara region of Turkey. Not. Bot. Horti Agrobot. Cluj-Napoca 43: 138–145. doi:10.15835/nbha4319602. De Klein, C.A.M., Cameron, K.C., Di, H.J., Rys, G., Monaghan, R.M., and Sherlock, R.R. 2011. Repeated annual use of the nitrification inhibitor dicyandiamide (DCD) does not alter its effectiveness in reducing N2O emissions from cow urine. Anim. Feed Sci. Technol. 166–167: 480–491. doi:10.1016/j.anifeedsci.2011.04.076. Kleinman, P.J.A., Sharpley, A.N., Moyer, B.G., and Elwinger, G.F. 2002. Effect of Mineral and Manure Phosphorus Sources on Runoff Phosphorus. J. Environ. Qual. 31: 2026. doi:10.2134/jeq2002.2026. Kuo, S., Huang, B., and Bembenek, R. 2005. Effect of Winter Cover Crops on Soil Nitrogen Availability, Corn Yield, and Nitrate Leaching. Sci. World J. 1: 22–29. doi:10.1100/tsw.2001.267. Kuznetsova, A., Brockhoff, P.B., and Christensen, R.H.B. 2017. {lmerTest} Package: Tests in Linear Mixed Effects Models. J. Stat. Softw. 82: 1–26. doi:10.18637/jss.v082.i13. Langdale, G.W., Blevins, R.L., Karlen, D.L., McCool, D.K., Nearing, M.A., Skidmore, E.L.,   103 Thomas, A.W., Tyler, D.D., and Williams, J.R. 1991. Cover crop effects on soil erosion by wind and water. Cover Crop. clean water: 15–22. Citeseer. Ledgard, S.F., Luo, J., Sprosen, M.S., Wyatt, J.B., Balvert, S.F., and Lindsey, S.B. 2014. Effects of the nitrification inhibitor dicyandiamide (DCD) on pasture production, nitrous oxide emissions and nitrate leaching in Waikato, New Zealand. New Zeal. J. Agric. Res. 57: 294–315. Taylor & Francis. doi:10.1080/00288233.2014.928642. Lemke, R.L., Izaurralde, R.C., and Nyborg, M. 1998a. Seasonal Distribution of Nitrous Oxide Emissions from Soils in the Parkland Region. Soil Sci. Soc. Am. J. 62: 1320. doi:10.2136/sssaj1998.03615995006200050025x. Lemke, R.L., Izaurralde, R.C., Nyborg, M., and Solberg, E.D. 1998b. Tillage and N source influence soil-emitted nitrous oxide in the Alberta Parkland region. Can. J. Soil Sci. 79: 15–24. doi:10.4141/s98-013. Lenth, R. 2018. emmeans: Estimated Marginal Means, aka Least-Squares Means. [Online] Available: https://cran.r-project.org/package=emmeans. Linn, D.M., and Doran, J.W. 1984. Effect of Water-Filled Pore Space on Carbon Dioxide and Nitrous Oxide Production in Tilled and Nontilled Soils1. Soil Sci. Soc. Am. J. 48: 1267. doi:10.2136/sssaj1984.03615995004800060013x. Liu, A., Ma, B.L., and Bomke, A.A. 2005. Effects of Cover Crops on Soil Aggregate Stability, Total Organic Carbon, and Polysaccharides. Soil Sci. Soc. Am. J. 69: 2041. doi:10.2136/sssaj2005.0032. Ludwig, B., Wolf, I., and Teepe, R. 2004. Contribution of nitrification and denitrification to the emission of N 2O in a freeze-thaw event in an agricultural soil. J. Plant Nutr. Soil Sci. 167: 678–684. doi:10.1002/jpln.200421462.   104 Luo, D., Ganesh, S., and Koolaard, J. 2018. predictmeans: Calculate Predicted Means for Linear Models. [Online] Available: https://cran.r-project.org/package=predictmeans. Luo, J., Ledgard, S., Wise, B., Welten, B., Lindsey, S., Judge, A., and Sprosen, M. 2015. Effect of dicyandiamide (DCD) delivery method, application rate, and season on pasture urine patch nitrous oxide emissions. Biol. Fertil. Soils 51: 453–464. doi:10.1007/s00374-015-0993-4. Luttmerding, H.A. 1981. Soils of the Langley-Vancouver Map Area. Volume 6: Technical Data: Soil Profile Descriptions and Analytical Data. Province of British Columbia, Ministry of Environment, Assessment and~…. Luttmerding, H.A., and Sprout, P.N. 1967. Soil survey of Agassiz area. British Columbia Department of Agriculture. Magdoff, F. 1991. Understanding the Magdoff Pre-Sidedress Nitrate Test for Corn. Jpa 4: 297. doi:10.2134/jpa1991.0297. Van Man, N., and Wiktorsson, H. 2003. Forage yield , nutritive value , feed intake and digestibility of three grass species as affected by harvest frequency. Trop. Grasslands 37: 101–110. de Mendiburu, F. 2017. agricolae: Statistical Procedures for Agricultural Research. [Online] Available: https://cran.r-project.org/package=agricolae. Misselbrook, T., Prado, A., and Chadwick, D. 2013. Opportunities for reducing environmental emissions from forage- based dairy farms. Agric. food Sci.: 93–107. Mitchell, R.J., Babcock, R.S., Gelinas, S., Nanus, L., and Stasney, D.E. 2003. Nitrate distributions and source identification in the Abbotsford-Sumas Aquifer, northwestern Washington State. J. Environ. Qual. 32: 789–800. doi:10.2134/jeq2003.0789.   105 Mørkved, P.T., Dörsch, P., Henriksen, T.M., and Bakken, L.R. 2006. N2O emissions and product ratios of nitrification and denitrification as affected by freezing and thawing. Soil Biol. Biochem. 38: 3411–3420. doi:10.1016/j.soilbio.2006.05.015. Odhiambo, J.J.O., and Bomke, A.A. 2001. Grass and Legume Cover Crop Effects on Dry Matter and Nitrogen Accumulation. 1: 299–307. [Online] Available: file:///C:/Users/sethapp/Downloads/aj-93-2-299.pdf. Pandey, R.K., Maranville, J.W., and Chetima, M.M. 2000. Deficit irrigation and nitrogen effects on maize in a Sahelian environment II. 2-Shoot growth, nitrogen uptake and water extraction. Agric. Water Manag. 46: 15–27. doi:10.1016/S0378-3774(00)00074-3. Paul, J.W., and Beauchamp, E.G. 1989. soIL compare denitrification rates. 61: 49–61. Paul, J.W., and Zebarth, B.J. 1996. Denitrification and nitrate leaching during the fall and winter following dairy cattle slurry application(MISSING). Can. J. Soil Sci.: 2240–2321. Perego, A., Basile, A., Bonfante, A., De Mascellis, R., Terribile, F., Brenna, S., and Acutis, M. 2012. Nitrate leaching under maize cropping systems in Po Valley (Italy). Agric. Ecosyst. Environ. 147: 57–65. Elsevier B.V. doi:10.1016/j.agee.2011.06.014. Petersen, S.O., and Sommer, S.G. 2011. Ammonia and nitrous oxide interactions: Roles of manure organic matter management. Anim. Feed Sci. Technol. 166–167: 503–513. doi:10.1016/j.anifeedsci.2011.04.077. Pontes, L.S., Carrère, P., Andueza, D., Louault, F., and Soussana, J.F. 2007. Seasonal productivity and nutritive value of temperate grasses found in semi-natural pastures in Europe: Responses to cutting frequency and N supply. Grass Forage Sci. 62: 485–496. doi:10.1111/j.1365-2494.2007.00604.x. Powell, J.M., Gourley, C.J.P., Rotz, C.A., and Weaver, D.M. 2010. Nitrogen use efficiency: A   106 potential performance indicator and policy tool for dairy farms. Environ. Sci. Policy 13: 217–228. Elsevier Ltd. doi:10.1016/j.envsci.2010.03.007. R Core Team 2017. R: A Language and Environment for Statistical Computing. Vienna, Austria. [Online] Available: https://www.r-project.org/. Rhoads, F.M., Bennett, J.M., and others 1990. Corn. Agronomy: 569–596. Robertson, G.P., and Groffman, P.M. 2015. Ch14: Nitrogen Transformations. Soil Microbiology, Ecology and Biochemistry, 4th edition. Elsevier Inc. doi:10.1016/B978-0-12-415955-6.00014-1. Rochette, P., Bertrand, N., Carter, M., and Gregorich, E.G. 2008. Soil-surface gas emissions. Soil Sampl. methods Anal. CRC Press. Boca Raton, FL: 851–861. Rotz, C.A., Corson, M.S., Chianese, D.S., Hafner, S.D., Jarvis, R., and Coiner, C.U. 2011. The Integrated Farm System Model (IFSM) Reference Manual. Rotz, C.A., Oenema, J., and van Keulen, H. 2013. Whole Farm Management To Reduce Nutrient Losses From Dairy Farms: a Simulation Study. Appl. Eng. Agric. 22: 773–784. doi:10.13031/2013.21992. Rotz, C.A., Satter, L.D., Mertens, D.R., and Muck, R.E. 2010. Feeding Strategy, Nitrogen Cycling, and Profitability of Dairy Farms. J. Dairy Sci. 82: 2841–2855. doi:10.3168/jds.s0022-0302(99)75542-6. Rubæk, G.H., Henriksen, K., Petersen, J., Rasmussen, B., and Sommer, S.G. 1996. Effects of application technique and anaerobic digestion on gaseous nitrogen loss from animal slurry applied to ryegrass ( Lolium perenne). J. Agric. Sci. 126: 481. Cambridge University Press. doi:10.1017/S0021859600075572. De Santis, G., Iannucci, A., Dantone, D., and Chiaravalle, E. 2004. Changes during growth in the   107 nutritive value of components of berseem clover (Trifolium alexandrinum L.) under different cutting treatments in a Mediterranean region. Grass Forage Sci. 59: 378–388. doi:10.1111/j.1365-2494.2004.00439.x. Schröder, J.J., Vermeulen, G.D., van der Schoot, J.R., van Dijk, W., Huijsmans, J.F.M., Meuffels, G.J.H.M., and van der Schans, D.A. 2015. Maize yields benefit from injected manure positioned in bands. Eur. J. Agron. 64: 29–36. Elsevier. doi:10.1016/J.EJA.2014.12.011. Shaffer, M.J., Halvorson, A.D., Keeney, D.R., Pierce, F.J., Follett, R.F., and Cruse, R.M. 1991. Nitrate Leaching and Economic Analysis Package (NLEAP): Model Description and Application. Soil Sci. Soc. Am. journaloil. doi:10.2136/1991.managingnitrogen.c13. Sharpley, A.N. 1995. Soil phosphorus dynamics: agronomic and environmental impacts. Ecol. Eng. 5: 261–279. doi:10.1016/0925-8574(95)00027-5. Sharpley, A.N., Kleinman, P.J.A., Allen, A.L., Bergström, L., and Jordan, P. 2009. Evaluating the Success of Phosphorus Management from Field to Watershed. J. Environ. Qual. 38: 1981. doi:10.2134/jeq2008.0056. Shepherd, M., Barzetti, S., and Hastie, D. 1991. The Production of Atmospheric NOx and N2O From A Fertilized Agricultural Soil. Atmos. Environ.: 1961–1969. Shepherd, M.A., and Newell-Price, J.P. 2016. The effects of pig manure type and application timing and frequency on nitrate leaching from a seven-course arable rotation on a retentive alluvial soil. Soil Use Manag. 32: 117–126. doi:10.1111/sum.12223. Skiba, U., Smith, K.A., and fowler, D. 1993. Nitrification and denitrification as sources of nitric oxide and nitrous oxide in a sandy loam soil. Soil Biol. Biochem. 25: 1527–1536. doi:https://doi.org/10.1016/0038-0717(93)90007-X.   108 Sogbedji, J.M., Geohring, L.D., Magdoff, F.R., van Es, H.M., Yang, C.L., Geohring, L.D., and Magdoff, F.R. 2000. Nitrate leaching and nitrogen budget as affected by maize nitrogen rate and soil type. J. Environ. Qual. 29: 1813–1820. doi:10.2134/jeq2000.00472425002900060011x. Soil Classification Working Group 1998. The Canadian System of Soil Classification. Third Edit. Agriculture and Agri-Food Canada (Publication 1646), Ottawa, Ontario. Sommer, S.., and Hutchings, N.. 2001. Ammonia emission from field applied manure and its reduction—invited paper. Eur. J. Agron. 15: 1–15. Elsevier. doi:10.1016/S1161-0301(01)00112-5. Stewart, J.W.B., and Tiessen, H. 1987. Dynamics of soil organic phosphorus. Biogeochemistry 4: 41–60. doi:10.1007/BF02187361. Subbarao, G., Ito, O., Sahrawat, K., Berry, W., Nakahara, K., Ishikawa, T., Watanabe, T., Suenaga, K., Rondon, M., and Rao, I. 2006. Scope and strategies for regulation of nitrification in agricultural systems - Challenges and opportunities. CRC. Crit. Rev. Plant Sci. 25: 303–335. doi:10.1080/07352680600794232. Tenuta, M., and Sparling, B. 2011. A laboratory study of soil conditions affecting emissions of nitrous oxide from packed cores subjected to freezing and thawing. Can. J. Soil Sci. 91: 223–233. doi:10.4141/cjss09051. Turner, L.R., Donaghy, D.J., Lane, P.A., and Rawnsley, R.P. 2006. Effect of defoliation management, based on leaf stage, on perennial ryegrass (Lolium perenne L.), prairie grass (Bromus willdenowii Kunth.) and cocksfoot (Dactylis glomerata L.) under dryland conditions. 2. Nutritive value. Grass Forage Sci. 61: 175–181. doi:10.1111/j.1365-2494.2006.00524.x.   109 Vallejo, A., García-Torres, L., Díez, J.A., Arce, A., and López-Fernández, S. 2005. Comparison of N losses (NO3-, N2O, NO) from surface applied, injected or amended (DCD) pig slurry of an irrigated soil in a Mediterranean climate. Plant Soil 272: 313–325. doi:10.1007/s11104-004-5754-3. VandeHaar, M.J., and St-Pierre, N. 2010. Major Advances in Nutrition: Relevance to the Sustainability of the Dairy Industry. J. Dairy Sci. 89: 1280–1291. Elsevier. doi:10.3168/jds.s0022-0302(06)72196-8. Veldkamp, E., Keller, M., and Nuñez, M. 1998. O and NO emissions from soils in the humid tropics of Costa Rica. Global Biogeochem. Cycles 12: 71. doi:10.1029/97GB02730. Vinther, F.P. 2006. Effects of cutting frequency on plant production, N-uptake and N2 fixation in harvested and below-harvest plant biomass of grass clover. Grass Forage Sci. 61: 154–163. [Online] Available: http://orgprints.org/10066/. van Vliet, L.J.P., and Zebarth, B.. 2004. Relay crop reduces over-winter runoff from a silage corn field. Page 95 in S. Bittman and G. Kowalenko, eds. Advanced silage corn management. Pacific Field Corn Association, Agassiz, BC. Webb, J., Pain, B., Bittman, S., and Morgan, J. 2010. The impacts of manure application methods on emissions of ammonia, nitrous oxide and on crop response-A review. Agric. Ecosyst. Environ. 137: 39–46. Elsevier B.V. doi:10.1016/j.agee.2010.01.001. Zaman, M., and Blennerhassett, J.D. 2010. Effects of the different rates of urease and nitrification inhibitors on gaseous emissions of ammonia and nitrous oxide, nitrate leaching and pasture production from urine patches in an intensive grazed pasture system. Agric. Ecosyst. Environ. 136: 236–246. Elsevier B.V. doi:10.1016/j.agee.2009.07.010. Zebarth, B.., Hii, B., Liebscher, H., Chipperfield, K., Paul, J.., Grove, G., and Szeto, S.. 1998.   110 Agricultural land use practices and nitrate contamination in the Abbotsford Aquifer, British Columbia, Canada. Agric. Ecosyst. Environ. 69: 99–112. doi:10.1016/S0167-8809(98)00100-5. Zhou, X., Madramootoo, C.A., MacKenzie, A.F., Kaluli, J.W., and Smith, D.L. 2000. Corn yield and fertilizer N recovery in water-table-controlled corn-rye-grass systems. Eur. J. Agron. 12: 83–92. doi:10.1016/S1161-0301(99)00048-9. Zhu, K., Bruun, S., and Jensen, L.S. 2016. Nitrogen transformations in and N2O emissions from soil amended with manure solids and nitrification inhibitor. Eur. J. Soil Sci. 67: 792–803. doi:10.1111/ejss.12385.   111 Appendices Appendix A  Integration of beneficial management practices (BMPs). BMPs (in bold) were added to each scenario incrementally.  Management Scenarios Farmlet Nutrient Management Crop Management Advanced Techniques Corn plots Grass plots Corn plots Grass plots Corn plots Grass plots Conventional F1 Whole Manure Broadcasting Starter N & P fertilizers Whole Manure Broadcasting Long-season corn hybrid  (CHU 2750) 5 harvest per year - - Innovative Nutrient F2 Manure sludge injection Separated liquid manure surface banding Long-season corn hybrid  (CHU 2750) 5 harvest per year - - Innovative Cropping F3 Manure sludge injection Separated liquid manure surface banding Short-season corn hybrid  (CHU 2150) Italian ryegrass as a relay crop 3 harvest per year - - Advanced Techniques F4 Manure sludge injection Separated liquid manure surface banding Short-season corn hybrid  (CHU 2150) Italian ryegrass as a relay crop 3 harvest per year DCD Irrigation (2018) DCD Irrigation     112 Appendix B  Soil sampling scheme for pre-sidedressing nitrate test (PSNT). Twenty cores of 0-15 cm deep soil samples were taken from each corn plot (6.1 m x 18.3 m) to form one composite sample.    113  Appendix C  Illustration of taking a soil water sample from ceramic-cup suction lysimeter.   114 Appendix D  Basal crop coefficient (Kcb), plant height (h), and maximum rooting depth (Zr) of each crop type at different growing stage throughout two years.  Crops Growing	stage2017 2018Start	DateEnd	Date DurationStart	DateEnd	Date DurationStart	KcbEnd	KcbStart	hEnd	hStart	ZrEnd	ZrCorn	 Fallow Jan-1 May-13 133 Jan-1 May-8 128 0.0 0.0 0.0 0.0 0.0 0.0(Late) Initial May-14 Jun-3 20 May-9 May-29 20 0.2 0.2 0.1 1.0 0.1 0.2Development Jun-4 Jul-16 42 May-30 Jul-11 42 0.2 1.2 1.0 2.8 0.2 1.0Mature Jul-17 Sep-2 47 Jul-12 Aug-28 47 1.2 1.2 2.8 2.8 1.0 1.0End Sep-3 Sep-26 23 Aug-29 Sep-24 25 1.2 0.5 2.8 2.8 1.0 0.6Fallow Sep-27 Dec-30 94 Sep-25 Dec-30 96 0.0 0.0 0.0 0.0 0.0 0.0Corn	 Fallow Jan-1 May-13 133 Jan-1 May-8 128 0.0 0.0 0.0 0.0 0.0 0.0(Early) Initial May-14 Jun-3 20 May-9 May-29 20 0.2 0.2 0.1 1.0 0.1 0.2Development Jun-4 Jul-11 37 May-30 Jul-4 35 0.2 1.2 1.0 2.8 0.2 1.0Mature Jul-12 Aug-21 40 Jul-5 Aug-14 40 1.2 1.2 2.8 2.8 1.0 1.0End Aug-22 Sep-10 19 Aug-15 Sep-11 25 1.2 0.5 2.8 2.8 1.0 0.6Fallow Sep-11 Dec-30 110 Sep-12 Dec-30 109 0.0 0.0 0.0 0.0 0.0 0.0Tall	fescue	 End Jan-1 Mar-31 90 Jan-1 Mar-31 90 0.9 0.9 0.3 0.3 0.6 0.6(5-cuts) Development Apr-1 Apr-19 19 Apr-1 Apr-18 18 0.9 1.0 0.3 1.3 0.6 0.6Mature Apr-20 May-8 18 Apr-19 May-6 17 1.0 1.0 1.3 1.3 0.6 0.6Development May-9 May-23 14 May-7 May-21 14 0.4 1.0 0.1 0.8 0.6 0.6Mature May-24 Jun-20 27 May-22 Jun-17 26 1.0 1.0 0.8 0.8 0.6 0.6Development Jun-21 Jul-5 14 Jun-18 Jul-2 14 0.4 1.0 0.1 0.8 0.6 0.6Mature Jul-6 Jul-30 24 Jul-3 Jul-22 19 1.0 1.0 0.8 0.8 0.6 0.6Development Jul-31 Aug-14 14 Jul-23 Aug-6 14 0.4 1.0 0.1 0.8 0.6 0.6Mature Aug-15 Sep-12 28 Aug-7 Sep-2 26 1.0 1.0 0.8 0.8 0.6 0.6Development Sep-13 Sep-27 14 Sep-3 Sep-17 14 0.4 1.0 0.1 0.3 0.6 0.6Mature Sep-28 Oct-10 12 Sep-18 Oct-8 20 1.0 1.0 0.3 0.3 0.6 0.6Development Oct-11 Nov-1 21 Oct-9 Oct-30 21 0.4 0.9 0.1 0.3 0.6 0.6End Nov-2 Dec-30 58 Oct-31 Dec-30 60 0.9 0.9 0.3 0.3 0.6 0.6Tall	fescue	 End Jan-1 Mar-31 90 Jan-1 Mar-31 90 0.9 0.9 0.3 0.3 0.6 0.6(3-cuts) Development Apr-1 Apr-19 19 Apr-1 Apr-18 18 0.9 1.0 0.3 1.3 0.6 0.6Mature Apr-20 May-27 37 Apr-19 May-22 33 1.0 1.0 1.3 1.3 0.6 0.6Development May-28 Jun-10 13 May-23 Jun-6 14 0.4 1.0 0.1 0.8 0.6 0.6Mature Jun-11 Jul-30 49 Jun-7 Jul-22 45 1.0 1.0 0.8 0.8 0.6 0.6Development Jul-31 Aug-14 14 Jul-23 Aug-6 14 0.4 1.0 0.1 0.8 0.6 0.6Mature Aug-15 Oct-10 56 Aug-7 Oct-8 62 1.0 1.0 0.8 0.8 0.6 0.6Development Oct-11 Nov-1 21 Oct-9 Oct-30 21 0.4 0.9 0.1 0.3 0.6 0.6End Nov-2 Dec-30 58 Oct-31 Dec-30 60 0.9 0.9 0.3 0.3 0.6 0.6Italian	Ryegrass End Jan-1 Mar-31 90 Jan-1 Mar-31 90 0.9 0.9 0.3 0.3 0.3 0.3Development Apr-1 Apr-19 19 Apr-1 Apr-18 18 0.9 1.0 0.3 0.6 0.3 0.6Mature Apr-20 Apr-29 9 Apr-19 Apr-24 5 1.0 1.0 0.6 0.6 0.6 0.6Fallow Apr-30 Jun-20 51 Apr-25 Jun-18 54 0.0 0.0 0.0 0.0 0.0 0.0Initial Jun-21 Sep-10 81 Jun-19 Sep-11 84 0.0 0.4 0.0 0.1 0.0 0.1Development Sep-11 Oct-2 21 Sep-12 Oct-3 21 0.4 0.9 0.1 0.3 0.1 0.3End Oct-3 Dec-30 88 Oct-4 Dec-30 87 0.9 0.9 0.3 0.3 0.3 0.3Basal	crop	coeffecient	(Kcb)Plant	height	(h)Max.	rooting	depth	(Zr)  115 Appendix E  List of parameters for calculation of evapotranspiration coefficient (ETc). Parameters Source Reference evapotranspiration (ETO) Penman-Monteith equation modified by FAO (Equation 3.2, Allen et al. 1998) Vapor Pressure Deficit (VPD) Equation from FAO Paper 56 (Example 6, Allen et al. 1998) and relative humidity data from ECCC weather station (Agassiz CDA) (ECCC, 2018). Slope of saturation vapor pressure curve (∆) Equation from FAO Paper 56 (Equation 13, Allen et al. 1998) and temperature data from ECCC weather station (Agassiz CDA) (ECCC, 2018). Psychometric constant (γ) Equation from FAO Paper 56 (Equation 8, Allen et al. 1998) and atmospheric pressure data from ECCC weather station (Agassiz CDA) (ECCC, 2018). Basal crop coefficient (Kcb) Equation from FAO Paper 56 (Equation 70, Allen et al. 1998)  Table value of Kcb (Kcb (Tab)) FAO Paper 56 (Table 17, Allen et al. 1998) and Personal communication with Derek Hunt Wind speed at 2 m above ground surface (u2) ECCC weather station (Agassiz CDA) (ECCC, 2018). Relative humidity (RH) ECCC weather station (Agassiz CDA) (ECCC, 2018). Plant height (h) Personal communication with Derek Hunt Evaporation coefficient (Ke) !" = !$(!&	()* − !&,) ≤ /"0!&	()*; FAO Paper 56 (Equation 71, Allen et al. 1998) Upper limit Kc  (Kc max) !&	()* = 123 1.2 + 0.04 :; − 2 − 0.04 <=(>? − 45 AB C.B , !&, + 0.05 ,	FAO Paper 56 (Equation 72, Allen et al. 1998) Soil evaporation reduction coefficient (Kr) !$ = EFGHIJ,KLMEFGHNFG 	/OP	Q",>HR > <TU;	FAO Paper 56 (Equation 74 example 35, Allen et al. 1998) Maximum cumulative depth of evaporation (TEW) WTU = 1000(XYZ − 0.5X0[)\"; soil water content at field capacity and wilting point were determined using measured soil texture and organic matter content. Ze was set at 10 cm as suggested by FAO paper 56 (Equation 73 and table 19, Allen et al. 1998) Cumulative depth of evaporation (REW) Simulated at daily step as summation of daily depletion and depletion from previous day following complete wetting from the exposed and wetted fraction of soil (De, i and De, i-1) Fraction of soil exposed and wetted (few) /"0 = min	(1 − /&, /0); FAO paper 56 (Equation 75 and table 20, Allen et al. 1998) Net radiation on top of crop (Rn) <? = <?` − <?a; FAO paper 56 (Equation 40, Allen et al. 1998)  Net shortwave radiation (Rns) Obtained from the NASA Langley Research Center (LaRC) POWER Project  Net longwave radiation (Rnl) Obtained from the NASA Langley Research Center (LaRC) POWER Project   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0378545/manifest

Comment

Related Items