THE USE OF FALL - PLANTED BRASSICA CEAE COVER CROP MONO - AND BICULTURES FOR NUTRIENT CYCLING AND WEED SUPPRESSION By Victoria Joy Ackroyd A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Crop and Soil Sciences Doctor of Philosophy 2015 ABSTRACT THE USE OF FALL - PLANTED BRASSICA CEAE COVER CROP MONO - AND BICULTURES FOR NUTRIENT CYCLING AND WEED SUPP RESSION By Victoria Joy Ackroyd Cover crops have the potential to increase the sustainability of agronomic cropping systems. Farmers are increasingly interested in using oilseed radish ( Raphanus sativus L. var. oleiformis Pers. ), both alone and in mixtures, to suppress weeds, reduce fertilizer inputs, and improve crop yields. However, there is limited information to guide cover crop species selection. To evaluate differences between and within species, we evaluated biomass accumulation of six oilse ed radishes, two brown mustards ( Brassica juncea [L.] Czern.), two white mustards ( Sinapis alba L.), one rapeseed ( Brassica napus L. ), and one hybrid turnip ( Brassica rapa L. x B. napus L. ) in field trials . Cover crop biomass accumulation within and between species was similar. The accessions provided rapid ground cover and accumulated biomass at similar rates. In Minnesota, dry aboveground biomass ranged between 3410 - 5542 kg ha - 1 , while in Michigan biomass ranged between 2545 - 3572 kg ha - 1 . There were no differences in N uptake for any of the accessions in either trial. Brassicaceae cover crops accumulated 100 - 131 kg N ha - 1 and 81 - 109 kg N ha - 1 in aboveground tissues in Minnesota and Michigan, respectively. Experiments were then conducted to investigat e the growth and weed suppression of oilseed radish, annual ryegrass ( Lolium multiflorum Lam.), cereal rye ( Secale cereale L.), oats ( Avena sativa L.), crimson clover ( Trifolium incarnatum L.), hairy vetch ( Vicia villosa Roth.), and winter pea [ Pisum sativum var. arvense (L.) Poir.] both in monocultures and in biculture mixtures of oilseed radish plus each species . Cover crop and weed biomass varied across years. Oilseed radish comprised the majority of biculture fall biomass, and was more competitive in biculture with legumes than grasses. Grass monoculture and grass biculture treatments were more effe ctive at weed suppression in fall 2012, fall 2013, and spring 2014 than legume monoculture treatments. Crimson clover failed to establish in two out of t he three years, and winter pea failed to survive the winter in two out of three years. This study also evaluated the impact of the cover crop monocultures and bicultures on a following corn crop in the absence of applied fertilizer. Overall, the cover crop s did not reduce corn grain yield, with the exception of annual ryegrass and cereal rye treatments each in one of three years. Annual ryegrass and cereal rye reduced corn yield by 51% and 24%, respectively, compared with the weedy control. An additional ex periment was conducted in Lansing and Hickory Corners, MI to determine the impact of fall - planted oilseed radish, annual ryegrass, and radish + ryegrass cover crops on nitrous oxide (N 2 O) emissions . There were no differences between the cover crop treatmen ts and the bare ground control for fall and spring - summer cumulative N 2 O N emissions. It appears nitrous oxide emissions did not represent a major pathway for N loss in this study. This work adds to the cover crop body of knowledge and provides informatio n which will be of use when making recommendations to farmers. iv This work is ten years a college student, but the or Aunt Ma Watts, who maybe does understand. v ACKNOWLEDGEMENTS I would like first and foremost to thank my graduate program committee members: Drs. Christy Sprague, Dale Mutch, Dan Brainard, and Kurt Steinke. My thanks also to Dr s . Dean Baas and Neville Millar , de facto committee member s. A student could no t hope for a more helpful or knowledgeable group of mentors. Over the course of my graduate career I have enjoyed the camaraderie and guidance and Dr. Lin da Hanson. This research would no t have been possible without the assistance of Todd Martin, Dr. Erin Hill, Josh Dykstra, Kevin Kahmark, Jon Dahl, Gary Powell, Mia Mutch, Briar Adams, and P hil Kantola. I am grateful to the rest of the previous and current graduate students in the weeds and cover crops research groups for their friendships and insights. Many thanks to the denizens of the Crop Barn: Brian Graff, Tom Galecka, Randy Lorenz, John Boyce, Bill Widdicombe, Joe Paling, and Lori Williams. Many thanks also to the denizens of the Plant, Soil, and Microbial Sciences office and other departmental staff: Dr. Jim Kells, Theresa Iadipolo, Sandie Litchfield, Darlene Johnson, Cal Bricker, Linda Colon, and Jessica Dunckel. I greatly appreciated the opportunity to collaborate with USDA - NRCS personnel in the mid - Michigan area including Dr. John Durling, Dr. John Leif, and Sergio Perez. I would like to thank Drs. Steven Mirsky and Harry Schomberg of USDA - ARS Beltsville Agricultural Research Center for their patience and assistance with stat istics, respectively, as I completed the writing of this dissertation. I owe many thanks to my funding sources: the Corn Marketing Program of Michigan; the Great L akes Regional Water Program; Michigan Corn Growers Association; MSU Project GREEEN; MSU Dept. of Plant, Soil and Microbial Sciences; MSU Council of Graduate vi Students and the MSU Graduate School; USDA - NIFA (Grant # ORG 2011 - 51106 - 31046), and NCR - SARE (Gradu ate Student Grant # GNC 13 - 164). And then there i s Dr. Ronald Nussbaum, persistently long - term nuisance and distraction. Trouble from the start. Ronald - I graduated last between us for the last time, so mine is the last word: love you most. vii TAB LE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ .......................... x LIST OF FIGURES ................................ ................................ ................................ ..................... xii KEY TO ABBREVIATIONS ................................ ................................ ................................ ...... x i v CHAPTER 1 LITERATURE REVIEW ................................ ................................ ................................ ................ 1 Cropping System Overview ................................ ................................ ................................ . 1 Benefits of Cover Crop Use ................................ ................................ ................................ . 2 Improved soil physical properties ................................ ................................ ....................... 2 Nutrient retention ................................ ................................ ................................ ................. 3 Weed suppression ................................ ................................ ................................ ................ 5 Increased cash crop yield ................................ ................................ ................................ ..... 6 Difficulties Associated with Cover Crop Use ................................ ................................ ...... 7 Interference with cash crop ................................ ................................ ................................ .. 8 Decreased cash crop yield ................................ ................................ ................................ .... 8 Types of Cover Crops ................................ ................................ ................................ .......... 9 Grasses ................................ ................................ ................................ ................................ . 9 Legumes ................................ ................................ ................................ ............................. 11 Brassicaceae species ................................ ................................ ................................ .......... 13 Cover crop mixtures ................................ ................................ ................................ ........... 15 Dissertation Objectives ................................ ................................ ................................ ...... 17 APPENDIX ................................ ................................ ................................ ................................ .... 19 LITERATURE CITED ................................ ................................ ................................ .................. 28 CHAPTER 2 GROWTH CHARACTERISTICS OF BRASSICACEOUS SPECIES USED AS FALL - PLANTED COVER CROPS IN MINNESOTA AND MICHIGAN ................................ ............ 36 ABSTRACT ................................ ................................ ................................ ....................... 36 INTRODUCTION ................................ ................................ ................................ ............. 37 MATERIALS AND METHODS ................................ ................................ ....................... 41 Site description, experimental design, and field management ................................ ........... 41 Data collection ................................ ................................ ................................ ................... 42 Weather data and growing degree day calculations ................................ ........................... 43 Data analysis ................................ ................................ ................................ ...................... 44 RESULTS AND DISCUSSION ................................ ................................ ........................ 45 Weather ................................ ................................ ................................ .............................. 45 Ground cover and biomass accumulation of four Brassicaceae accessions ...................... 46 Characteristics of all twelve Brassicaceae accessions ................................ ...................... 49 Flowering and seed production ................................ ................................ ......................... 52 Winter hardiness ................................ ................................ ................................ ................ 53 viii CONCLUSIONS ................................ ................................ ................................ ................ 54 APPENDICES ................................ ................................ ................................ ............................... 55 APPENDIX A CHAPTER 2 TABLES AND FIGURES ................................ .................. 56 APPENDIX B NON - OILSEED RADISH BRASSICACEAE ROOT N CONTENT ANALYSIS OF VARIANCE ................................ ................................ ........................... 66 LITERATURE CITED ................................ ................................ ................................ ............... ... 67 CHAPTER 3 PERFORMANCE OF COVER CROP MONOCULTURES VS. BICULTURES: COVER CROP GROWTH AND IMPACT ON WEED BIOMASS AND CORN ( Zea mays L.) YIELD ............ 73 ABSTRACT ................................ ................................ ................................ ...................... 73 INTRODUCTION ................................ ................................ ................................ ............ 74 MATERIALS AND METHODS ................................ ................................ ...................... 81 Site characteristics and field operations ................................ ................................ ............ 81 Cover crop, weed, and corn data collection ................................ ................................ ...... 83 Biomass relative yield calculation ................................ ................................ .................... 84 Statistical analysis ................................ ................................ ................................ .............. 84 RESULTS AND DISCUSSION ................................ ................................ ....................... 85 Cover crop biomass production ................................ ................................ ......................... 85 Cover crop impact on weeds ................................ ................................ ............................. 91 Biomass relative yields (RY) ................................ ................................ ............................ 94 Cover crop impact on corn ................................ ................................ ................................ . 95 CONCLUSIONS ................................ ................................ ................................ ............... 98 APPENDICES ................................ ................................ ................................ ............................. 100 APPENDIX A CHAPTER 3 TABLES AND FIGURES ................................ ................ 101 APPENDIX B CHAPTER 3 SOIL, WEED, AND CORN SUPPLEMENTARY DATA COLLECTION METHODS ................................ ................................ ............................ 110 APPENDIX C CHAPTER 3 SUPPLEMENTARY TABLES ................................ ........ 113 LITERATURE CITED ................................ ................................ ................................ ................ 119 CHAPTER 4 IMPACT OF FALL - PLANTED COVER CROPS ON NITROUS OXIDE EMISSIONS ........ 127 ABSTRACT ................................ ................................ ................................ ..................... 127 INTRODUCTION ................................ ................................ ................................ ........... 12 8 MATERIALS AND METHODS ................................ ................................ ..................... 130 Site characteristics and field operations ................................ ................................ ........... 130 Data collection ................................ ................................ ................................ ................. 132 Data analysis ................................ ................................ ................................ .................... 135 RESULTS AND DISCUSSION ................................ ................................ ...................... 136 Cover crop biomass ................................ ................................ ................................ .......... 136 Cover crop C:N ratios, N and soil N ................................ ................................ ............... 138 Nitrous oxide emissions ................................ ................................ ................................ ... 141 CONCLUSIONS ................................ ................................ ................................ .............. 144 APPENDICES ................................ ................................ ................................ ............................. 146 AP PENDIX A CHAPTER 4 TABLES AND FIGURES ................................ ................ 1 47 ix APPENDIX B CORN GROWTH AND YIELD MATERIALS AND METHODS ...... 156 APPENDIX C CORN GROWTH AND YIELD TABLES ................................ ............. 158 LITERATURE CITED ................................ ................................ ................................ ................ 161 x LIST OF TABLES Table 1.1 . Overview of fall - planted cover crop impact on corn yield in the literature (assuming appropriate management of cover crop such as allowing adequate time between cover crop termination and corn planting). ................................ ................................ ................................ ...... 20 Table 1.2. Overview of cover crop dry aboveground biomass in the literature ... .......................... 22 Table 1.3 Cover crop mixture dry aboveground biomass, as reported in the literature. .. .............. 26 Table 2.1. Brassicaceae accessions evaluated in trials in St. Paul, MN and Bath, MI in 2010 and 2011 ................................ ................................ ................................ ................................ ................ 56 Table 2.2. Brassicaceae variety trial site characteristics at St. Paul, MN and Bath, MI in 2010 and 2011. .. ................................ ................................ ................................ ................................ ............. 5 7 Table 2.3. Final aboveground biomass production and N uptake of twelve Brassicaceae . ................................ ................................ .. 58 Table 2.4. Fin al dry root biomass production and N uptake of six oilseed radish accessions ................................ ................................ ..................... 59 Table 2.5. Flowering initiation date and winter survival of Brassicaceae cover crops in MN and MI 2010 - 2011 and 2011 - 2012... ................................ ................................ ................................ .... 60 Table 2.6 . Root tissue N concentrations of non - oilseed radish Brassicaceae cover crop accessions ................................ ................................ ...................... 66 Table 3.1. Estimated cost of cover crop seed in an experiment in Lansing, MI... ....................... 101 Table 3.2. Field characteristics, field operation and data collection dates for experiments conducted in Lansing, MI ................................ ................................ ................................ ........... 102 Table 3.3. Maximum annual dry aboveground biomass produced by cover crop monocultures and bicultures in Lansing, MI (2012 - 2014). ................................ ................................ ................ 103 Table 3.4. Fall and spring total dry aboveground weed biomass in Lansing, MI (2012 - 2014)... 104 Table 3.5. - 2014)... ................................ ................................ ................................ ................................ ......... 105 Table 3.6 . Data collection dates for experiments conducted in Lansing, MI... ........................... 113 Table 3.7 . Fall oilseed radish root: shoot ratios in Lansing, MI (2012 - 2014) . ......................... 11 4 Table 3.8 . Spring pre - side dress nitrate - N (PSNT), weed counts conducted at the time of the first herbicide application, and post corn harvest residual soil nitrate levels in experiments conducted in Lansing, MI (2012 - 2014). ................................ ................................ ................................ ........ 115 xi Table 3.9 . from experiments conducted in Lansing, MI (2012 - 2014)... ................................ ................................ ................................ ................................ ......... 116 Table 3.10 . Relative N status at corn stages V6 and VT from experiments conducted in Lansing, MI (2011 - 2014)... ................................ ................................ ................................ ......................... 117 Table 3.11 . ex of corn at stage - 2014). ................................ ............... 118 Table 4.1. Lansing, MI (MSU). . ................................ ................................ ................................ ................... 147 Table 4.2. Field operation and data collection dates in Hickory Corners, MI (KBS) and Lansing, MI (MSU). ................................ ................................ ................................ ................................ ... 148 Table 4.3. Corners, MI (KBS) and Lansing, MI (MSU) (2012 - 2014). ................................ ......................... 149 Table 4.4. Fall and spring cover crop C:N ratios a Hickory Corners, MI (KBS) and Lansing, MI (MSU) (2012 - 2014). ................................ ........... 150 Table 4.5. Soil NO 3 values during fall and spring cover crop phases and after corn harvest in Hickory Corners, MI (KBS) and Lansing, MI (MSU) (2012 - 20 14). ................................ ........... 151 Table 4.6. Cumulative, daily, and scaled N 2 O - N emissions in Hickory Corners, MI (KBS) in 2012 - 2014... ................................ ................................ ................................ ................................ . 152 Table 4.7. Cumulative, daily, and scaled N 2 O - N emissions in Lansing, MI (MSU) in 2012 - 2014 ... ................................ ................................ ................................ ................................ ........... 153 Table 4.8 . Field operation and corn data collection dates in Hickory Corners, MI (KBS) and Lansing, MI (MSU). ................................ ................................ ................................ .................... 158 Table 4.9 . Mean ± SE relative N status at stages V6 and VT in Hickory Corners, MI (KBS) and Lansing, MI (MSU).... ................................ ................................ ................................ ................. 159 Table 4.10 . Corn height, ear leaf N, leaf area index (LA I), and grain yield in Hickory Corners, MI (KBS) and Lansing, MI (MSU) (2013 - 2014).... ................................ ................................ .... 160 xii LIST OF FIGURES Figure 2.1. Monthly and 30 - yr average growing degree days (base 5 C) and precipitation for St. Paul, MN (top) and Bath, MI (bottom) in 2010 and 2011. Growing degree days are represented on the left y - axis and by lines, while precipitation is represented on the right y - axis and by bars. ................................ ................................ ................................ ................................ ........................ 61 Figure 2.2. oilseed radish accumulation in Minnesota in 2011 and 2012. Polynomial regression was conducted with growing degree days (GDD) (base 5 °C) and percent ground cover as covariates. Means ± SE ... ................................ ................................ ................................ ................................ ........................ 62 F igure 2.3. accumulation in Michigan in 2011 and 2012. Polynomial regression was conducted with growing degree days (GDD) (base 5 °C) and percent ground cover as covariates. Means ± SE... ............. 63 F igure 2.4. oilseed radish (gray triangle) dry aboveground accumulation in Minnesota in 2011 and 2012. Polynomial regression was conducted with growing degree days (GDD) (base 5 °C) and biomass as covariates. Means ± SE. Response of production to GDD: y = 0.0099x 2 + 1.84x - 1102, r 2 = 0.91, P to GDD: y = - 0.0034x 2 + 11.45x - 2754, r 2 = 0.94, P = <0.0001... ................................ ................ 64 Figure 2.5. accumulation in Michigan in 2011 and 2012. Polynomial regression was conducted with growing degree days ( 2 + 1.61x - 587, r 2 = 0.91, P = GDD: y = - 0.0058x 2 + 11.88x - 2254, r 2 = 0.97, P = <0.0001... ................................ .................... 65 Figure 3.1. Monthly and 30 - yr mean air temperature and precipitation for Lansing, MI 2011 - 2014. Precipitation is represented on the left y - axis and by bars. Temperature is represented on t he right y - axis and by lines... ................................ ................................ ................................ ...... 106 Figure 3.2. Mean (±SE) dry aboveground fall biomass of legume and grass cover crops in monoculture and in biculture with oilseed radish in Lansing, MI in 2011 - 2013 ......................... 107 xiii Figure 3.3. Mean (±SE) dry aboveground spring biomass of legume and grass cover crops in monoculture and in biculture with oilseed radish, weeds, and volunteer wheat in Lansing, MI in 2012 - 2014 ................................ ................................ ................................ ................................ .... 108 Figure 3.4 . Fall relative yields of oilseed radish and complementary species (species B) in each cover crop biculture treatment. Bars represent one SE from the mean. To the left of the diagonal line RY oilseed radish =RY species B , oilseed radish had the competitive adva ntage over species B while to the right of the diagonal line, the reverse was true. In the upper right quadrant, the species interaction was mutually beneficial. In the lower right quadrant, species B suppressed oilseed radish growth. The lower left quadra nt indicates competition or interference between the two species. In the upper left quadrant, oilseed radish suppressed the growth of species B. Figure adapted from Williams and McCarthy (2001) . ................................ ................................ ............ 109 Figure 4.1 . Temperature (°C), daily precipitation (mm), and nitrous oxide emissions (g N 2 O - N ha - 1 day) from 29 Oct. 2012 18 Sep. 2013 (year 1) and 28 Oct. 2013 17 Sep. 2014 (year 2) in Hickory Corners, MI (KBS). Each error bar represents one standard error from the mean ........ 154 Figure 4.2 . Temperature (°C), daily precipitation (mm), and nitrous oxide emissions (g N 2 O - N ha - 1 ) from 11 Nov. 2012 27 Sep. 2013 and 24 Oct. 2013 19 Sep. 2014 in Lansing, MI (MSU). Note the right (2013 - 2014, year 2) y - axis is one order of magnitude larger than the left y - axis. Eac h error bar represents one standard error from the mean ................................ ....................... 155 xiv KEY TO ABBREVIATIONS DAP: days after planting GDD: growing degree days NIRS: near infrared spectroscopy PSNT: pre - side dress nitrogen - N test RY: relative yield VNS: variety not specified WAP: weeks after planting 1 CHAPTER 1 LITERATURE REVIEW Cropping System Overview generates over $90 billion each year (USDA - NASS 2013). In commodity group in terms of cash receipts was field crops. Corn ( Zea mays L.), soybeans [ Glycine max (L.) Merr.], and winter wheat ( Triticum aestivum L.) were respectively ranked the 1 st , 2 nd , and 5 th largest commodities in that g roup (USDA - NASS 2015). Michigan farmers planted 2.6 million acres of corn, 1.9 million acres of soybeans, and 620,000 acres of winter wheat in 2013 (USDA - NASS 2015). As the value of these commodities has increased over the last five years, so has the cost of inputs. Farmers have minimal control over some expenses such as cropland rent. However, the use of other inputs such as fertilizer can be creatively managed to decrease costs and increase profitability. One large and increasingly expensive input for far mers is nitrogen (N) fertilizer. It is also particularly well - suited to creative management. The goal of fertility management strategies is to increase the nitrogen use efficiency (NUE) of the system by synchronizing fertilizer application with peak crop n eed, while minimizing N losses to the environment (Ribaudo et al., 2011) . In addition to careful manipulation of N fertilization, Robertson and Vitousek (2009) advocated the incorporation of cover crops into crop rotations to improve the uptake of N fertilizer into the system and increase the sustainability of cropping systems. Cover crops are generally defined as species which are grown between cash crops, when the ground would otherwise lie fallow. In Michigan, one of the best windows to plant cover crops in a winter wheat - corn - soybean rotation is after wheat harvest. Winter wheat is typically harvested in 2 July, which allows adequate time to establish a cover c rop before the onset of freezing temperatures. Research into the use of cover crops after wheat harvest would be of interest to a wide audience. Benefits of Cover Crops Use Cover crop use has grown steadily among farmers over the last five years (CTIC, 20 15). This is a result of both more farmers using cover crops and current users increasing their cover crop acreage. The top f benefits that cover crop users seek from cover crops were increased overall soil health, increased soil organic matter, reduced s oil erosion, weed suppression, and reduced soil compaction (CTIC, 2015). These results echo those of a survey by Singer et al. (2007) , who found that the top two benefits ascribed to cover crops by farmers were the reduction of soil erosion (96% of respond ents) and increases in soil organic matter (74% of respondents). Improved soil physical properties Cover crops have been documented to decrease soil erosion and improve soil physical properties (Meisinger et al., 1991) . Oats ( Avena sativa L.) and cereal ry e ( Secale cereale L.) each decreased rill and inter - rill erosion in at least one year of a three year study (Kaspar et al., 2001) . In this study, cereal rye decreased erosion by 48 - 62% and oats by 51% during simulated rainfall events on ground with an aver age slope of 4.4%. The authors hypothesized that the decrease in erosion was due to decreased sediment detachment and an increase in ponding and sediment deposition. Hively and Cox (2001) found annual ryegrass ( Lolium multiflorum Lam.) to be an acceptable choice for erosion control after soybean harvest as it reliably provided over 75% ground cover. Villamil et al. (2006) found that after two years, crop rotations including both hairy vetch ( Vicia villosa Roth) and cereal rye had a higher soil organic matte r (SOM) content 3 within the top 30 cm of soil than other rotations. They postulated that this was due to the addition of C from the cereal rye, plus N from the hairy vetch that fueled microbial activity which lead to the increased SOM levels. The same study found that rotations which included cereal rye or hairy vetch had 9 - 17% greater wet aggregate soil stability, depending on the rotation (Villamil et al., 2006) . Rotations which include a cover crop can benefit from decreased soil bulk density, which means improved aeration and water infiltration as well as better cash crop establishment and root growth (Villamil et al., 2006) . Williams and Weil (2004) observed soybean roots growing down into channels created by the roots of canola ( Brassica rapa L.), oilse ed radish ( Raphanus sativus L.) , cereal rye, and cereal rye plus oilseed radish cover crops. Nutrient retention Cover crops can supply N to the following cash crop and help retain nutrients in the system (Meisinger et al., 1991) . Nitrogen is of particular concern due to its mobility and activity in the soil. Aside from incorporation into soil organic matter, N can leach, run off, or be lost to volatilization and other soil gaseous emissions (greenhouse gas emissions). Roughly 50% of the N applied to agricu ltural systems is lost through these pathways (Tonitto et al., 2006) . In the best - case scenario, cover crops take up residual soil N and then that N is released from the biomass in synchrony with cash crop N demands. Wagger (1989) found that while cereal r ye, crimson clover ( Trifolium incarnatum L.), and hairy vetch accumulated more tissue N when terminated later in the spring, early termination resulted in faster N release. The author suggested that this faster N release offset the lower amount of N accumu lated by the cover crops. In the Midwest, N leaching is of most concern from November May. Cover crops can decrease leaching by decreasing soil moisture levels via evapotranspiration and by taking up residual soil nitrate (NO 3 ) (Meisinger et al., 1991) . Cereal rye is particularly effective at 4 decreasing nitrate leaching (Kaspar et al., 2007; Ruffo et al., 2004) . Kaspar et al. (2007) found cereal rye decreased nitrate leaching by at least 50%, compared with a no - cover control, in all four years of a stud y. Tonitto et al. (2006) calculated that legumes reduced nitrate leaching by 40% while non - legume cover crops reduced leaching by an average of 70%. Collins et al. (2007) found mustard ( Brassica hirta - 142 kg N ha - 1 and decreased nit rate leaching. Aside from leaching and runoff, N could be lost from the system in the form of greenhouse gas emissions. Not much literature exists on the topic of cover crops and their impact on greenhouse gas emissions. However, Parkin and Kaspar (2006) found that while corn plots emitted significantly more N 2 O than soybean plots, the inclusion of a cereal rye winter cover crop after both corn and soybean harvest each year did not affect N 2 O emissions from each cash crop. Robertson et al. (2000) stated th at the primary driver of N 2 O fluxes in agricultural systems is the amount of N available in the soil. The more N available in the soil, the higher the flux. This assertion is supported by research done by Gomes et al. (2009) , in which leguminous cover crop s including vetch ( Vigna sativa L.), cowpea ( Vigna unguiculata L. Walp), pigeon pea ( Cajanus cajan L. Millsp.), and lablab ( Dolichos lablab L. ) had larger cumulative N 2 O emissions during the 45 days after cover crop residue management than a black oat ( Ave na strigosa Schreb) cover crop in no - till maize in a subtropical climate. In spite of this, the authors calculated that less than one percent of the N added to the soil by the legumes was subsequently lost as greenhouse gas emissions. Gomes et al. (2009) a lso determined that N 2 O emissions in the 45 days after cover crop residue management correlated with the N added to the soil by the cover crop biomass, while after that period total soil N content drove N 2 O emissions. The N 2 O emissions took longer to peak in the black oat + vetch cover crop mixture treatment than in the 5 C:N ratio can improve N synchrony by delaying N loss from the system. Weed suppression Weeds can significantly reduce crop yield and thus their control is of considerable concern to farmers (Walsh et al., 2013) . Farmers currently control weeds primarily through the use of herbicides and cultivation; both tactics have detrimental aspects. The intensive use of herbicides can lead to the development of herbicide - resistant weeds, while cultivation increases soil erosion. Cover crops are thus of interest as a weed - suppression tool. They can make environmental conditions unfavorable for weed germi nation and growth, increase the activity of weed seed predators and weed seedling pathogens, act as a mulch that smothers weeds, outcompete weeds for nutrients or light, and some are allelopathic (Conklin et al., 2002; Creamer et al., 1996; Pullaro et al., 2006) . Stivers - Young (1998) found that oilseed radish and mustards suppressed weeds in the fall and their residues shaded out weeds in the spring. Cereal rye, crimson clover, and red clover planted during the wheat phase of a rotation were believed to als o decrease weed biomass by shading (Smith et al., 2008) . Weed species richness was likewise decreased, an effect the authors attributed specifically to the cover crops and not to rotational diversity. Several common cover crops are believed to be allelopat hic including cereal rye, hairy vetch, and oilseed radish. For example, laboratory studies indicate that substances produced by Brassicaceae cover crops such as oilseed radish inhibit weed seed germination (Norsworthy and Meehan IV, 2005; Norsworthy and Me ehan IV, 2009; Norsworthy et al., 2006) . Promising laboratory results do not always translate to the field. In one study, cereal rye and hairy vetch had no impact on weed density (Davis, 2010) . Wang et al. (2008) found that sorghum - sudangrass [ Sorghum bic olor (L.) Moench X S. sudanese (P.) Stapf ] and Brassicaceae 6 cover crops reduced weed density and altered the composition of weed species, but did not eliminate the need for other methods of weed control. De Bruin et al. (2005) determined that when weed pre ssure was high, a late - season herbicide application in addition to the cereal rye cover crop was necessary for adequate weed control. Fall - planted cereal rye and crimson clover were more effective at suppressing weed growth than hairy vetch in a North Caro lina study, though herbicides were still necessary for optimum corn yields (Yenish et al., 1996) . Increased cash crop yield Cover crops can increase the yield of a following cash crop (Table 1.1). In a meta - analysis by Miguez and Bollero (2005) , corn grown after legume cover crops yielded 24% more than that grown without a cover crop when no inorganic fertilizer was applied. Another meta - analysis found that as long as a legume cover crop provided at least 110 kg ha - 1 of N, cash crop yields did not differ be tween cover crop plots and those in which an inorganic fertilizer had been applied (Tonitto et al., 2006) . The use of Dutch white clover ( Trifolium repens L.) and medium red clover ( Trifolium pratense L.) has also been shown to increase corn yield, as comp ared with a no - cover crop control (Hively and Cox, 2001) . In a study by Williams and Weil (2004) , soybean yields were greater at one site following an oilseed radish plus cereal rye mixture than after cereal rye and no - cover crop treatments. At another sit e they were greater following cereal rye than after a cereal rye plus oilseed radish mixture, oilseed radish, and no - cover crop treatments. The authors postulated that the cover crops provided the most benefit at the site with lower precipitation levels an d more compacted soil. Smith et al. (2008) found a linear relationship between crop rotation diversity and corn yield. The more cash and cover crop species included in the rotation, the higher the yield. They attributed this effect to the N provided by leg ume species. Soybean and wheat yields also benefited from increased rotational diversity, 7 though not to the same extent as corn (Smith et al., 2008) . Other studies have shown cover crops to generally have no impact on corn yield (Delate et al., 2003; Yenis h et al., 1996) or soybean yield (De Bruin et al., 2005; Delate et al., 2003; Kaspar et al., 2007; Reddy, 2001; Ruffo et al., 2004) (Table 1.1) . A meta - analysis by Tonnito et al. (2006) determined that when inorganic fertilizer was applied, cash crop yield s did not differ between bare fallow plots and those in which a non - legume cover crop had been grown. The impact of the cover crop on cash crop yield is dependent on the interaction among a variety of variables including cover crop management, weed pressur e, soil fertility, and the weather (De Bruin et al., 2005) . Kaspar et al. (2007) found cereal rye to have no impact on corn yield so long as sufficient time was allowed between cover crop termination and cash crop planting. Difficulties Associated with Cover Crop Use In spite of the benefits provided by cover crops, cover crop adoption remains relatively low. Singer et al. (2007) found that in the last five years, only 11% of surveyed Corn Belt farmers had used cover crops. Several factors limit cover cr op use, and they all relate to economics (Snapp et al., 2005). These factors can be loosely grouped into physical costs (e.g., of seed), time costs (e.g., cover crop planting), management concerns (e.g., cover crops may need to be planted at a busy point i n the season like during cash crop harvest), interference with the cash crop (e.g., causing delayed planting), and difficulty choosing the right cover crop. In the CTIC - user s cited were the time and labor associated with planting and management, the cost to plant/manage the cover crops, seed cost, difficulty getting the cover crop to establish, and concerns about delayed planting in the spring. Among cover crop users, the top five challenges associated with cover 8 crops were establishment, seed cost, the time and labor associated with planting and Interference with cash crop Cover crops may negativel y impact cash crops in a number of ways. In years or locations where moisture is limited, cover crops may deplete soil water reserves needed by the cash crop (Meisinger et al., 1991; Ruffo et al., 2004) . Cover crops with high C:N ratios and large biomass p roduction such as cereal rye can immobilize soil N (De Bruin et al., 2005) to the detriment of a cash crop. When weather or soil conditions are unfavorable for field work, the need to terminate cover crops prior to cash crop planting may lead to delayed pl anting (Long et al., 2013) . Some cover crops can become weeds which compete with the cash crop for resources. Williams and Weil (2004) observed this interaction between canola and soybeans. They attributed poor soybean yield to a canola cover crop which wa s not adequately controlled prior to soybean planting. Living mulches such as alfalfa can also compete with the cash crop to decrease yield (Schmidt et al., 2007) . When intercropped, cover crops such as cereal rye and winter pea can interfere with cash cro p harvest by physically impeding harvest (Hively and Cox, 2001) . Cover otherwise subside during a fallow period (Odhiambo et al., 2012) . Decreased cash crop yield While some studies have shown cover crops to be of benefit to the yield of a following cash crop, other studies have shown the opposite (Table 1.1). In a review by Miguez and Bollero (2005) , corn following cereal rye yielded 1% less than that grown following no cover crop. The meta - analysis by Tonitto et al. (2006) determined that a legume cover crop had to provide at least 110 kg ha - 1 of N for a cash crop to yield the same as plots fertiliz ed with inorganic fertilizer. 9 Otherwise, there was a yield reduction of 10%. Corn following cereal rye suffered a yield depression when insufficient time (eight days) was allowed between cover crop termination and corn planting (Kaspar et al., 2007) . The a uthors suggested that the large amount of cereal rye biomass incorporated into the soil may have interfered with the corn. Terminated cover crops can also attract pests such as seed corn maggot [ Delia platura (Meigen)] that can damage cash crop stand (Cull en and Holm, 2013) . Westgate et al. (2005) observed a decrease in soybean dry matter accumulation and a delay in soybean maturity when soybeans were planted after cereal rye. In one of two years of a study, corn following annual ryegrass yielded less than the no - cover crop control (Hively and Cox, 2001) . Work done in Ontario, Canada, found corn following oats or cereal rye to yield less than a no - cover control (Vyn et al., 2000) . Because of these reductions in yield, a number of researchers contend that cov er crop use is not economically feasible, particularly in the case of non - legume cover crops (Reddy, 2001; Reddy, 2009; Tonitto et al., 2006) . These studies generally examined the economics of cover crop use in the context of a rotation with minimal divers ity, such as continuous corn. In contrast, Smith et al. (2008) assert that the higher cash crop yields seen in the most diverse crop rotations (those including three cash crops and three cover crops) could offset the opportunity cost of planting high - retur n cash crops less frequently. Rotational diversity could act as a buffer against sudden changes in the commodity - market values of cash crops. Types of Cover Crops In the Midwest, three of the most commonly grown groups of cover crops are fall - planted grass es, legumes, and Brassicaceae species. Each type of cover crop has its advantages and its drawbacks. Grasses 10 Grasses are a good choice for fall planting because their large biomass production allows them to take up residual soil NO 3 that might otherwise le ach away over the winter. Their deep roots and large root biomass make them able N scavengers (Meisinger et al., 1991) . In the Midwest, three of the most common grass cover crops are cereal rye, annual ryegrass, and oats. Cereal rye is the hardiest of the three, reliably survives the winter, and is one of the easiest cover crops to establish. Cereal rye is also one of the cover crops that is recognized as being the most likely to interfere with a following cash crop. This observation is due to its large pot ential dry aboveground biomass production, which ranges from 610 - 4640 kg ha - 1 in the Midwest in the fall (Kaspar et al., 2001; Vyn et al., 2000) and up to 6095 kg ha - 1 by the following spring (Ruffo et al., 2004) (Table 1.2). Cereal rye can cause delayed cash crop planting due to the difficulty of terminating it in the spring under adverse weather or soil conditions. It is also known to be allelopathic to following cash crops (Burket et al., 1997) . Further, its re latively high C:N ratio and large biomass production increases the likelihood of N immobilization in the soil after cover crop termination, to the potential detriment of the following cash crop (Crandall et al., 2005) . However, cereal rye remains a commonl y used cover crop in spite of its potential risks because that same cover crop biomass which can make termination difficult also adds C to the soil and acts as a mulch. Its seed is relatively inexpensive and it is one of the hardiest cover crop species, ma king it a likely choice for late fall planting when weather delays the harvest of a preceding crop such as corn or soybean. As with cereal rye, annual ryegrass may not be the best choice for a novice cover crop user due to concerns over termination in th e spring (Creamer et al., 1997; Madden et al., 2004) . When planted in the fall, it survives the winter and can produce 1240 6241 kg ha - 1 of dry aboveground biomass by the following spring (Francis et al., 1998; Kuo and Jellum, 2000) 11 (Table 1.2). One issu e related to annual ryegrass is that of herbicide resistance. The resistance of annual ryegrass to several herbicide classes has been documented in the U.S. (Martins et al., 2012; Perez - Jones et al., 2005) . This is no small concern given that 59% of respon dents in the CTIC survey who grew row crops (2015) relied on herbicides for cover crop termination. Nonetheless, annual ryegrass is one of the few cover crop choices for planting in wet areas, and farmers have been experimenting with its use. There are sev eral ryegrass species ( Lolium spp.) Lolium multiflorum Lam. Like cereal rye, oats has the potential for large biomass production in the fall. It has been reported to produce 36 70 kg ha - 1 of dry aboveground biomass (Kaspar et al., 2001) and 12,500 kg ha - 1 of biomass in regions where it overwinters (Brennan and Smith, 2009) (Table 1.2). Unlike cereal rye, oats does not overwinter in the upper Midwest, and may be a good choice for farmers concerned with spring cover crop termination. Because oats winterkills several weeks to months prior to cash crop planting, there is little risk of N immobilization in the soil negatively impacting a following cash crop. However, oats has been docu mented to negatively impact corn growth in regions where it does overwinter, possibly partially due to allelopathy (Norsworthy, 2004) . When purchased as bin - run seed, oats has one of the lowest seed costs of all the commonly grown cover crops. Legumes Legu me cover crops can form symbiotic relationships with Rhizobacteria spp., which fix atmospheric N in nodules on the legume roots. Farmers make use of this symbiosis to add N to cropping systems. The shallower root systems of legumes means they are less like ly to deplete soil moisture (Nielsen, 2001) , which is of benefit in dry years or locations. In the Midwest, some 12 of the legume cover crops that can be grown include hairy vetch, winter pea ( Pisum sativum var. arvense (L.) Poir. ), and crimson clover. Hairy vetch is the hardiest of the three, and the easiest to establish. It has been documented to produce up to 8900 kg ha - 1 of dry aboveground biomass by May when planted the previous September (Singogo et al., 1996) (Table 1.2). In an Ohio study examining hair y vetch as part of a cover crop mixture, it produced 5500 - 7810 kg ha - 1 of biomass (Creamer et al., 1996) . It is also has some of the highest N - fixation rates: 160 - 265 kg N ha - 1 compared with 10 - 30 kg N ha - 1 accumulated by crimson clover in a study by Cream er et al. (1996) . There is a need for research into ways to successfully establish legume cover crops (Hively and Cox, 2001) . Crimson clover and winter pea are more difficult to establish than hairy vetch. Fall - planted crimson clover has been documented t o produce 968 - 8000 kg ha - 1 of dry aboveground biomass by the following spring (Daniel et al., 1999; Decker et al., 1994) (Table 1.2). As with the other cover crops listed in Table 1.2, there is a large variation in biomass production due to differences i n cover crop genetics, planting date, and site - specific factors such as the weather and soil fertility. In spite of its name, winter pea does not reliably survive the winter in Michigan. Under favorable weather conditions, winter pea can produce similar am ounts of biomass as the other cover crops discussed in this chapter. In a study in North Carolina, winter pea produced 2400 - 6300 kg ha - 1 of biomass and accumulated 67 - 208 kg N ha - 1 (Parr et al., 2011) . While hairy vetch is the most reliable of the three le gumes, its large biomass production can make it difficult to terminate in the spring. Hairy vetch which was not successfully terminated has been documented to negatively impact corn yield through competition (Parr et al., 2011) . Further, it is not suitable for some crop rotations. Farmers who grow small grains are advised to avoid hairy vetch as a cover crop, as it readily volunteers and can be a contaminant of 13 harvested seed. Given these issues and concerns over increasing weather variability, it is sensib le to investigate the use of other legume options such as winter pea and crimson clover. Brassicaceae species Cover crops with rapid fall growth and large biomass production are particularly useful as N scavengers (Meisinger et al., 1991) , and Brassicaceae cover crop species neatly fit these criteria. In a study conducted by Weil and Dean (2009) , rapeseed and oilseed radish shoots scavenged more N in the fall than cereal rye. Brassicaceae cover crop species were another commonly used cover crop type in a re cent survey, with 61% of respondents saying they had planted these cover crops (CTIC, 2015). Among the Brassicaceae species, oilseed radish has become a particularly popular choice in the Midwest (Ngouajio and Mutch, 2004; Sundermeier, 2008) . With regard t o nomenclature, the radishes grown as cover crops are closely related and are usually either oilseed radish ( Raphanus sativus L. var. oleiformis) or forage oilseed radish ( R. sativus L. var. longipinnatus ) (Weil et al., 2009) . The subspecies readily interbreed and management recommendations for both are the same (Weil et al., 2009) . For the purpose of this dissertation, cover crop radishes will all be referred to as oilseed radish. Other Brassicaceae species of interest to far mers include rapeseed or canola (both Brassica napus L.), turnips ( B. rapa L.), white mustards ( Sinapis alba L.), and brown mustards ( B. juncea [L.] Czern.). As with oilseed radish, there is confusion with regard to the common names associated with B. rapa L. A search of the Integrated Taxonomic Information System (itis.gov) reveals that common names for this species include field mustard, rape, mustard rape, turnip rape, and wild turnip. For the purposes of this dissertation, B. rapa L. will be referred to as turnip. Fall - planted Brassicaceae cover crops grow quickly. Rapid and large biomass production is one of several attributes that farmers desire from oilseed radish. It has been shown to produce 14 close to 5000 kg ha - 1 of dry aboveground biomass in the fa ll when planted in August (Table 1.2) (Dean and Weil, 2009; Vyn et al., 1999) . Other reported values range from 890 3947 kg ha - 1 (Stivers - Young, 1998; Vyn et al., 2000) . White mustard has been documented to produce 1626 6397 kg ha - 1 of dry aboveground biomass in the fall (Collins et al., 2006; Stivers - Young, 1998) . Table 1.2 provides an overview of biomass production values found in the literature of yellow mustard, rapeseed, and turnip. In spite of farmer interest in, and marketing efforts involving, Brassicaceae cover crops there has been relatively little research in the Midwest to compare their growth both within and between species. In New York, Stivers - Young (1998) white mustards (referred to as B. hirta L. b ut now called S. alba L.), forage kale ( B. oleracea L.), turnip, and canola and found that the accessions produced the same amount of dry aboveground biomass in the fall, though oilseed radish plots had less surface cover crop residue in the spring than th e other Brassicaceae treatments and oats. All of the cover crop treatments had lower fall soil NO 3 levels than the no - cover crop control plots. Some studies have examined a few accessions of one species such as oilseed radish, or a few accessions from the plant family such as oilseed radish and mustard. Dean and Weil (2009) found no difference in fall dry aboveground - years, oilseed radish produced more dry aboveground biomass and accumulated more shoot tissue N than canola in a study conducted in Quebec, Canada (Isse et al., 1999) . Biomass production is not the only Brassicaceae cover crop attribute of interest. Oilseed radish is reported to have the ability to alleviate soil compaction while suppressing weeds and scavenging nutrients (Snapp et al., 2005) . The latter benefit has caug ht the attention of farmers, 15 who have also expressed interest in the use of Brassicaceae cover crops to help retain N in the system. Thorup - Kristensen (2001) determined that oilseed radish and canola roots grew faster and deeper than those of annual ryegra ss, cereal rye, and oats and took up more soil NO 3 than the grass cover crops. Meisinger et al. (1991) noted that Brassicaceae cover crops are not generally winter hardy, and that they decompose more easily than grass cover crops, allowing rapid re - mineral ization of N the following spring. Fall - planted oilseed radish accessions have been shown to accumulate 100 - 170 kg N ha - 1 (Allison et al., 1998; Axelsen and Kristensen, 2000; Isse et al., 1999; Thorup - Kristensen, 1994; Thorup - Kristensen, 2001; Thorup - Krist ensen, 2006) . Other Brassicaceae species likewise rapidly accumulate soil N. Stivers - Young (1998) found that five weeks after planting, turnip had accumulated 69 kg N ha - 1 while two accessions of white mustards accumulated 51 - 62 kg N ha - 1 . This accumulated N could be of benefit to a following cash crop. There are concerns, however, as to the fate of the N accumulated by fall - planted oilseed radish. Because oilseed radish winter - kills, and because it has a relatively low C:N ratio, there is a window between decomposition in late winter/early spring and cash crop planting during which N released by the cover crop biomass could be lost from the system via surface runoff, leaching, or denitrification and subsequent diffusion into the atmosphere. Cover crop mixtu res Farmers have long been known as innovators (Carlson and Stockwell, 2013) and have recently begun to experiment with the use of cover crop mixtures (multiple species grown at the same time). In spite of farmer interest in cover crop mixtures, there has been relatively little research on the topic (Carlson and Stockwell, 2013) . Research that has been conducted has focused typically on a grass such as cereal rye mixed with a legume such as hairy vetch (Table 1.3). The trend towards the use of cover crop mi xtures is driven by the desire to take advantage 16 of the complementary or synergistic benefits of cover crop species. A slow - to - establish N fixing legume such as hairy vetch pairs well with a rapid - growing Brassicaceae species such as oilseed radish. Furthe rmore, the combination of high and low C:N ratio species may improve N synchrony via reduced N immobilization (Creamer et al., 1997) . Mixing a cover crop with a high C:N ratio such as cereal rye with one with a lower C:N ratio such as oilseed radish may immobilization and too low of a ratio such as too rapid/early N release and subsequent leaching (Odhiambo and Bomke, 2001; Rosecrance et al., 2000) . There is evidence in the literature to support this idea. Gomes et al. (2009) observed that in a vetch ( Vigna sativa L.) plus black oat ( Avena strigosa Schreb) treatment, N 2 O emissions took longer to peak after cover crop termination than in pure legume treatments. Several co ver crops such as hairy vetch, oilseed radish, and cereal rye, are believed to be allelopathic; therefore, using a mixture of cover crops offers the possibility of a wider spectrum of weed suppression (Creamer et al., 1997) . Mixtures have been found to be more productive than cover crop monocultures, and are resilient in the face of extreme weather events (Wortman et al., 2012) . This resiliency may contribute to higher and more consistent cash crop yields. While cover crop mixtures can convey many benefits , there are potential disadvantages to their use. Some cover crops may be better candidates for inclusion in mixtures than others. Furthermore, there is the question of planting rates and proportions. Wortman et al. (2012) found mustard species to be more competitive and productive than legumes when grown in mixtures in Nebraska. Creamer et al. (1997) surveyed a range of cover crop species in 13 multi - species mixtures in Ohio and observed that some species were not suitable due to failure to overwinter or l ow biomass production. When a component species dominates in a mixture, it can also lessen 17 the benefits provided by the other cover crops. In both years of a study by Brainard et al. (2011), soybeans grown as a cover crop in a mixture with sorghum - sudangra ss produced less biomass, nodulated less, and fixed less N than soybeans grown in monoculture. Furthermore, the cost of seed was greater for the mixture than for the sorghum - sudangrass monoculture treatment. When sown later in the fall (mid - to late Septem ber), hairy vetch planted in mixture with cereal rye produced less biomass than when planted in late August in Michigan (Hayden et al., 2015). It may be more difficult for a farmer to predict the effect of a cover crop mixture on a cropping system. Hayden et al. (2014) did not find evidence of synergy between cereal rye and hairy vetch when grown in mixture. The authors noted that altering the seeding proportions of the two species in mixture resulted in tradeoffs among the agroecosystem benefits provided by the cover crops. For example, higher proportions of hairy vetch lead to greater seed costs and less weed suppression but greater amounts of N fixed. More research is needed to determine which species are the best candidates for inclusion in mixtures in Michigan, and which mixtures provide the best combination of benefits and the least likelihood of risk to a following cash crop. Dissertation Objectives In spite of the popularity of cover crops, many questions remain about their growth characteristics and use. Research has not kept pace with farmer innovations on the use of cover crop mixtures. Farmers have requested more data on the benefits of cover crops, and how to successfully implement their use (Carlson and Stockwell, 2013) . Key research priorities of interest to farmers include environmental impacts of cover crops, cover crop effects on cash crop yield, the performance of cover crop mixtures, and the synchronization of nutrient release in cover crop systems (Carlson and Stockwell, 2013) . Kaspar et a l. (2007) likewise identified a need for more research into the management of cover crops to avoid the risk of cash crop yield 18 reduction. Many questions remain to be answered about cover crops. Given the rising popularity of both oilseed radish and cover c rop mixtures, these questions should be addressed with mixtures and oilseed radish in mind. Research topics of interest include: a) cover crop growth potential and cover crop interactions, especially with regard to the difficulty of establishing a mixed co ver crop stand, b) cover crop impact on weeds, c) cover crop impact on cash crop yield as informed by the risk of N immobilization by grass cover crops, and d) the synchronization of N release by cover crops to match the N uptake demands of the following c ash crop. Much has been said about oilseed radish, but there is relatively little research to support or disprove the claims. In particular, there are questions as to the benefits provided by oilseed radish in mixtures with other cover crops. Also of conce rn is the fate of N released by oilseed radish after decomposition. 19 APPENDIX 20 Table 1.1. Overview of fall - planted cover crop impact on corn yield in the literature (assuming appropriate management of cover crop such as allowing adequate time between cover crop termination and corn planting). Common name Scientific name Overall impact on corn yield compared to no - cover crop Reference Annual ryegrass Lolium multiflorum Lam. Negative to neutral Dapaah and Vyn, 1998 Neutral to negative Hively and Cox, 2001 Neutral Isse et al., 1999 Neutral Kuo and Jellum, 2002 Negative to neutral Vyn et al., 1999 Cereal rye Secale cereale L. Neutral Crandall et al., 2005 Neutral to negative Kaspar et al., 2007 Neutral Kuo and Jellum, 2002 Neutral Vyn et al., 2000 Neutral Yenish et al., 1996 Oats Avena sativa L. Neutral Vyn et al., 2000 Crimson clover Trifolium incarnatum L. Neutral to positive Decker et al., 1994 Neutral Isse et al., 1999 Neutral to negative Parr et al., 2011 Neutral Yenish et al., 1996 Hairy vetch Vicia villosa Roth. Positive to neutral Decker et al., 1994 Positive Kuo and Jellum, 2002 Neutral to positive Parr et al., 2011 Neutral Yenish et al., 1996 Winter pea Pisum sativum var. arvense (L.) Poir. Positive to neutral Decker et al., 1994 Neutral to negative Parr et al., 2011 21 Table 1.1 Common name Scientific name Overall impact on corn yield compared to no - cover crop Reference Oilseed radish Raphanus sativus L. var. oleiformis or R. sativus L. var. longipinnatus Positive to neutral Dapaah and Vyn, 1998 Neutral Isse et al., 1999 Neutral to positive Vyn et al., 1999 Neutral Vyn et al., 2000 Annual ryegrass + hairy vetch Lolium multiflorum Lam. + Vicia villosa Roth. Neutral to positive Kuo and Jellum, 2002 Cereal rye + hairy vetch Secale cereale L. + Vicia villosa Roth. Positive to neutral Kuo and Jellum, 2002 neutral Parr et al., 2011 Cereal rye + winter pea Secale cereale L. + Pisum sativum var. arvense (L.) Poir. neutral Parr et al., 2011 - added fertilizer treatment or the treatments averaged across rates (when presented) are summarized here. 22 Table 1.2. Overview of cover crop dry aboveground biomass in the literature. Common name Scientific name Planted Harvested Total dry aboveground biomass (kg ha - 1 ) Reference Annual ryegrass Lolium multiflorum Lam. Jul. Nov. 1280 - 2530 Dapaah and Vyn, 1998 Jun. 1685 Francis et al., 1998 Oct. 3558 - 6241 Francis et al., 1998 Sep. Nov. 302 - 1691 Isse et al., 1999 Sep. - Oct. Apr. - May 1240 4650 Kuo and Jellum, 2000 Oct. Apr. - May 4420 - 4650 Kuo et al., 1997 Aug. Apr. 4310 Odhiambo and Bomke, 2001 Jul. - Aug. Nov. 3500 Thorup - Kristensen, 2001 Apr. - Aug. Nov. 660 - 2620 Vyn et al., 1999 Cereal rye Secale cereale L. Jul. Nov. 1680 Axelsen and Kristensen, 2000 Jul. Mar. 2660 Axelsen and Kristensen, 2000 Oct. Apr. 680 - 2660 Crandall et al., 2005 Oct. Mar. - Apr. 528 - 4600 Daniel et al., 1999 Aug. Oct. - Nov. 1910 - 4640 Kaspar et al., 2001 Sep. - Oct. Apr. - May 250 - 2740 Kaspar et al., 2007 Sep. - Oct. Apr. - May 1420 - 4190 Kuo and Jellum, 200 Oct. Apr. - May 4050 - 4190 Kuo et al., 1997 Nov. Mar. 4970 Norsworthy, 2004 Aug. Apr. 4240 - 5670 Odhiambo and Bomke, 2001 Sep. - Oct. Apr. - May 1800 - 12600 Parr et al., 2011 Oct. Apr. 1500 - 5730 Ranells and Wagger, 1996 Oct. Dec. 600 - 1000 Ranells and Wagger, 1997 Oct. Apr. 3360 - 4630 Ranells and Wagger, 1997 Oct. May 2236 6095 Ruffo et al., 2004 Oct. Nov. Apr. 2280 - 6070 Sainju et al., 2005 Jul. - Aug. Nov. 2100 Thorup - Kristensen, 2001 Aug. Oct. 610 - 1480 Vyn et al., 2000 23 Table 1.2 Common name Scientific name Planted Harvested Total dry aboveground biomass (kg ha - 1 ) Reference Cereal rye Secale cereale L. Aug. Apr. 650 - 2120 Vyn et al., 2000 Oct. Mar. - Apr. 4540 - 5140 Yenish et al., 1996 Oats Avena sativa L. Oct. Dec. - 2000 Brennan and Smith, 2009 Oct. Feb. - 12,500 Brennan and Smith, 2009 Mar.* Jun. 3108 Francis et al., 1998 Mar.* Oct. 9908 Francis et al., 1998 Aug. Oct. - Nov. 2170 - 3670 Kaspar et al., 2001 Nov. Mar. 3690 Norsworthy, 2004 Aug. - Sep. Oct. - Nov. 957 - 3922 Stivers - Young, 1998 Jul. - Aug. Nov. 3100 Thorup - Kristensen, 2001 Aug. Oct. 970 - 1630 Vyn et al., 2000 Crimson clover Trifolium incarnatum L. Oct. - Nov. Apr. 1914 - 5824 Bauer et al., 1993 Oct. Mar. - Apr. 968 - 4270 Daniel et al., 1999 Sep. - Oct. Apr. - May 2100 - 8000 Decker et al., 1994 Sep. Nov. 141 - 1041 Isse et al., 1999 Aug. Apr. 1460 - 4930 Odhiambo and Bomke, 2001 Sep. - Oct. Apr. - May 2000 - 7800 Parr et al., 2011 Oct. Apr. 1420 - 4980 Ranells and Wagger, 1996 Oct. Dec. 90 - 140 Ranells and Wagger, 1997 Oct. Apr. 1120 - 1170 Ranells and Wagger, 1997 Oct. Mar. - Apr. 3500 - 3690 Yenish et al., 1996 Hairy vetch Vicia villosa Roth. Jul. Nov. 1910 Axelsen and Kristensen, 2000 Jul. Mar. 2390 Axelsen and Kristensen, 2000 Oct. Mar. - Apr. 738 - 2890 Daniel et al., 1999 Sep. - Oct. Apr. - May 2700 - 7200 Decker et al., 1994 Sep. - Oct. Apr. - May 2800 - 6600 Parr et al., 2011 24 Table 1.2 Common name Scientific name Planted Harvested Total dry aboveground biomass (kg ha - 1 ) Reference Hairy vetch Vicia villosa Roth. Sep. - Oct. Apr. - May 910 - 3480 Kuo and Jellum, 200 Oct. Apr. - May 2700 - 3480 Kuo et al., 1997 Oct. May 4150 - 4380 Norsworthy et al., 2010 Oct. Apr. 2920 - 4760 Ranells and Wagger, 1996 Oct. Nov. Apr. 2440 - 5100 Sainju et al., 2005 Sep. May 5600 - 8900 Singogo et al., 1996 Oct. Mar. - Apr. 2190 - 2380 Yenish et al., 1996 Winter pea Pisum sativum var. arvense (L.) Poir. Oct. - Nov. Apr. 2673 - 4587 Bauer et al., 1993 Sep. - Oct. Apr. - May 1800 - 6000 Decker et al., 1994 Oct. May 4350 - 4910 Norsworthy et al., 2010 Sep. - Oct. Apr. - May 2400 - 6300 Parr et al., 2011 Sep. May 3200 7600 Singogo et al., 1996 Oilseed radish Raphanus sativus L. var. oleiformis or R. sativus L. var. longipinnatus Jul. Nov. 4750 Axelsen and Kristensen, 2000 Aug. Nov. 2400 - 3640 Dapaah and Vyn, 1998 Aug. Oct. - Nov. 1993 - 4912 Dean and Weil, 2009 Sep. Nov. 176 - 3120 Isse et al., 1999 Aug. - Sep. Oct. - Nov. 1560 - 3947 Stivers - Young, 1998 Jul. - Aug. Nov. 4700 Thorup - Kristensen, 2001 Aug. Nov. 1250 - 4840 Vyn et al., 1999 Aug. Oct. 890 - 1270 Vyn et al., 2000 Mar. May - 3000 Wortman et al., 2012b Rapeseed/canola Brassica napus L. Aug. Oct. - Nov. 2987 - 5053 Dean and Weil, 2009 Nov. Mar. 6800 Hartz et al., 2005 Sep. Oct. 181 - 1960 Isse et al., 1999 Aug. - Sep. Oct. - Nov. 1462 - 4010 Stivers - Young, 1998 Mar. May - 2400 Wortman et al., 2012 25 Table 1.2 Common name Scientific name Planted Harvested Total dry aboveground biomass (kg ha - 1 ) Reference Turnip Brassica rapa L. Aug. - Sep. Oct. - Nov. 1974 - 3140 Stivers - Young, 1998 Jul. - Aug. Nov. 4000 Thorup - Kristensen, 2001 Yellow mustard Brassica juncea [L.] Czern. Nov. Mar. 6700 - 9200 Hartz et al., 2005 Mar. May - 3000 Wortman et al., 2012 White mustard Sinapis alba L. Aug. Oct. 2360 - 6397 Collins et al., 2006 Nov. Mar. 8300 - 9200 Hartz et al., 2005 Aug. - Sep. Oct. - Nov. 1626 - 4358 Stivers - Young, 1998 Mar. May - 3100 Wortman et al., 2012 26 Table 1.3 Cover crop mixture dry aboveground biomass, as reported in the literature. Common name Scientific name Planted Harvested Total dry aboveground biomass (kg ha - 1 ) Reference Annual ryegrass + hairy vetch Lolium multiflorum Lam. Vicia villosa Roth. Fall Apr. 1880 - 2920 Kuo and Jellum, 2002 Annual ryegrass + crimson clover Lolium multiflorum Lam. + Trifolium incarnatum L. Aug. Apr. 3400 Odhiambo and Bomke, 2001 Cereal rye + crimson clover Secale cereale L. + Trifolium incarnatum L. Aug. Apr. 5060 - 6220 Odhiambo and Bomke, 2001 Oct. Apr. 2300 - 5180 Ranells and Wagger, 1996 Oct. Dec. 200 - 700 Ranells and Wagger, 1997 Oct. Apr. 3120 - 3340 Ranells and Wagger, 1997 Cereal rye + hairy vetch Secale cereale L. + Vicia villosa Roth. Oct. Mar. - Apr. 605 - 3650 Daniel et al., 1999 Sep. - Oct. Apr. - May 2700 - 9700 Parr et al., 2011 Oct. Apr. 3010 - 5420 Ranells and Wagger, 1996 Oct. Nov. Apr. 5720 - 8180 Sainju et al., 2005 Cereal rye + winter pea Secale cereale L + Pisum sativum var. arvense (L.) Poir. Sep. - Oct. Apr. - May 4000 - 9600 Parr et al., 2011 Oilseed radish + cereal rye Raphanus sativus L. var. oleiformis + Secale cereale L. Oct. Nov. 667 - 4406 Cavadini, 2013 Aug. - Sep. Nov. 1640 - 2584 White and Weil, 2010 27 Common name Scientific name Planted Harvested Total dry aboveground biomass (kg ha - 1 ) Reference Oilseed radish + oats Raphanus sativus L. var. oleiformis + Avena sativa L. Oct. Nov. 592 - 5201 Cavadini, 2013 Oilseed radish + pea Raphanus sativus L. var. oleiformis + Pisum sativum L. Aug. Oct. 2860 - 3180 Möller and Reents, 2009 Oilseed radish + common vetch Raphanus sativus L. var. oleiformis + Vicia sativa L. 3300 Möller and Reents, 2009 28 LITERATURE CITED 29 LITERATURE CITED Allison, M., M. Armstrong, K. Jaggard and A. Todd. 1998. Integration of nitrate cover crops into sugarbeet ( Beta vulgaris ) rotations. I. Management and effectiveness of nitrate cover crops. J Agr Sci 130:53 - 60. Axelsen, J.A. and K.T. Kristensen. 2000. Collembola and mites in plots fertilised with different types of green manure. Pedobiologia 44:556 - 566. Bauer, P., J. Camberato and S. Roach. 1993. Cotton yield and fiber quality response to green manures and nitrogen. Agron J 85:1019 - 1023. Brainard, D. C., R.R. Bellinder and V. Kumar. 2011. Grass - legume mixtures and soil fertility affect cover crop performance and weed seed production. Weed Technol 25 :473 - 479. Brennan, E.B. and R.F. Smith. 2009. Winter co ver crop growth and weed suppression on the Central Coast of California. Weed Technol 1017 - 1024. Burket, J.Z., D.D. Hemphill and R.P. Dick. 1997. Winter cover crops and nitrogen management in sweet corn and broccoli rotations. HortSci 32:664 - 668. Carlson, S. and R. Stockwell. 2013. Research priorities for advancing adoption of cover crops in agriculture - intensive regions. J Agric Food Syst Community Dev 3:125 - 129. Cavadini, J.S. 2013. Cover crop bicultures and their effects on phosphorus cycling and soil co Publishing # 1549309. Collins, H., J. Delgado, A. Alva and R. Follett. 2007. Use of nitrogen - 15 isotopic techniques to estimate nitrogen cycling from a mustard cover crop to potato es. Agron J 99:27 - 35. Collins, H., A. Alva, R. Boydston, R. Cochran, P. Hamm, A. McGuire and E. Riga. 2006. Soil microbial, fungal, and nematode responses to soil fumigation and cover crops under potato production. Biol Fertil Soils 42:247 - 257. Conklin, A. E., M.S. Erich, M. Liebman, D. Lambert, E.R. Gallandt and W.A. Halteman. 2002. Effects of red clover ( Trifolium pratense ) green manure and compost soil amendments on wild mustard ( Brassica kaber ) growth and incidence of disease. Plant and Soil 238:245 - 256. Crandall, S., M. Ruffo and G. Bollero. 2005. Cropping system and nitrogen dynamics under a cereal winter cover crop preceding corn. Plant and Soil 268:209 - 219. Creamer, N.G., M.A. Bennett and B.R. Stinner. 1997. Evaluation of cover crop mixtures for use i n vegetable production systems. HortSci 32:866 - 870. 30 Creamer, N.G., M.A. Bennett, B.R. Stinner and J. Cardina. 1996. A comparison of four processing tomato production systems differing in cover crop and chemical inputs. J Am Soc Hortic Sci 121:559 - 568. CTIC (Conservation Technology Information Center) and NCR - SARE. 2015. 2014 - 2015 annual report cover crop survey. Available at: http://www.sare.org/Learning - Center/From - the - Field/North - Central - SARE - From - the - Field/2015 - Cover - Crop - Survey - Analysis Accessed 20 July 2015. Cullen, E.M. and K.M. Holm. 2013. Aligning insect IPM programs with a cropping systems perspective: Cover crops and cultural pest control in Wisconsin organic corn and soybean. Agroecol Sust Food Syst 37:550 - 577. Daniel, J., A. Abaye, M. Alley, C. Adcock and J. Maitland. 1999. Winter annual cover crops in a Virginia no - till cotton production system: I. Biomass produ ction, ground cover, and nitrogen assimilation. J Cotton Sci 3:74 - 83. Dapaah, H. and T. Vyn. 1998. Nitrogen fertilization and cover crop effects on soil structural stability and corn performance. Commun Soil Sci Plant Anal 29:2557 - 2569. Davis, A.S. 2010. Cover - crop roller - crimper contributes to weed management in no - till soybean. Weed Sci 58:300 - 309. De Bruin, J.L., P.M. Porter and N.R. Jordan. 2005. Use of a rye cover crop following corn in rotation with soybean in the upper Midwest. Agron J 97:587 - 598. D ean, J.E. and R.R. Weil. 2009. Brassica cover crops for nitrogen retention in the Mid - Atlantic Coastal Plain. Journal of Environmental Quality 38:520 - 528. Decker, A.M., A.J. Clark, J.J. Meisinger, F.R. Mulford and M.S. McIntosh. 1994. Legume cover crop con tributions to no - tillage corn production. Agron J 86:126 - 135. Delate, K., M. Duffy, C. Chase, A. Holste, H. Friedrich and N. Wantate. 2003. An economic comparison of organic and conventional grain crops in a long - term agroecological research (LTAR) site in Iowa. Am J Alternative Agric 18:59 - 69. Francis, G., K. Bartley and F. Tabley. 1998. The effect of winter cover crop management on nitrate leaching losses and crop growth. J Agric Sci 131:299 - 308. Gomes, J., C. Bayer, F. de Souza Costa, M. de Cássia Piccol o, J.A. Zanatta, F.C.B. Vieira and J. Six. 2009. Soil nitrous oxide emissions in long - term cover crops - based rotations under subtropical climate. Soil Till Res 106:36 - 44. Hartz, T., P. Johnstone, E. Miyao and R. Davis. 2005. Mustard cover crops are ineffec tive in suppressing soilborne disease or improving processing tomato yield. HortSci 40:2016 - 2019. 31 Hayden, Z. D., M. Ngouajio and D.C. Brainard. 2014. Rye vetch mixture proportion tradeoffs: Cover crop productivity, nitrogen accumulation, and weed suppressi on. Agron J 106 :904 - 914. Hayden, Z. D., M. Ngouajio and D.C. Brainard. 2015. Planting date and staggered seeding of rye vetch mixtures: Biomass, nitrogen, and legume winter survival. Agronomy Journal 107 : 33 - 40. Hively, W.D. and W.J. Cox. 2001. Interseedin g cover crops into soybean and subsequent corn yields. Agron J 93:308 - 313. Isse, A., A.F. MacKenzie, K. Stewart, D.C. Cloutier and D.L. Smith. 1999. Cover crops and nutrient retention for subsequent sweet corn production. Agron J 91:934 - 939. Kaspar, T., J. Radke and J. Laflen. 2001. Small grain cover crops and wheel traffic effects on infiltration, runoff, and erosion. J Soil Water Conserv 56:160 - 164. Kaspar, T., D. Jaynes, T. Parkin and T. Moorman. 2007. Rye cover crop and gamagrass strip effects on NO con centration and load in tile drainage. J Environ Qual 36:1503 - 1511. Kuo, S. and E. Jellum. 2000. Long - term winter cover cropping effects on corn ( Zea mays L.) production and soil nitrogen availability. Biol Fertil Soils 31:470 - 477. Kuo, S. and E. Jellum. 20 02. Influence of winter cover crop and residue management on soil nitrogen availability and corn. Agron J 94:501 - 508. Kuo, S., U. Sainju and E. Jellum. 1997. Winter cover crop effects on soil organic carbon and carbohydrate in soil. Soil Sci Soc Am J 61:14 5 - 152. Long, E., Q. Ketterings and K. Czymmek. 2013. Survey of cover crop use on New York dairy farms. Crop Management doi: 10.1094/CM - 2013 - 0019 - RS . Madden, N.M., J. Mitchell, W. Lanini, M. Cahn, E. Herrero, S. Park, S. Temple and M. Van Horn. 2004. Evalua tion of conservation tillage and cover crop systems for organic processing tomato production. Hort Technol 14:243 - 250. Martins, B.A.B., E. Sanchez - Olguin, A. Perez - Jones and C. Mallory - Smith. 2012. ACCase cross - resistance in Italian ryegrass ( Lolium multif lorum ) sub - populations from Oregon, US. Julius - Kühn - Archiv 434:177 - 182. Meisinger, J., W. Hargrove, R. Mikkelsen, J. Williams and V. Benson. 1991. Effects of cover crops on groundwater quality, In: W.L. Hargrove (ed.) Cover Crops for Clean Water. Proceedin gs of an international conference. April 9 - 11, 1991., Soil and Water Conservation Society, Ankeny, IA. Jackson, TN. pp. 57 - 68. Miguez, F.E. and G.A. Bollero. 2005. Review of corn yield response under winter cover cropping systems using meta - analytic method s. Crop Sci 45:2318 - 2329. 32 Möller, K. and H.J. Reents. 2009. Effects of various cover crops after peas on nitrate leaching and nitrogen supply to succeeding winter wheat or potato crops. J Plant Nut Soil Sci 172:277 - 287. Ngouajio, M. and D. Mutch. 2004. Oilseed oilseed radish: A new cover crop for Michigan. MSU Extension Bulletin E2907. Nielsen, D.C. 2001. Production functions for chickpea, field pea, and lentil in the central Great Plains. Agron J 93:563 - 569. Norsworthy, J.K. 2004. Small - grain cover crop interaction with glyphosate - resistant corn ( Zea mays ). Weed Technol 18:52 - 59. Norsworthy, J.K. and J.T. Meehan IV. 2005. Use of isothiocyanates for suppression of Palmer amaranth ( Amaranthus palmeri ), pitted morningglory ( Ipomoea lacunosa ), and yellow nutsedge ( Cyperus esculentus ). Weed Sci 53:884 - 890. Norsworthy, J.K. and J.T. Meehan IV. 2009. Herbicidal activity of eight isothiocyanates on Texas panicum ( Panicum texanum ), large crabgrass ( Digitaria sanguinalis ), and sicklepod ( Senna obtusifolia ). Weed Sci 53: 515 - 520. Norsworth, J.K., M.S. Malik, P. Jha and M.J. Oliveira. 2006. Effects of isothiocyanates on purple ( Cyperus rotundus L.) and yellow nutsedge ( Cyperus esculentus L.). Weed Biol Manag 6:131 - 138. Norsworthy, J.K., M. McClelland, G. Griffith, S.K. Bangarwa and J. Still. 2010. Evaluation of legume cover crops and weed control programs in conservation - tillage, enhanced glyphosate - resistant cotton. Weed Technol 24:269 - 274. Odhiambo, J.J. and A.A. Bomke. 2001. Grass and legume cover crop effects on dry matter and nitrogen accumulation. Agron J 93:299 - 307. Odhiambo, J.J., W.D. Temple and A.A. Bomke. 2012. Managing Cover Crops for Conservation Purposes in the Fraser River Delta, British Columbia. In: Crop Management Cases and Tools for Higher Yield and Sustainability. Available at: http://ainfo.cnptia.embrapa.br/digital/bitstream/item/59108 /1/Crop - Management - Cases - and - Tools - for - Higher - Yield - and - Sustainability.pdf#page=49 Accessed 1 February 2015. Parkin, T.B. and T.C. Kaspar. 2006. Nitrous oxide emissions from corn soybean systems in the Midwest. J Environ Qual 35:1496 - 1506. Parr, M., J. Gro ssman, S. Reberg - Horton, C. Brinton and C. Crozier. 2011. Nitrogen delivery from legume cover crops in no - till organic corn production. Agron J 103:1578 - 1590. Perez - Jones, A., K.W. Park, J. Colquhoun, C. Mallory - Smith and D. Shaner. 2005. Identification of glyphosate - resistant Italian ryegrass ( Lolium multiflorum ) in Oregon. Weed Sci 53:775 - 779. 33 Pullaro, T. C., P.C. Marino, D.M. Jackson, H.F. Harrison and A.P. Keinath. 2006. Effects of killed cover crop mulch on weeds, weed seeds, and herbivores. Agric Ecos ys Environ 115 :97 - 104. Ranells, N.N. and M.G. Wagger. 1996. Nitrogen release from grass and legume cover crop monocultures and bicultures. Agron J 88:777 - 882. Ranells, N.N. and M.G. Wagger. 1997. Nitrogen - 15 recovery and release by rye and crimson clover c over crops. Soil Scie Soc Am J 61:943 - 948. Reddy, K.N. 2001. Effects of cereal and legume cover crop residues on weeds, yield, and net return in soybean ( Glycine max ) 1. Weed Technol 15:660 - 668. Reddy, K.N. 2009. Impact of rye cover crop and herbicides on weeds, yield, and net return in narrow - row transgenic and conventional soybean ( Glycine max ). Weed Technol 17:28 - 35 Ribaudo, M., J. Delgado, L. Hansen, M. Livingston, R. Mosheim and J. Williamson. 2011. Nitrogen in agricultural systems: Implications for co nservation policy. USDA - ERS Economic Research Report 127. Robertson, G.P. and P.M. Vitousek. 2009. Nitrogen in agriculture: Balancing the cost of an essential resource. Annu Rev Environ Res 34:97 - 125. Robertson, G.P., E.A. Paul and R.R. Harwood. 2000. Gree nhouse gases in intensive agriculture: Contributions of individual gases to the radiative forcing of the atmosphere. Science 289:1922 - 1925. Rosecrance, R., G. McCarty, D. Shelton and J. Teasdale. 2000. Denitrification and N mineralization from hairy vetch ( Vicia villosa Roth) and rye ( Secale cereale L.) cover crop monocultures and bicultures. Plant and Soil 227:283 - 290. Ruffo, M.L., D.G. Bullock and G.A. Bollero. 2004. Soybean yield as affected by biomass and nitrogen uptake of cereal rye in winter cover cr op rotations. Agron J 96:800 - 805. Sainju, U.M., B.P. Singh and W.F. Whitehead. 2005. Tillage, cover crops, and nitrogen fertilization effects on cotton and sorghum root biomass, carbon, and nitrogen. Agron J 97:1279 - 1290. Schmidt, N.P., M.E. O'Neal and J.W . Singer. 2007. Alfalfa living mulch advances biological control of soybean aphid. Environ Entomol 36:416 - 424. Singer, J., S. Nusser and C. Alf. 2007. Are cover crops being used in the US corn belt? J Soil Water Conserv 62:353 - 358. Singogo, W., W.J. Lamont and C.W. Marr. 1996. Fall - planted cover crops support good yields of muskmelons. Hort Sci 31:62 - 64. 34 Smith, R.G., K.L. Gross and G.P. Robertson. 2008. Effects of crop diversity on agroecosystem function: Crop yield response. Ecosystems 11:355 - 366. Snapp, S ., S. Swinton, R. Labarta, K. O'Neil, J. Nyiraneza, R. Leep, D. Mutch and J. Black. 2005. Evaluating cover crops for benefits, costs and performance within cropping system niches. Agron J 97:322 - 332. Stivers - Young, L. 1998. Growth, nitrogen accumulation, a nd weed suppression by fall cover crops following early harvest of vegetables. Hort Sci 33:60 - 63. Sundermeier, A. 2008. Oilseed oilseed radish cover crop. Ohio State University Extension Fact Sheet SAG - 5 - 08. Thorup - Kristensen, K. 1994. The effect of nitrog en catch crop species on the nitrogen nutrition of succeeding crops. Nut Cycl Agroecosys 37:227 - 234. Thorup - Kristensen, K. 2001. Are differences in root growth of nitrogen catch crops important for their ability to reduce soil nitrate - N content, and how ca n this be measured? Plant and Soil 230:185 - 195. Thorup - Kristensen, K. 2006. Effect of deep and shallow root systems on the dynamics of soil inorganic N during 3 - year crop rotations. Plant and Soil 288:233 - 248. Tonitto, C., M. David and L. Drinkwater. 2006. Replacing bare fallows with cover crops in fertilizer - intensive cropping systems: A meta - analysis of crop yield and N dynamics. Agric Ecosys Environ 112:58 - 72. USDA - NASS. 2013. Crop production 2012 summary. Available at: http://usda01.library.cornell.edu/usda/current/CropProdSu/CropProdSu - 01 - 11 - 2013.pdf . Accessed 1 February 2015. USDA - NASS. 2015. 2013 State agriculture overview: Michigan. Available at: http: //www.nass.usda.gov/Quick_Stats/Ag_Overview/stateOverview.php?state=MICHIG AN Accessed 30 January 2015. Villamil, M., G. Bollero, R. Darmody, F. Simmons and D. Bullock. 2006. No - till corn/soybean systems including winter cover crops. Soil Sci Soc of Am J 70 :1936 - 1944. Vyn, T.J., K.J. Janovicek, M.H. Miller and E.G. Beauchamp. 1999. Soil nitrate accumulation and corn response to preceding small - grain fertilization and cover crops. Agron J 91:17 - 24. Vyn, T.J., J.G. Faber, K.J. Janovicek and E.G. Beauchamp. 200 0. Cover crop effects on nitrogen availability to corn following wheat. Agron J 92:915 - 924. Wagger, M. 1989. Time of desiccation effects on plant composition and subsequent nitrogen release from several winter annual cover crops. Agron J 81:236 - 241. 35 Walsh, M., P. Newman and S. Powles. 2013. Targeting weed seeds in - crop: A new weed control paradigm for global agriculture. Weed Technol 27:431 - 436. Wang, G., M. Ngouajio and D.D. Warncke. 2008. Nutrient cycling, weed suppression, and onion yield following brass ica and sorghum sudangrass cover crops. Hort Technol 18:68 - 74. Weil, R., C. White and Y. Lawley. 2009. Forage oilseed radish: A new multi - purpose cover crop for the Mid - Atlantic. Maryland Cooperative Extension Fact Sheet Fact Sheet 824. Weil, R.R. and J.E. Dean. 2009. Brassica cover crops for nitrogen retention in the Mid - Atlantic coastal plain. J Environ Qual 38:520 - 528. Westgate, L.R., J.W. Singer and K.A. Kohler. 2005. Method and timing of rye control affects soybean development and resource utilization. Agron J 97:806 - 816. White, C.M. and R.R. Weil. 2010. Forage oilseed radish and cereal rye cover crop effects on mycorrhizal fungus colonization of maize roots. Plant and Soil 328:507 - 521. Williams, S.M. and R.R. Weil. 2004. Crop cover root channels may al leviate soil compaction effects on soybean crop. Soil Sci Soc Am J 68:1403 - 1409. Wortman, S.E., C. Francis and J.L. Lindquist. 2012. Cover crop mixtures for the western corn belt: Opportunities for increased productivity and stability. Agron J 104:699 - 705. Yenish, J.P., A.D. Worsham and A.C. York. 1996. Cover crops for herbicide replacement in no - tillage corn ( Zea mays ). Weed Technol 10:815 - 821. 36 CHAPTER 2 GROWTH CHARACTERISTICS OF BRASSICACEOUS SPECIES USED AS FALL - PLANTED COVER CROPS IN MINNESOTA AND MICHIGAN ABSTRACT Research was conducted to evaluate the growth characteristics of two brown mustards ( Brassica juncea rapeseed ( Brassica napus urnip ( Brassica rapa L. x B. napus L. Raphanus sativus L. var. oleiformis and R. sativus L. var. longipinnatus Ampac Seed Co., Tangent, OR) two white mustards ( Sinapis alba . Trials were conducted in St. Paul, MN and Bath, MI for two years. Cover crop performance among Brassicaceae species was generally similar. The only radish. Differences in final oilseed radish dry root biomass also occurred in the Minnesota trial, boveground biomass N uptake was similar among Brassicaceae species, ranging from 100 - 131 kg N ha - 1 in Minnesota and 81 - 109 kg N ha - 1 in Michigan. In both Minnesota and Michigan, ground cover accumulation was similar for the four accessions, which can be expected to provide 50% ground cover within 315 accumulated growing degree days 37 (GDD) in Minnesota and 335 GDD in Michigan. In an average year, 315 - 335 GDD accumulate during the first three weeks of August in both states, indicating how rapidly these accessions can oilseed radishes mustard and hybrid turnip in both trials. Flowering varied considerably among access ions in both hybrid turnip, and the named oilseed radishes did not. There were notable differences in winter survival, as well. The white mustard accessions did not survive the winter. Zero to ten % of the - 88 and 1 - 59 %. The results from this research suggest that in the Midwestern U.S., Brassicaceae cover crops are similar in terms of time to canopy closure, aboveground biomass production, and aboveground N accumulation. If these characteristics are the only ones of concern to a producer, the producer would do well to choose the cheapest or most readily - available seed. INTRODUCTION Cover crops can contribute to the profitability and sustainability of crop production systems (Cherr et al., 2006; O'Reilly et al., 2011) as the costs of inputs such as fertilizer rise. The use of cover crops in the Brassicaceae (mustard family) has become increasingly common (Ngouajio and Mutch, 2004) . Brassicaceae cover crops include mustards (e.g., Brassica juncea [L.] Czern. and Sinapis alba L.) , radishes ( Raphanus sativus L. var. oleiformis and R. sativus L. var. longipinnatus ), rapeseed ( Brassica napus L. ), and turnip ( Brassica rapa L. and B. rapa L. x 38 B. napus L.) . Researchers and university extension educators view these species as valuable additions to the spectrum of av ailable cover crops (Ngouajio and Mutch, 2004; Snapp et al., 2006; Sundermeier, 2008) . With regard to nomenclature, the radishes grown as cover crops are closely related and are usually either oilseed radish ( Raphanus sativus L. var. oleiformis) or forage radish ( R. sativus L. var. longipinnatus ) (Weil et al., 2009) . The subspecies readily interbreed and management recommendations for both are the same (Weil et al., 2009) . Both Brassicaceae cove r crops generally have small seeds, broad leaves, and deep roots (Snapp et al., 2007). Oilseed radish roots can reach a depth of 2.4 m in 11 weeks (Thorup - Kristensen and Kristensen, 2004) . Since they are cool - season annual plants, Brassicaceae cover crops are useful in the often cool and unpredictable climate of the upper U.S. Midwest. These species can germinate at soil temperatures as low as 4°C soil (Snapp et al., 2006). When planted in the fall Brassicaceae cover crops are killed by cold winter temperat ures in climates like Minnesota and Michigan, which allows for easy spring tillage operations (Stivers - Young, 1998) . Producers have expressed interest in the use of Brassicaceae cover crops to help manage N in production systems. Meisinger et al. (1991) st ated in a research review that Brassicaceae cover crops on average reduced the amount of N leached by 60 - 75% as compared with non - cover crop controls. The researchers also noted that Brassicaceae cover crops are not winter hardy, and that they decompose mo re easily than grass cover crops, allowing for rapid re - mineralization of N. Fall - planted oilseed radishes have been found to assimilate 80 - 170 kg N ha - 1 (Allison et al., 1998; Axelsen and Kristensen, 2000; Dean and Weil, 2009; Isse et al., 1999; Thorup - Kr istensen, 1994; Thorup - Kristensen, 2001; Thorup - Kristensen, 2006) . Fall - planted mustards, turnips, and rapeseed have been shown to accumulate 50 - 170 kg N ha - 1 (Brennan and 39 Boyd, 2012; Dean and Weil, 2009; Muir and Bow, 2009; Stivers - Young, 1998) . Even over short growing periods, Brassicaceae species are efficient N scavengers. Stivers - Young (1998) found that five weeks after planting, turnip ( B. rapa ) had taken up 69 kg N ha - 1 while two accessions of mustard ( S. alba, formerly B. hirta ) had accumulated 51 - 6 2 kg N ha - 1 . In a study by Brennan and Boyd (2012) , mustards (a mixture of S. alba and B. juncea ) accumulated N at a faster rate than cereal rye and a legume/cereal rye mixture. Brassicaceae cover crops can improve soil physical properties. Oilseed radish, in (Chen and Weil, 2010) . In a study using a mini rhizotron camera, William s and Weil (2004) observed soybean ( Glycine max (L.) Merr.) roots growing through channels in compacted soil created by oilseed radish and canola roots. Lehrsch and Gallian (2010) found that oilseed radish improved soil characteristics related to water dyn amics (e.g., hydraulic conductivity and water infiltration through pores). Oilseed radish can increase soil aggregate stability, though this effect may be short - lived in conventional tillage systems (Dapaah and Vyn, 1998) . Brassicaceae cover crops are a lso of interest as a weed - suppression tool. Cover crops can create environmental conditions unfavorable for weed germination and growth. They can encourage the activity of weed seed predators and weed seedling pathogens (Conklin et al., 2002; Pullaro et a l., 2006); act as a mulch that smothers or shades out weeds (Smith et al., 2008; Stivers - Young, 1998); compete with weeds for space, nutrients, or light (Linares et al., 2008); and some cover crops are allelopathic (Weston, 1996) . Among Brassicaceae cover crops, oilseed radish is a good choice for the suppression of winter annual weeds in the fall due to quick growth and large biomass accumulation (Ngouajio and Mutch, 2004; Stivers - Young, 1998; Sundermeier, 2008) - 1 dry 40 above - ground biomass in a study by Dean and Weil (2009) . Other studies have found oilseed radish can produce 3,000 - 5,600 kg ha - 1 of biomass when planted in the summer or fall (Allison et al., 1998; Axelsen and Krist ensen, 2000; Isse et al., 1999; Thorup - Kristensen, 2001; Thorup - Kristensen, 2006) . Brassicaceae cover crops also have the potential to reduce nematode and soil - borne , can be used as a trap - crop for sugar beet cyst nematodes (Smith et al., 2004) . Brassicaceae cover crops produce glucosinolates, which upon breakdown form compounds [including isothiocyanates (ITCs)] that have been shown to be biocidal against soil - borne pathogens (Angus et al., 1994; Brown and Morra, 1997; Dunne et al., 2003; Kirkegaard et al., 1996; Sarwar et al., 1998) . Greenhouse and field studies have shown that Brassicaceae residues can also reduce disease incidence in some cases (Blok et al., 2000; Larkin and Griffin, 2007; Snapp et al., 2007), but not others (Bensen et al., 2009; Hartz et al., 2005; Wiggins and Kinkel, 2005). Research into Brassicaceae cover crops ranges across the U.S. and Canada (Collins et al., 2007 ; Dean and Weil, 2009; Vyn et a l., 2000) . Kaspar et al. (2007) identified several priorities for cover crop research, including the need for more information on specific cover crop cultivars. Furthermore, Cherr et al. (2006) argued that studies in which cover crop growth parameters are sampled multiple times over the course of the growing season may provide more meaningful information to producers than those which only collect data once. Finally, the U.S. Department of Agriculture Natural Resources Conservation Service (USDA - NRCS) has in dicated a need for data on the growth of Brassicaceae cover crop species for inclusion into planning tools such as RUSLE2 (Revised Universal Soil Loss Equation, Version 2) and WEPS (Wind Erosion Prediction System), which assist producers in the selection o f cover crops (J. Douglas, pers. 41 comm., 2010). The objectives of this research were to: a) compare the performance of Brassicaceae cover crop accessions within and between species, b) evaluate Brassicaceae cover crop performance in two Midwestern states (M innesota and Michigan), and c) collect Brassicaceae cover crop growth data for use in USDA - NRCS models. MATERIALS AND METHODS Four mustards (two accessions of brown mustard and two accessions of white mustard), one rapeseed, one hybrid turnip, and six accessions of oilseed radish were evaluated in Minnesota and Michigan during the falls of 2010 and 2011. Individual accession names can be found in Table 2.1. Site description, experimental design, and field management The Minnesota trial was located on the University of Minnesota St. Paul campus (44.99 o N; 93.19 o W), where the soil was a well - drained Waukegan silt - loam (fine - silty over sandy mixed super active mesic Typic Hapludolls). The Michigan trial was located at t he Natural Resources Conservation Service (USDA - NRCS) Rose Lake Plant Materials Center (PMC) (42.81 o N; 84.44 o W) in Bath, MI. The soil was a poorly drained Colwood loam (fine - loamy mixed active mesic Typic Endoaquolls). Field preparation and planting date s and methods are listed in Table 2.2. Both trials were conducted using a randomized complete block design with four replicates. Experimental units were a minimum of 3 x 10 m. Planting rates were: mustards 9.0 kg ha - 1 , oilseed radishes 11.2 kg ha - 1 , rapese ed 5.6 kg ha - 1 , and turnip 2.2 kg ha - 1 . In Minnesota in 2011 planting rates were adjusted so that each planting rate was of pure live seed (PLS). Adjustments were based on germination tests. Cover crops were drilled in August 42 September of each year (Tabl e 2.2). In Minnesota, plots were hand - weeded 7 d after planting (DAP) in 2010. Volunteer oat and annual grasses were controlled in the Michigan trial with quizalofop (quizalofop p - ethyl ethyl(r) - 2 - [4 - (6 - chloroquinoxalin - 2 - yloxy) - phenoxypropionate) at 0.0 7 kg a.i. ha - 1 + 1% v v - 1 crop oil concentrate. Data collection trial, % ground cover (amount of canopy closure) was estimated every 10 - 14 d by counting the number of centimeter marks visible on a meter stick placed horizontally on the ground under the canopy. In the Michigan trial % cover was estimated every 10 - 14 d using the grazin g stick method (Smith et al., 2010) . The grazing stick had a grid on which ten dots were printed. % cover was estimated by placing the grazing stick below the cover crop canopy, counting the number of visible dots on the grid, subtracting the result from t en, and then multiplying by 100. Above - and belowground biomass data were also collected every 10 - 14 d. In the Minnesota trial, one subsample per plot was collected by harvesting plants in a 0.25 m 2 quadrat. In the Michigan trial, one subsample per plot wa s collected using a 0.2 m 2 quadrat. Quadrat locations were flagged to avoid sampling in the same area multiple times. Biomass samples were collected from all twelve Brassicaceae accessions once each year in the fall, before the onset of a hard freeze. In t he Minnesota trial, two 0.25 m 2 subsamples were collected per plot. In Michigan, one 0.2 - m 2 subsample was collected per plot. In both trials, plants in each quadrat were pulled by hand and separated into root and shoot fractions. Samples - 21 d, and weighed. These samples were then analyzed for tissue N 43 concentrations. In the Minnesota trial, tissue N was determined by combustion using a ThermoFinnigan FlashEA organic elemental analyzer (Thermo Fisher Scientific Inc., Waltham, MA) rape seed in 2010. For all other accessions, tissue N was determined with near infrared spectroscopy (NIRS) using a Perten DA 7250 NIRS analyzer (Perten Instruments, Inc., Springfield, IL) . Different analysis methods were employed because of cost - effectiveness concerns. In the Michigan trial in 2010, samples were analyzed for total Kjeldahl nitrogen (TKN) content using the Hach method (Watkins et al., 1987) . In 2011 samples were tested for TKN by A&L Great Lakes Laboratories (Fort Wayne, IN) using plant TKN AOAC method 978.04 (AOAC, 2012) . Root and shoot N were analyzed separately in both Minnesota and Michigan. Flowering status of the Brassicaceae species was noted every 10 - 14 d throughout the fall. Winter survival was determined in March - April by counting the n umber of green cover crop plants and calculating the % survival from cover crop stand counts the previous fall. The exception was in spring 2012 in Minnesota where % survival was estimated visually. Weather data and growing degree day calculations Daily a ir temperature and precipitation data for 2010 - 2012 in St. Paul, MN were obtained from the Minnesota Department of Natural Resources interactive online tool (MDNR, 2015) (Figure 2.1). Weather data for Bath, MI were obtained from the Michigan State Universi ty Enviro - weather website (MSUEW, 2015), with the exception of December - April precipitation data which were collected from the Haslett, MI weather station via the NOAA website (NOAA, 2015). The 30 - yr monthly averages (1981 - 2010) for both locations were obt ained from the National Oceanic and Atmospheric Administration (NOAA) website (NOAA, 2015). 44 Temperature data were used to calculate growing degree days (GDD) from the time of planting using the standard formula GDD = (T max + T min )/2 T base where T max is the maximum daily temperature and T min is the minimum daily temperature. T base (Morrison et al., 1989) . T max and T min values less than 5°C were set to 5°C before calculating GDD (McMaster and Wilhelm, 1997) . Data analysis Data were checked for normality and homogeneity of variance. Aboveground final biomass, N accumulation, and oilseed radish root N accumulation were log base 10 transformed to improve normality. Data were analyzed using PROC MIXED in SAS Enterprise Guide v. 6.1 (SAS Institute Inc., Cary, NC). Replication was treated as a random factor. The Tukey - Kramer Means were separated using the PDMIX 800 macro (Saxton, 1998) . Nitro gen accumulation was determined by multiplying tissue N content by biomass. Percent ground cover and biomass accumulation data were collected season - long for turnip. PROC GLM in SAS Enterprise Guide v. 6.1 (SAS Institute Inc., Cary, NC) was used to fit these data to second order polynomial (quadratic) equations in the form of y = ax 2 + ax + b where y is biomass accumulated or percent ground cover for a given accumulati on of GDD, x is GDD accumulated, and b is the y - intercept of the equation. Data for each accession were combined across years to investigate the effect of GDD on biomass accumulation and percent ground cover 45 (Hayden, 2014). GraphPad Prism v.6.02 (GraphPad Software, La Jolla, CA) was used to conduct a n extra sum - of squares F - test to determine whether one line could be fitted to the ground cover - test was significant, the nul only equations for the individual accessions are presented. When the F - test was not significant, the general equation for the pooled data is presented. In all cases, the dependen t variable was either percent cover or dry aboveground biomass, while the independent variable was GDD. RESULTS AND DISCUSSION Weather The Brassicaceae cover crop growing season generally encompasses August November, depending on planting date and the w eather. September 2010 was 19% cooler than average in Minnesota (284 accumulated GDD compared with 338 GDD), but only 12% cooler than average in Michigan (319 vs. 358 GDD) (Figure 2.1). Growing degree days are calculated from temperature data and thus can serve as a rough proxy for these data. Winter and early spring temperatures are of interest because they can affect the overwintering of cover crops. March of 2012 was warmer than average in both Minnesota and Michigan (178 and 211 GDD, respectively, vs. a n average of 0 GDD for both locations). There were some differences in precipitation (Figure 2.1). It should be noted that the first planting date in each trial - year did not occur before mid - August, and the last data collection date in each trial - year occ urred between 1 November and 14 November. In Minnesota, August - November 2010 was 25% wetter than average with a total precipitation of 399.3 mm as compared with the norm of 297.3 mm. In contrast, in Michigan August November 2010 was 46 20% drier than aver age with 205.2 mm of precipitation as compared with 246.6 mm. The trends reversed in 2011, when Minnesota experienced a dry period from August November in which total precipitation was 62% of average. Precipitation in Michigan during August November 20 11 was within norms and totaled 253.8 mm, compared with the average of 246.6 mm. Ground cover and biomass accumulation of four Brassicaceae accessions Ground cover accumulation. An extra - sum of squares F - test (H 0 : one equation fits pooled accession data) was not significant for both the Minnesota and Michigan data (p = 0.18 and 0.71, respectively). A single second - order polynomial equation was thus fitted to the pooled and fitted equation is y= - 0.0006x 2 + 0.80x 142 where y is percent ground cover at a given number of GDD and x is GDD (Figure 2.2). The r 2 for this equation is 0.85, P = <0.0001. According to this equation, all four Brassicaceae accessions would be expected to reach 50% ground cover at 315 accumulated GDD. Based on the 30 - yr average monthly GDDs (Figure 2.1), if these cover crops were planted on 1 August they wo uld achieve 50% ground cover within three weeks of planting. While a single line fits all four declined earlier than the other Brassicaceae accessions (Figure 2.2). Mus tards are more cold - sensitive than oilseed radish and hybrid turnip (data not shown). Freezing weather at the end of 47 In the Michigan trial, the equation fitt ed to the percent ground cover data for all four accessions across two years is y= - 0.0005x 2 + 0.65x 112 where y is percent ground cover and x is GDD (Figure 2.3). The r 2 for this equation is 0.89, P = <0.0001. Based on this equation, the four accessions could be expected to reach 50 % ground in Minnesota it could reach 50% ground cover within three weeks of planting, based on 30 - yr average monthly GDD accumulation (Figure 2.1). Averaged across the two years, all four accessions reached peak ground cover (90% or greater) by 550 - 555 GDD in both Minnesota and Michigan. Rapid co ver crop canopy closure is one mechanism through which cover crops confer benefits to a crop production system. Good canopy cover suppresses weeds by shading the ground and preventing emergence of winter annual weeds (Lawley et al., 2012; O'Reilly et al., 2011) and also helps decrease soil erosion by protecting the soil from the impact of raindrops (Dabney et al., 2001) . While the equations in our study were generated from data collected from four accessions, rapid ground cover was also observed in the full set of twelve Brassicaceae accessions studied (data not shown). Progression of % ground cover by Brassicaceae cover crops is not commonly reported in the literature. Stivers - Young (1998) found that in the northeastern USA, R. sativus and S. alba (formerly B. hirta ) reached 23 - 33% ground cover at 20 DAP and 100% ground cover by seven WAP; in another year of that study, ground cover for six different Brassicaceae accessions (including R. sativus , B. napus , B. rapa , and S. alba [formerly B. hirta ]) ranged fro m 40 - 68% at 18 DAP. 48 Aboveground biomass accumulation. An extra - sum of squares F - test (H 0 : one equation fits pooled accession data) was significant for both the Minnesota and Michigan data (p = 0.006 and 0.02, respectively). A single second - order polynomial equation thus could not be fitted to the pooled biomass accumulation data in each trial. However, in Minnesota a single line was fitted to lseed radish data (p = 0.99). The equations and r 2 and P values for the Minnesota trial are provided in Figure 2.4. Based on these equations, it would take roughly 485 GDD for the oilseed radishes to accumulate 2,000 kg ha - 1 - 1 biomass. In seed radishes and thus had steeper sloping lines. Growth rates for both groups were similar until near 500 GDD, at which radish accessions. As in Minnesota, in Mi 0.88).The equations and r 2 and P values for the Michigan trial are provided in Figure 2.5. Based on oilseed radishes to accumulate 2,000 kg ha - 1 of dry aboveground biomass, vs. 540 GDD for higan the 49 turnip, up until near 500 accumulated GDD when oilseed radish growth rate slowed as the mustard and turnip growth rate accelerated. Rapid and large cover crop biomass production is of interest for several reasons. It allows cover crops to outcompete weeds (O'Reilly et al., 2011) . Fast root growth makes Brassicaceae cover crops good N scavengers (Meisinger et al., 1991) . Since Brassicaceae cover crops can also be used as forage (Barry, 2013) , biomass production potential could be of interest to producers whose systems include livestock. Characteristics of all twelve Brassicaceae accessions Final biomass. Fall aboveground biomass data were combined over years for each trial since the accession*year interactions were not significant (P = 0.41 and 0.32 in the Minnesota and differences were detected. Nor were there any differences among the Brassicaceae accessions in final dry aboveground biomass produced in Michigan (Table 2.3). The failure to detect differences in the Michigan data may have been the result of insufficient subsample size and number. In Minnesota, biomass ranged from 3311 kg ha - 1 - 1 ard) (Table 2.3). In Michigan biomass ranged from 2476 kg ha - 1 rapeseed) to 3747 kg ha - 1 were likely the result of varying weather and soil fertility. Mustard, rapeseed, and hybrid turnip final root biomass data are not presented because the sampling protocol was insufficient to capture small, fibrous roots. Fall final oilseed radish 50 dry root biomass data were combined over years for each trial because the accession*year i nteraction was almost not significant (P = 0.04) in Minnesota and not significant (P = 0.78) in ed radishes (Table 2.4). Oilseed radish root biomass ranged from 1226 - 2483 kg ha - 1 in Minnesota and 846 - 1294 kg ha - 1 in Michigan. No differences were detected in root biomass production in the Michigan trial. The failure to detect differences may have been the result of insufficient subsample size and number in Michigan, or because roots were hand - pulled instead of excavated with a shovel. Variability in the data may also have impacted the ability to detect differences. Aboveground biomass production was in general accord with the literature. As in our Michigan trial, Stivers - Young (1998) detected no differences in biomass in northeastern U.S. early - fall plantings of oilseed radish ( R. sativus ), two accessions of mustards ( S. alba , formerly B. hirta ), turnip ( B. rapa ), and rapeseed ( B. napus ). However, for a later planting of Brassicaceae cover crops, by late October turnip had more biomass than the other Brassicaceae species, a lead it maintained at the mid November biomass sampling (Stivers - Young, 1998) . Ha d biomass sampling in our study continued later into the fall, it is possible that differences in Michigan biomass might have been detected. Dry matter production in the Stivers - Young (1998) study ranged from 2525 kg ha - 1 to 2826 kg ha - 1 by late October, l ess than many accessions produced in our Minnesota trial in the same time frame, but close to the range produced in our Michigan trial (Table 2.6). In our Minnesota trial, biomass production by mid - October ranged from 3311 to 5542 kg ha - 1 , while that found by Stivers - Young (1998) in mid - November ranged from 3947 kg ha - 1 to 4358 kg ha - 1 (Stivers - Young, 1998) . 51 Nitrogen content. Tissue N concentration data allow for a partial estimation of N accumulated ha - 1 when combined with biomass production data (Table 2 .3). There was no treatment*year interaction for either location (P = 0.17 and 0.33 in Minnesota and Michigan, respectively), so aboveground N accumulation data were combined over years. There were no differences detected in aboveground N accumulation for the different Brassicaceae accessions in both Minnesota and Michigan. In Minnesota, N accumulation ranged from 100 - 131 kg N ha - 1 and in Michigan, N accumulation ranged from 81 - 109 kg N ha - 1 . Nitrogen uptake was lower in the Michigan trial than the Minneso ta trial, possibly due biomass production differences caused by varying soil fertility and weather conditions. The treatment*year interaction was barely significant in the Minnesota trial (P = 0.04) and not significant in Michigan (0.40), so oilseed radish root N accumulation data were lated 39 - 59 % more N than the other oilseed radish accessions. This is not a surprise, given the large root the Minnesota trial. Nor were differences detected in the Michigan trial. Aboveground tissue N accumulation in our trials were similar to those found in other studies. Fall - planted oilseed radishes have been shown to accumulate 100 - 170 kg N ha - 1 (Allison et al., 1998; Axelsen and Kristensen, 2000; Isse e t al., 1999; Thorup - Kristensen, 1994; Thorup - Kristensen, 2001; Thorup - Kristensen, 2006). In our Minnesota trial, all accessions had accumulated at least 100 kg N ha - 1 by roughly eight weeks after planting, while in Michigan the smallest amount of N accumul ated by that point was 81 kg ha - 1 (Tables 2.1 and 2.3). Five weeks after planting, Stivers - Young (1998) found turnip to have accumulated 69 kg N ha - 1 while two white mustard accessions had accumulated 51 kg N ha - 1 . 52 Flowering and seed production Flowering varied among cover crop accessions in both trials (Table 2.7). The brown mustard accessions flowered within 45 days after planting (DAP), except in the Michigan trial in (Table 2.5). The white mustard accessions flowered within 37 - 65 DAP, with the exception of VNS oilseed radish accessions flowered within 37 - 78 DAP. Daikon VNS #1 di d not flower in the Michigan trial in 2011. Rapeseed, hybrid turnip, and the named oilseed radish accessions did not flower in either location in either year. Differences in flowering were likely due to the interaction between accession genotype and the en vironment. Factors affecting time to flowering include temperature and photoperiod (Nanda et al., 1996; Robertson et al., 2002). In a study of four Brassicaceae species including B. juncea (brown mustard), cooler temperatures were calculated to shorten the time to flowering by 22 - 41 growing degree days (base temperature 0 °C) for every 1 °C decline in average temperature (Nanda et al., 1996). Within a Brassicaceae species, different genotypes respond differently to temperature and photoperiod (Robertson et al., 2002). Accessions that flower may be of interest to producers who want to provide a nectar source for beneficial insects. Insects were observed visiting mustard accession flowers as late as November in the Michigan trial (data not shown). Time to flo wering is also of interest because plants which have reached the reproductive phase cease allocation of resources to vegetative structures (Rossato et al., 2001). Early flowering Brassicaceae accessions could provide reduced ecosystem services such as weed suppression due to potentially lower biomass production and 53 less canopy closure. Time to flowering is also of concern because Brassicaceae cover crop plants may produce seed to add to the weed seed bank. Although none of the accessions in this study set m ature seed (data not shown), producers who grow Brassicaceae cover crops should monitor their crop and be prepared to control them with mowing or herbicides before seed set occurs. Brassicaceae cover crops could also become established as weeds via seeds w hich are planted but fail to germinate immediately, instead persisting as part of the soil seed bank. In Michigan, newly - germinated oilseed radish seedlings were observed in the spring both years, seven months after the cover crops were planted (data not s hown). Winter hardiness - 59% and 7 - 88%, respectively, across years in Minnesota and Michigan. In the Minnesota t - 10% of several oilseed radish accessions and both brown mustards also survived winter 2011 - 2012 (Table 2.5). Increased winter survival in 2011 - 2012 was likely due to the weather. From December 2011 - March 2012, there were 188 and 248 GDD accumulated in Minnesota and Michigan, respectively, compared to 15 and 50 GDD over the same period in 2010 - 2011 (Figure 2.1). Cover crops that overwinter may provide bene fits such as continuous ground cover in the spring that decreases soil erosion. However, the need to terminate surviving cover crops may delay cash crop planting in wet years. If spring logistics are of concern to producers, cold - sensitive cover crop speci es such as mustards are a better choice. 54 CONCLUSIONS The Brassicaceae accessions in these trials had similar growth characteristics and were suitable for use as fall - planted cover crops. With regard to percent ground cover and biomass accumulation, final aboveground biomass, and N uptake, little differentiation was detected. If these characteristics are the only ones of concern to a producer, the producer would do well to choose the cheapest or most readily - available seed. However, Brassicaceae cover crop do differ with regard to other traits, which could affect cover crop selection. In our trials, white and brown mustards flowered before winter - kill, while rapeseed, hybrid turnip, and the named oilseed radish accessions did not. In terms of winter surviva l, rapeseed and hybrid turnip had the highest incidence of over - wintering while mustards and oilseed radishes had the lowest incidence. Though not quantified in our trials, there are also known differences in Brassicaceae cover crop pest suppression. For e - crop for sugar beet cyst nematode. Plant breeders continue to develop and refine Brassicaceae germplasm, so future work could continue to evaluate oilseed radish, turnip, and mustard co ver crop accessions. Root biomass could be better quantified by excavating roots with a shovel, rather than pulling the plants by hand. It would be interesting to trial accessions under varying fertility management regimes, to better estimate how fall - plan ted Brassicaceae cover crops perform under limited vs. optimal nutrient conditions. 55 APPENDICES 56 APPENDIX A CHAPTER 2 TABLES AND FIGURES 57 58 59 60 61 Figure 2.1. Monthly and 30 - yr average growing degree days (base 5 C) and precipitation for St. Paul, MN (top) and Bath, MI (bottom) in 2010 and 2011. Growing degree days are represented on the left y - axis and by lines, while precipitation is represented on the right y - axis and by bars. 0 40 80 120 160 200 0 100 200 300 400 500 600 Precipitation (mm) Monthly growing degree days 2010-2011 2011-2012 30-yr ave. 2010-2011 2011-2012 30-yr ave. GDD precip 0 40 80 120 160 200 0 100 200 300 400 500 600 Precipitation (mm) Monthly growing degree days 62 Figure 2.2. ray triangle) ground cover accumulation in Minnesota in 2011 and 2012. Polynomial regression was conducted with growing degree days (GDD) (base 5 °C) and percent ground cover as covariates. Means ± SE. y= - 0.0006x 2 + 0.80x 142 r 2 = 0.85 P = <0.0001 63 Figure 2.3. accumulation in Michigan in 2011 and 2012. Polynomial regression was conducted with growing degree days (GDD) (base 5 °C) and percent ground cover as covariates. Means ± SE. y = - 0.0005x 2 + 0.65 x 112 r 2 = 0.89 P = <0.0001 64 Figure 2.4. oilseed radish (white diamond), and Groundh accumulation in Minnesota in 2011 and 2012. Polynomial regression was conducted with growing degree days (GDD) (base 5 °C) and biomass as covariates. Means ± SE. Response of ass production to GDD: y = 0.0099x 2 + 1.84x - 1102, r 2 = 0.91, P to GDD: y = - 0.0034x 2 + 11.45x - 2754, r 2 = 0.94, P = <0.0001. 65 Figure 2.5. accumulation in Michigan in 2011 and 2012. Polynomial regression was conducted with growing de 2 + 1.61x - 587, r 2 = 0.91, P = duction to GDD: y = - 0.0058x 2 + 11.88x - 2254, r 2 = 0.97, P = <0.0001. 66 APPENDIX B NON - OILSEED RADISH BRASSICACEAE ROOT N CONTENT ANALYSIS OF VARIANCE 67 LITERATURE CITED 68 LITERATURE CITED 69 Hartz, T.K., P.R. Johnstone, E.M. Miyao and R.M. Davis. 2005. Mustard cover crops are ineffective in suppressing soilborne dis ease or improving processing tomato yield. HortScience 40: 2016 - 2019. Hayden, Z. D. 2014. Optimizing rye - vetch cover crop mixture management in vegetable cropping systems: Opportunities and tradeoffs. Michigan State University Ph.D. dissertation. ProQuest LLC UMI number 3618536 . 70 MDNR ( Minnes ota Dept. of Natural Resources). 2015. Retrieve climate data from National Weather Service reporting stations. Available at: http://www.dnr.state.mn.us/climate/historical/acis_stn_meta.html Accessed 29 June 2015. MSUEW ( Michigan State University Enviro - weather). 2015. Enviro - weather automated weather station network. Available at: http://www.agweather.geo.msu.edu/mawn/ Accessed 1 July 2015. NOAA ( National Oceanic and Atmospheric Administration). 2015. Climate data online search. Available at: http://www.ncdc.noaa.gov/cdo - web/search?datasetid=GHCND Accesse d 1 July 2015. Pullaro T. C., Marino P. C., Jackson D. M., Harrison H. F., Keinath A. P. (2006) Effects of killed cover crop mulch on weeds, weed seeds, and herbivores. Agric Ecosys Environ 115 : 97 - 104. 71 Robertson, M. J., S. Asseng, J.A. Kirkegaard, N. Wratten , J.F. Holland, A.R. Watkinson, T.D. Potter, W. Burton, G.H. Walton, D.J. Moot and I. Farre. 2002. Environmental and genotypic control of time to flowering in canola and Indian mustard. Crop Pasture Sci 53 : 793 - 809. Rossato, L., P. Laine, P. and A. Ourry. 2001. Nitrogen storage and remobilization in Brassica napus L. during the growth cycle: nitrogen fluxes within the plant and changes in soluble protein patterns. J Exp Bot 52 : 1655 - 1663. Smith R.G., K.L. Gross and G.P. Robertson. 2008. Effects of crop diversity on agroecosy stem function: Crop yield response. Ecosys 11: 355 - 366. 72 Wiggins, B.E. and L.L. Kinkel. 2005. Green manures and crop sequences influence potato diseases and pathogen inhibitory activity of indigenous Streptomycetes. Phytopathology 95: 178 - 185. 73 CHAPTER 3 PERFORMANCE OF COVER CROP MONOCULTURES VS. BICULTURES: COVER CROP GROWTH AND IMPACT ON WEED BIOMASS AND CORN ( Zea mays L.) YIELD ABSTRACT A three - year experiment was conducted in Lansing, MI to quantify biomass production of cover crop monocultures and biculture mixtures and investigate the impact of the cover crop treatments on fall and spring weed biomass production and corn ( Zea mays L.) grain yield in a conventionally tilled system with no additional applied fertilizer. Cover crops investigated consisted of annual ryegrass ( Lolium multiflorum Lam.), cereal rye ( Secale cereale L.), oat s ( Avena sativa L.), crimson clover ( Trifolium incarnatum L.), hairy vetch ( Vicia villosa Roth.), and winter pea [ Pisum sativum var. arvense (L.) Poir.] grown in monoculture and in biculture with oilseed radish ( Raphanus sativus L. var. oleiformis Pers. ). Oats and oilseed radish produced 2 - 11 times more fall aboveground biomass than crimson clover, the cover crop treatment with the smallest biomass production in all three years. Crimson clover failed to establish in two of the three years, while winter p ea failed to survive the winter in two of the three years. Within each year, biomass production was similar across cover crop biculture treatments. Oilseed radish biomass exceeded that of the second species in each biculture, accounting for 54 - 99% of total biomass. Monoculture treatments produced more spring biomass than bicultures because the seeding rate of the complementary species in biculture was half that in monoculture. In addition, biomass relative yield values indicate that the oilseed radish compe ted with each complementary species in biculture. Oilseed radish competition with complementary species was greatest in the oilseed radish + crimson clover and oilseed radish + winter pea treatments and lowest in the 74 oilseed radish + cereal rye and oilseed radish + oats treatments. Overall, the cover crops did not reduce corn grain yield, with the exceptions of annual ryegrass and cereal rye each in one of three years. Annual ryegrass and cereal rye reduced corn grain yield by 51 and 24%, respectively, comp ared with the weedy control. INTRODUCTION The environmental services provided by cover crops are well documented. Cover crops can prevent soil erosion and improve soil physical properties (Chen and Weil, 2010; Hively and Cox, 2001; Kaspar et al., 2001; M eisinger et al., 1991; Villamil et al., 2006; Williams and Weil, 2004) . Cover crops provide food and habitat for micro - and macro - organisms ranging from collembola to birds (Axelsen and Kristensen, 2000; Blanchart et al., 2006; Decourtye et al., 2010; Hend erson et al., 2004) . Cover crops can decrease the amount of N leached from agricultural systems (Kaspar et al., 2007; Ruffo et al., 2004; Tonitto et al., 2006) . The use of cover crops has been suggested as one way to help decrease the nutrient load in the Mississippi River drainage basin that contributes to the hypoxic zone in the Gulf of Mexico (Strock et al., 2004) . In spite of the benefits listed above, cover crop adoption remains low. In a 2006 survey of U.S. Corn Belt farmers, only 11% had used cover c rops in the previous five years (Singer et al., 2007) . Farmers have cited economic factors including time and monetary costs of planting and managing cover crops as key barriers to cover crop use (CTIC, 2015). There is a need for more data regarding cover crop use and ways to minimize the risk of cash crop yield reduction (Carlson and Stockwell, 2013; Kaspar et al., 2007) . The most commonly - grown cover crop species in the U.S. belong to three plant families: the Poaceae (grasses), Fabaceae (legumes), or Br assicaceae (mustards and oilseed radishes). 75 With regard to nomenclature, the oilseed radishes grown as cover crops are closely related and are usually either oilseed radish ( Raphanus sativus L. var. oleiformis Pers.) or forage radish ( R. sativus L. var. lo ngipinnatus L.H.Bailey) (Weil et al., 2009) . Said cover crop will be referred to as oilseed radish throughout this dissertation. Grass cover crops include cereal rye, annual ryegrass, and oats. Compared with a weedy control, cereal rye and annual ryegrass have been found to increase soil organic carbon 200 - 400% more than Austrian winter pea and hairy vetch (Kuo et al., 1997) . Kaspar et al. (2001) determined that cereal rye decreased inter - rill erosion up to 62%, while Ruffo et al. (2004) found cereal rye treatments to have 11% more ground cover residue than treatments that did not include cereal rye. Cereal rye has been found to take up residual soil N, decreasing the amount of N lost from corn/soybean [ Glycine max (L.) Merr.] cropping systems (Kaspar et al., 2007; Ruffo et al., 2004) . Due to ease of establishment and smaller seed costs (Table 3.1), grass species are commonly grown as cover crops. However, there are potential risks involved with grass cover crops. Annual ryegrass and cereal rye can cause delayed cash crop planting due to the difficulty of spring cover crop termination under adverse weather or soil conditions (Creamer e t al., 1997; Madden et al. 2004). Annual ryegrass in the U.S. has been documented to be resistant to several herbicide classes (Martins et al., 2012; Perez - Jones et al., 2005), which could make cover crop termination difficult in no - till systems . The relat ively large C:N ratio and biomass production of cereal rye increases the likelihood of N immobilization in the soil after cover crop termination (Crandall et al., 2005) . In the CTIC survey (2015), 57% of respondents reported the use of legume cover crops. Nitrogen fixed in the root nodules of legumes such as hairy vetch can be of benefit to cash crops directly as a fertilizer supplement (Ruffo et al., 2004; Sainju et al., 2001) and indirectly by 76 contributing to the formation of soil organic matter when in the presence of a C source (Villamil et al., 2006) . Similar to grasses, legumes also entail costs and risks. Hairy vetch and winter pea are more costly to plant, per - hectare, than annual ryegrass, cereal rye, oats, and oilseed radish (Table 3.1). The smal l size of crimson clover seed can make planting and stand establishment difficult. Winter pea does not always survive the winter in cold climates such as that of Michigan. Legume cover crops may serve as hosts to plant - parasitic nematodes and pathogens tha t infect cash crops (Dabney et al., 2001). As with cereal rye, the large biomass production of hairy vetch can make it difficult to terminate in the spring. Parr et al. (2011) documented the negative impact of unsuccessfully terminated hairy vetch on corn yield, as a result of inter - specific competition. Hairy vetch seed is a contaminant in harvested small grains and thus should not be grown in rotations that include wheat ( Triticum aestivum L.) or barley ( Hordeum vulgare L.) (Clark, 2008). Due to the poten tial benefits (e.g., decreased erosion), Brassicaceae species such as oilseed radish ( Raphanus sativus L. var. oleiformis Pers.) are also commonly grown as cover crops. Oilseed radish accumulates large quantities of biomass and provides dense ground cover (Lawley et al., 2011) , scavenges residual soil N (Weil and Dean, 2009) , may improve soil physical properties (Lehrsch and Gallian, 2010; Williams and Weil, 2004) , and can be used to help manage weeds and other pests (Lawley et al., 2011; Melakeberhan et al ., 2008; Stivers - Young, 1998; Wang et al., 2008) . Brassicaceae cover crops involve many of the same costs and risks as grass and legume species. Mustards and oilseed radish are susceptible to the same pests and pathogens as Brassicaceae cash crops and shou ld not be grown in close temporal proximity in a rotation (Snapp et al., 2006). The benefits of Brassicaceae cover crops, such as suppression of soil - borne pathogens, are sometimes overstated or oversimplified. For example, mustards and 77 oilseed radish are frequently noted for their bio - suppressive characteristics. Bensen et al. (2009), however, found no decrease in lettuce ( Lactuca sativa L.) disease with long - term mustard ( B. juncea and S. alba ) use, in spite of evidence of decreased lettuce drop incidence in the short - term. Brassicaceae species, along with other cover crops, have been documented to negatively impact cash crop stand (Haramoto and Gallandt, 2005). While Brassicaceae species such as oilseed radish are frequently planted in the fall to scaveng e residual soil N, there are concerns as to the fate of the accumulated N. Oilseed radish winter - kills and has a relatively small C:N ratio. After oilseed radish termination or winter kill, there is a window prior to cash crop planting during which N relea sed by the cover crop biomass could be lost from the system leaching or denitrification (Radersma and Smit, 2011). Cover crop impact on cash crop yield depends on a variety of factors including the specific cover crop and cash crop grown, cover and cash cr op management, soil fertility, and the weather. Interestingly, the literature reveals that when more than one type of impact of a given cover crop on cash crop yield is observed, the impacts range either from positive to neutral or neutral to negative. Cov er crop impact can vary across years. Even for a single specified cash crop such as corn, the literature is mixed as to the effect of a cover crop on grain yield. Some studies have shown that a preceding annual ryegrass, cereal rye, oats, oilseed radish, c rimson clover, hairy vetch, or winter pea cover crop generally had no impact on corn grain yield (Crandall et al., 2005; Isse et al., 1999; Kuo and Jellum, 2002; Parr et al., 2011; Vyn et al., 2000; Yenish et al., 1996) . However, some studies have found co ver crops to positively impact corn yield in at least some site - years (Dapaah and Vyn, 1998; Decker et al., 1994; Kuo and Jellum, 2002; Parr et al., 2011; Vyn et al., 1999) . Legume cover crops such as hairy vetch can contribute N to cropping systems for us e by the following cash crops (Miguez 78 and Bollero, 2005; Tonitto et al., 2006) . Cover crops with extensive root systems such as oilseed radish and cereal rye can create macropores in the soil which help cash crop roots access water (Williams and Weil, 2004) . Cover crops that have been terminated can conserve soil moisture by acting as a mulch (Unger and Vigil, 1998) . Cover crops also have the potential to decrease cash crop yield. Grass species such as annual ryegrass and cereal rye most often are obser ved to have negative or neutral impacts on corn yield (Dapaah and Vyn, 1998; Kaspar et al., 2007; Parr et al., 2011; Vyn et al., 1999) . Farmers have cited the risk of delayed cash crop planting as one barrier to cover crop adoption (CTIC, 2015). Delayed pl anting often correlates to decreased yields in agronomic systems. Cover crops can deplete limited soil moisture resources in dryland regions or drought years (Meisinger et al., 1991; Ruffo et al., 2004) . Cereal rye and other cover crops with large C:N rati os can immobilize soil N after cover crop termination (De Bruin et al., 2005) . Cover crops with hard seed, that produce viable seed before termination, or are not completely terminated prior to cash crop planting may act as weeds that compete with the cash crop for resources such as sunlight and N. Williams and Weil (2004) attributed decreased soybean yield to competition between soybeans and a poorly controlled canola ( Brassica napus L.) cover crop. Cover crops sects or pathogens whose populations would otherwise subside during a fallow period (Odhiambo et al., 2012) . From 2014 to 2015, the percentage of respondents who planted cover crop mixtures increased from 60% to 67% (CTIC, 2015). Research has been done on a variety of cover crop bicultures, often consisting of a legume combined with a grass. The results of different studies involving mixtures are difficult to compare since planting rates and proportions of component species are a confounding factor (Miyaza wa et al., 2014) . Cover crops grown in mixtures can 79 benefit each other by occupying complementary niches (Hauggaard - Nielsen et al., 2001; Teasdale and Abdul - Baki, 1998) . Kuo and Jellum (2002) suggested that upright grass cover crops provided vining hairy v etch with a support structure on which to grow and better access light. Cover crop mixtures can be more resilient and stable than pure stands when subjected to extreme weather events such as drought (Rusinamhodzi et al., 2012; Wortman et al., 2012) . Mixtur es may allow producers to take advantage of the complementary benefits provided by cover crops while mitigating potential risks (e.g., N immobilization during periods of high cash crop N demand) (Miyazawa et al., 2014; Smith et al., 2014; Teasdale and Abdu l - Baki, 1998) . For example, cereal rye provides fast ground cover and produces large amounts of biomass, protecting against erosion and building soil organic matter through the addition of C. The large C:N ratio of cereal rye, however, can lead to N immobi lization after cover crop termination (Clark et al., 1994) . Conversely, nodulating cover crops such as hairy vetch fix N that can offset the C produced by large C:N ratio species and contribute to cash crop N needs, but are slow to establish and provide gr ound cover. Combining a cover crop with a large C:N ratio such as avoiding the difficulties inherent in too high a ratio (i.e. N immobilization) and too small a ratio (i.e. early N release from decomposing cover crop tissue and subsequent leaching) (Odhiambo and Bomke, 2001; Rosecrance et al., 2000) . Clark et al. (1994) found that in cereal rye + hairy vetch bicultures, cereal rye scavenged residual soil N while hairy vetch fixed N, which increased corn grain yield an average of 14% compared with a no - cover crop control. The theoretical benefits provided by mixtures are not always realized in field studies. Species selection, and planting rates and proportions dictate to what extent potential benefits are realized. In Nebraska, Wortman et al. (2012) found mustard species t o be more competitive and 80 productive than legumes when grown in mixtures. In Ohio, Creamer et al. (1997) evaluated 13 multi - species mixtures and observed that some species were not suitable due to lack of winter hardiness or small biomass production. A dom inant component species in a mixture can also lessen the benefits provided by the other cover crops. Brainard et al. (2011) found that soybeans grown in a mixture with sorghum - sudangrass produced less biomass, nodulated less, and fixed less N than soybeans grown in monoculture in both years of a study. Furthermore, the sorghum - sudangrass monoculture treatment had a smaller seed cost than the mixture. Depending on the species and planting rates, seed mixtures can be more costly than monocultures (Table 3.1). Mixtures with a range of seed sizes may be difficult to plant, establish, and manage (Creamer et al., 1997). When sowed later in the fall (mid - to late September), hairy vetch planted in mixture with cereal rye produced less biomass than when planted in l ate August in Michigan (Hayden et al., 2015). Farmers may also have difficulty predicting the effect of a cover crop mixture on a cropping system. Hayden et al. (2014) did not find evidence of synergy between hairy vetch and cereal rye in mixture. The aut hors noted that altering the seeding proportions of the two species in mixture resulted in tradeoffs among the agroecosystem benefits provided by the cover crops. For example, larger proportions of hairy vetch lead to greater seed costs and less weed suppr ession but greater amounts of fixed N. Despite farmer interest, there has been relatively little research on cover crop mixtures (Carlson and Stockwell, 2013) . More research is needed to determine and which mixtures provide the best combination of benefit s and the least likelihood of risk to a following cash crop. In this study, we chose to evaluate cover crop bicultures rather than polycultures because the planting and management of polycultures can be impractical (Creamer et al., 1997) . Research also sug gests that polycultures are not more beneficial to a cropping system than bicultures 81 (Miyazawa et al., 2014; Smith et al., 2014; Teasdale and Abdul - Baki, 1998) . D ue to a lack of information on bicultures in literature, the current study focuses upon bicult ures of a Brassicaceae cover crop species ( specifically, oilseed radish) combined with either a grass or legume. Annual ryegrass, cereal rye, oats, crimson clover, hairy vetch, and winter pea were chosen as complementary species to provide a range of chara cteristics such as large vs. small C:N ratio, winter - hardy vs. frost sensitive, and N - fixers vs. non - N fixers. The objectives of this study were to: a) quantify biomass production of annual ryegrass, cereal rye, oats, crimson clover, hairy vetch, winter p ea, and oilseed radish both individually and in bicultures, b) determine the effect of the cover crops on fall and spring weed biomass, and c) evaluate cover crop impact on corn grain yield in a conventionally tilled system with no applied fertilizer. We h ypothesized that: a) bicultures will produce more biomass than monocultures, b) legume species will be less competitive in biculture with oilseed radish than grass species, c) cover crop treatments with the largest biomass production will be more effective at weed suppression than those with smaller biomass, d) the inclusion of oilseed radish in biculture with each grass will mitigate the negative impact of the grass on corn yield, and e) the inclusion of oilseed radish in biculture with each legume will re duce the positive impact of the legume on corn yield when compared with the legume monocultures. T o our knowledge this study is unique in that it investigated the performance of a variety of legumes and grasses included in biculture with oilseed radish. MATERIALS AND METHODS Site characteristics and field operations 82 A field experiment was conducted at the Michigan State University Agronomy Farm in Lansing, MI (42° 46' N, 84° 34' W) for three years. In 2011 - 2012 the soil was a Riddles - Hillsdale sandy loam (fine - loamy, mixed, mesic Typic Hapludalfs). In 2012 - 2013 the so ils were a Riddles - Hillsdale sandy loam and a Metea loamy sand (loamy, mixed, mesic Arenic Hapludalfs). In 2013 - 2014 the soil was a Capac loam (fine - loamy, mixed, mesic Aeric Ochraqualfs). Field characteristics are listed in Table 3.2. The experiment had r andomized complete block design with three (2012 - 2013) or four (2011 - 2012 and 2013 - 2014) replications. All plots were 3 x 12 m in size. Cover crop treatments were planted using a Great Plains no - till drill or a John Deere planter in 19 - cm wide rows into w heat stubble between mid - August and early - September (Table 3.2). Prior to planting, weeds were controlled with applications of glyphosate ( N - (phosphonomethyl)glycine) at 1.22 kg ae ha - 1 tank - mixed with 2% w w - 1 ammonium sulfate (AMS). Cover crop treatments (Ampac Seed Co., Tangent, OR) ( 11 kg ha - 1 ), variety not stated (VNS) crimson clover ( 11 kg ha - 1 ), VNS hairy vetch (34 kg ha - 1 ), VNS winter pea (67 kg ha - 1 ha - 1 ha - 1 ), and VNS oats (72 kg ha - 1 ). Biculture mixtures consisted 11 kg ha - 1 ) + VNS crimson clover ( 6 kg ha - 1 ) oilseed radish ( 11 kg ha - 1 ) + VNS hairy vetch ( 17 kg ha - 1 ) radish ( 11 kg ha - 1 ) + VNS winter pea ( 34 kg ha - 1 ) 11 kg ha - 1 ) ryegrass ( 22 kg ha - 1 ) 11 kg ha - 1 ) 63 kg ha - 1 ) 11 kg ha - 1 ) + VNS oats ( 36 kg ha - 1 ) . Additionally, a no - cover crop ( weedy) control and a bare ground control that was hand - weeded once in fall 2011 and sprayed with glyphosate ( N - (phosphonomethyl)glycine) once per fall in 2012 and 2013 were 83 included. Seeding rates were base d on Michigan State University Extension and seed distributor recommendations as outlined in the Midwest Cover Crops Council online decision tool species fact sheets (MCCC, 2015). To ensure nodulation for N fixation, Rhizobium leguminosarum bv. trifolii in oculant was mixed by hand with the crimson clover seed prior to planting, while Rhizobium leguminosarum bv. viciae inoculant was applied to the hairy vetch and winter pea seed. Cover crops which survived the winter were terminated the following spring wit h glyphosate ( N - (phosphonomethyl)glycine) + 2,4 - D ( 2,4 - dichlorophenoxyacetic acid) 8 - 10 d prior to corn planting . The fields were prepared for planting using a chisel plow and a minimum of two passes of a soil finisher. The 102 - day field corn cultivar - Louis, MO) was planted at a rate of 74,100 seeds per ha - 1 using a four - row planter with a row width of 76 cm. Weeds were controlled as necessary with glyphosate at corn stages V6 and V10. In order to avoid concealing the effec ts of cover crops, no irrigation or fertilizer was applied to either the cover crops or corn. Yearly weather data for Lansing, MI were obtained from the Michigan State University Enviro - weather website for the MSU - Hort station (MSUEW, 2015), with the excep tion of winter precipitation data which were collected from the Lansing, MI weather station via the NOAA website (NOAA, 2015). Temperature and precipitation 30 - yr monthly means (1981 - 2010) for Lansing, MI were obtained from the National Oceanic and Atmosph eric Administration (NOAA) website (NOAA, 2015). Weather conditions over the course of the experiment are presented in Figure 3.1. Cover crop, weed, and corn data collection Cover crop, weed, and volunteer wheat aboveground biomass were harvested from two randomly placed 0.25 m 2 quadrats per plot, with the exception of fall 2011 when one subsample 84 per plot was collected. Biomass was harvested in the fall prior to winter kill and in the spring prior to cover crop termination (Table 3.2). Oilseed radish root s were also pulled by hand from each quadrat in the fall. Tissue samples were separated into cover crop and weed material, dried at 70° C for 5 - 10 d, and weighed. Corn stand was assessed 2 - 4 wk after planting by counting the number of plants in 5.3 meter - l engths of each of two rows. Corn stand was again assessed prior to corn harvest. A plot combine was used to harvest the middle two rows of each plot. Corn yield was adjusted to 15.5% moisture. Biomass relative yield calculation Biomass relative yield (RY) was calculated using fall cover crop biomass. The RY index buffers the influence of population density on biomass and thus helps account for different planting rates between treatments (Bedoussac and Justes, 2011; Williams and McCarthy, 2001), allowing for a more distinct evaluation. Relative yield is calculated as RY A = Y AB /(P A * Y A ) where RY A was the relative biomass yield of species A, Y AB was the dry biomass yield of species A when grown with species B, P A was the proportional seeding rate of species A that is seeded in the biculture, and Y A was the dry biomass yield of species A when grown in monoculture [adapted from Bedoussac and Justes ( 2011 ) , who used the notation of Fowler ( 1982 ) ]. In this study, the oilseed radish seeding rate was held constant across both monoculture and b iculture treatments; P A was thus 1. The complementary species to oilseed radish in each biculture treatment was drilled at half the monoculture planting rate; P B for these species was thus 0.5. Statistical analysis 85 All data sets were analyzed with SAS Enterprise Guide v. 6.1 (SAS Institute Inc., Cary, NC). Analysis of variance was completed using PROC MIXED, the LSMEANS and PDIFF statements, and the Tukey - < 0.05). Dat a were tested for normality, equal variances, and treatment*year interactions prior to ANOVA and mean separation. Replication was deemed a random effect. Mean separation was conducted with the PDMIX 800 macro (Saxton, 1998) . Because some species failed to survive the winter in some years, some species did not establish well, and some species did not overwinter well, maximum annual aboveground biomass was calculated using a combination of fall and spring data. For example, in the annual ryegrass treatment fa ll biomass values were deemed the maximum annual biomass if spring biomass production was smaller than fall due to poor winter survival. If spring values were larger than fall, then the spring biomass was used. For bicultures, oilseed radish biomass was ad ded to the greater of the fall and spring biomass of the complementary species in each mixture. Corn yield data were square - root transformed to improve normality. RESULTS AND DISCUSSION Cover crop biomass production Fall cover crop biomass. There was a significant treatment*year interaction (P < 0.0001), so fall cover crop total aboveground dry biomass is presented by year (Figure 3.2). The oilseed radish root: shoot ratio treatment*year interaction was not significant (P = 0.21), so data wer e combined across years (Table F.2). Oats produced the most biomass in the fall of 2011 and 2012, while oilseed radish produced the most biomass in 2013 (Figure 3.2). Crimson clover produced the least amount of biomass in all three years. Oats produced app roximately 2.1 and 9.1 times more biomass than crimson clover in 2011 and 2012, respectively. Oilseed radish produced 11 times more biomass than crimson clover in 2013. Treatments that included oats consistently produced 86 some of the largest biomass across years. The large fall aboveground biomass production may have been the result of oats partitioning less biomass to the roots than annual ryegrass and cereal rye, which are more winter - hardy. The small fall biomass production of crimson clover was indicativ e of poor stand as a result of establishment issues. Crimson clover had the smallest - sized seed among the cover crops tested, which caused difficulties despite the use of a planter set up to handle small seeds. In addition, the crimson clover monoculture w as planted at 11 kg ha - 1 in this study, which is a small seeding rate even when using a drill (MCCC, 2015). November 2011 was 2.8 and 3.9 °C warmer than November 2012 and 2013, respectively (Figure 3.1). Biomass was collected on 14 November 2011 (Table 3.2 ). The larger amount of crimson clover biomass collected in 2011 may have been due to the warmer temperatures. In September 2013 precipitation totaled 23% of the 30 - yr average for the month (Figure 3.1). Dry conditions immediately following the 2 September cover crop planting (Table 3.2) may have delayed germination slowing cover crop growth and contributing to the decreased cover crop biomass observed in 2013 (Figure 3.2). In 2013, biomass production was smaller than in the other two years at least partly because biomass was collected 10 - 11 weeks after planting (WAP) in 2011 and 2012, but only eight WAP in 2013. While differences were observed in cover crop aboveground biomass, no differences were detected in oilseed radish root: shoot ratios (Appendix Tabl e 3.7 ). Differences may not have been detected because they did not exist (i.e. oilseed radish was highly competitive in biculture), or because root sampling protocols were inadequate. Variability in cover crop biomass across years and studies is common i n the literature. The fall biomass production found in this study is similar to that found in other studies for annual ryegrass (Dapaah and Vyn, 1998; Francis et al., 1998; Isse et al., 1999; Thorup - Kristensen, 2001; 87 Vyn et al., 1999), cereal rye (Axelsen and Kristensen, 2000; Kaspar et al., 2001; Ranells and Wagger, 1997; Thorup - Kristensen, 2001; Vyn et al., 2000) , and oats (Brennan and Smith, 2009; Francis et al., 1998; Kaspar et al., 2001; Stivers - Young, 1998; Thorup - Kristensen, 2001; Vyn et al., 2000) . Fall legume biomass production values are less commonly reported in the literature, but crimson clover has been noted to produce 90 - 1040 kg ha - 1 (Axelsen and Kristensen, 2000; Isse et al., 1999; Ranells and Wagger, 1996) , while Axelsen and Kristensen (2000 ) found hairy vetch to produce 1910 kg ha - 1 of dry aboveground biomass. Oilseed radish biomass ranges from 176 4840 kg ha - 1 (Axelsen and Kristensen, 2000; Dapaah and Vyn, 1998; Dean and Weil, 2009; Isse et al., 1999; Stivers - Young, 1998; Thorup - Kristensen, 2001; Vyn et al., 2000; Wortman et al., 2012) . Within each year, the cover crop biculture treatments generally produced similar amounts of fall biomass to each other (Figure 3.2). In all mixtures, oilseed radish biomass exceeded that of the second specie s in the biculture, accounting for 54 - 99% of total above ground biomass. Across the three years, oilseed radish dominated the oilseed radish + legume bicultures to a greater extent than the oilseed radish + grass bicultures (Figure 3.2). While there has b een no comprehensive evaluation of oilseed radish in biculture with grass and legume species, there are some reports of oilseed radish biculture performance in the literature. Cavadini (2013) found that fall - planted oilseed radish, oilseed radish + cereal rye, and oilseed radish + oats generally produced the same amount of dry aboveground biomass in the fall in each site - year. Oilseed radish was planted at a smaller rate in biculture than monoculture in that study. In one site - year, the oilseed radish in th e oilseed radish + cereal rye biculture composed 70% of the dry aboveground production of that mixture even though both crops were planted at half the rate in biculture that they were in monoculture (White and Weil, 2010) . In a 88 study in which seeding rates were the same in biculture for each species as in monoculture, Moller and Reents (2009) found that at one site there were no significant differences in cover crop dry aboveground biomass production of oilseed radish + pea ( Pisum sativum L.), oilseed radis h + common vetch ( Vicia sativa L.), and oilseed radish, pea, and common vetch monocultures while at the other site, oilseed radish + pea yielded more biomass than the oilseed radish monoculture. Spring cover crop biomass. Total spring cover crop biomass could not be combined across years due to a treatment*year interaction (P < 0.0001) caused by varying weather conditions across the three years of this study that impacted cover crop winter survival. February and March 2013 were 1.3 and 2.6 °C colder than the 30 - yr averages for those months, respectively (Figure 3.1). January, February, and March 2014 were 5.4, 3.0, and 5.6 °C colder than the corresponding 30 - yr monthly averages. The winter killed cover crops of oilseed radish and oats did not contribute to cover crop biomass in the spring. In spring 2012, crimson clover, hairy vetch, and annual ryegrass produced 1.5 - 3.9 times more biomass than the bicultures of oilseed radish + crimson clover, + hairy vetch, + winter pea, + cereal rye, + annual ryegrass, an d + oats (Figure 3.3). With the exception of cereal rye and oilseed radish + cereal rye, all of the other monoculture treatments produced 1.9 - 3.9 times more biomass than the corresponding biculture treatments. Monoculture treatments produced more biomass t han the bicultures partly because the seeding rate in biculture was half that in monoculture and partly because the oilseed radish did not overwinter. In spring 2013, cereal rye and oilseed radish + cereal rye respectively produced a minimum of 855 and 703 kg ha - 1 more biomass than all other non - oat, non - oilseed radish + oat, and non - radish monoculture treatments, except annual ryegrass. Unlike in 2012, in 2013 biomass production was the same within each monoculture/biculture treatment pair. In 2014, cereal rye 89 produced at least 1.3 times more biomass than all of the other cover crops that survived the winter. Hairy vetch, annual ryegrass, and cereal rye produced more biomass than the corresponding biculture treatments. The weather impacted spring biomass p roduction. Winter pea was not difficult to establish (data not shown) but did not survive the winter in two of the three years. March 2013 was 2.6°C colder than the 30 - yr average (Figure 3.1). November 2013 March of 2014 was an average of 4.3 °C colder t han the 30 - yr average. The cold winter conditions in the second and third year of this study may help to explain why winter pea overwintered the first year of the study but not the other two years (Figure 3.3). In addition, in 2012 - 2013 and 2013 - 2014 winte r pea was observed to have reached the reproductive stage prior to frost (data not shown), which would have further decreased the likelihood of winter survival. Weather clearly affected the spring biomass of the other cover crops, as well. For example, hai ry vetch and annual ryegrass produced up to an order of magnitude more biomass in spring 2012 compared with 2013 and 2014 (Figure 3.3), likely because November - March was an average of 5.5 °C warmer in 2011 - 2012 than winter of other two years. Our results show several cases where cover crops appear to have been suppressed by oilseed radish, such as crimson clover vs. oilseed radish + crimson clover in 2012, and crimson clover and hairy vetch vs. oilseed radish + crimson clover and oilseed radish + hairy vet ch in 2014. Moreover, it cannot be determined from our study whether mixtures per se or differences in seeding rates account for differences in the biomass of the complementary species in mixture. For example, had cereal rye been sown in monoculture at ha lf of the rate used in this study, reductions in intraspecific competition may have resulted in similar biomass as those observed at 90 the full monoculture rate, as has been documented in the literature (Clark et al., 1994; Hayden et al., 2014). As with fa ll cover crop biomass, there is a wide range of values reported for spring dry aboveground biomass production of fall - planted cover crops. Spring biomass values found in this study were in agreement with those found in the literature, except in the cases o f crimson clover and winter pea. Values reported in the literature for crimson clover and winter pea range from 968 - 8000 kg ha - 1 and 1800 - 7600 kg ha - 1 , respectively, and are greater than those found in this study (Bauer et al., 1993; Daniel et al., 1999; D ecker et al., 1994; Isse et al., 1999; Norsworthy et al., 2010; Odhiambo and Bomke, 2001; Parr et al., 2011; Ranells and Wagger, 1996; Ranells and Wagger, 1997; Singogo et al., 1996; Yenish et al., 1996) . As previously discussed, crimson clover did not est ablish well in two of the three years of the study, while winter pea failed to survive the winter in two of the three years of the study. To improve the probability of winter pea survival over the winter, it has been suggested that the seed be planted rela tively deep (5 cm) and not too early in the fall. Maximum annual biomass. Fall and spring aboveground dry biomass data were used to calculate the total maximum likely amount of biomass contributed by each cover crop treatment from fall through spring, rem oving the confounding effects caused by some species that failed to establish or overwinter well. Due to a treatment*year interaction (p < 0.0001), data were not combined across year. In 2011 - 2012, no differences between treatments were detected (Table 3.3 ). In 2012 - 2013, oilseed radish + crimson clover produced 3.2 times more biomass than crimson clover and oilseed radish + hairy vetch yielded 5.7 times more biomass than hairy vetch. In 2013 - 2014, oilseed radish + crimson clover, oilseed radish + winter pe a, and oilseed radish + annual ryegrass 91 respectively produced 4.6, 1.9, and 1.8 times more biomass than crimson clover, winter pea, and annual ryegrass. The biomass results in this study are of interest because the presence and quantity of cover crop biom ass is one of the mechanisms through which cover crops influence production systems. In some cases, this influence comes in the form of harm, such as when cereal rye biomass with a large C:N ratio immobilizes soil N (Crandall et al., 2005; Odhiambko and Bo mke, 2001). In other cases, cover crop biomass can convey benefits to productions systems (Kuo et al., 1997). Cover crops decrease erosion by buffering the ground from the impact of raindrops and slowing the flow of water over the ground (Dabney et al., 20 01) . The rapid production of large amounts of biomass helps cover crops suppress weeds (O'Reilly et al., 2011) . Non - legume cover crops scavenge residual soil nutrients, while legume cover crops fix atmospheric N (Dabney et al., 2001) . In order to estimate the impact of cover crops on nutrient cycling, knowledge of cover crop potential biomass production is necessary. Cover crop impact on weeds Fall weed biomass . Due to a treatment*year interaction (P < 0.0001), total fall weed dry aboveground biomass data Total weed densities were so small in fall 2011 that no weed data were collected. Total weed bio mass was over an order of magnitude larger in fall 2012 than in 2013 (Figure 3.2 and Table 3.4). In both 2012 and 2013, the predominant weed species were common chickweed ( Stellaria media (L.) Vill.) and Lamium spp. (data not shown). In fall 2012, there wa s 155 - 571 kg ha - 1 weed biomass in treatments with the least amount of weed biomass, less than the 1268 - 1586 kg ha - 1 biomass in the weedy control, crimson clover, hairy vetch, and winter pea treatments. In 92 2013, there was an order of magnitude less total we ed biomass in the bicultures of oilseed radish + crimson clover, + hairy vetch, + winter pea, + annual ryegrass, and + cereal rye, and the annual ryegrass and cereal rye monocultures than in the weedy control. When bicultures were compared with monoculture s, total weed biomass was smaller in the oilseed radish + crimson clover, oilseed radish + hairy vetch, and oilseed radish + winter pea treatments than in the corresponding legume monocultures and weedy control in fall 2012. The oilseed radish in these bic ulture treatments appeared to contribute to weed suppression. No differences in total weed biomass were detected among the annual ryegrass, cereal rye, oats, oilseed radish + annual ryegrass, oilseed radish + cereal rye, and oilseed radish + oats treatment s. All were equally effective at suppressing weeds when compared with the weedy control. Irrespective of oilseed radish presence, annual ryegrass and cereal rye suppressed weed biomass. In fall 2013 the biculture treatments were no more effective at suppre ssing weeds than the corresponding monocultures, with the exception of oilseed radish + hairy vetch and hairy vetch. Cover crops can suppress weeds (Conklin et al., 2002; Creamer et al., 1996; Stivers - Young, 1998; Hayden et al. 2012) . This ability is of i nterest because winter annual weeds can harbor pests and diseases of cash crops (Groves et al., 2001; Venkatesh et al., 2009) , they can grow too large to be easily controlled, and if they are left uncontrolled they may add to the weed seedbank and become p roblematic in future years of the rotation. In particular, the ability of cereal rye (Akemo et al., 2000; O'Reilly et al., 2011) and oilseed radish (Lawley et al., 2011; O'Reilly et al., 2011; Stivers - Young, 1998) to suppress weeds has been documented in t he literature. Since large cover crop biomass production has been linked to effective weed suppression (Teasdale, 1996) , it is not surprising that the treatments in this study which provided the most fall weed suppression also had some of the largest cover crop biomass (Figure 3.2). 93 There an average of at least 453 kg ha - 1 more total weed biomass across years in the crimson clover, hairy vetch, and winter pea treatments than in the annual ryegrass, cereal rye, oats, and oilseed radish treatments. The inclus ion of oilseed radish in biculture with the legumes greatly increased fall weed suppression in 2012, probably because of the large biomass contributed by the oilseed radish in the biculture treatments (Figure 3.2). Other studies have also found that cover crop mixtures controlled weeds better than legume monocultures. Bicultures of cereal rye + crimson clover and cereal rye + hairy vetch were more effective at suppressing weed emergence and biomass production than crimson clover and hairy vetch monocultures (Teasdale and Abdul - Baki, 1998) . Field pea ( Pisum sativum L.) was less effective at weed suppression than a barley + pea biculture evaluated by Hauggard - Nielsen et al. (2001) . Akemo et al. (2000) likewise found cereal rye and cereal rye + field pea to mor e effective at weed suppression than field pea grown in monoculture. Spring weed biomass . Due to a treatment*year interaction (P < 0.0001), total spring dry aboveground total weed biomass data could not be combined across years. Spring total weed biomass did not vary in magnitude across years to the extent that it did in the fall (Figure 3.3 and Table 3.4). In spring 2012 the dominant weed species were wheat, common chickweed and common speedwell ( Veronica arvensis L.) (data not shown). Total weed biomass was composed of 95% wheat (Figure 3.3). All of the cover crop treatments had at least 643 kg ha - 1 less total weed biomass than the weedy control; all treatments had significantly less total weed biomass than the weedy control (Table 3.4). In spring 2013, w ild violet ( Viola - purse [ Capsella bursa - pastoris (L.) Medik.], dandelion ( Taraxacum officinale F.H. Wigg.), and Lamium spp. were the major weed species. No differences in total weed biomass were detected. In spring 2014 common chickweed, Lamium spp., and common lambsquarters ( Chenopodium 94 album L.) dominated in the field. In spring 2014, there was 596 kg ha - 1 biomass in the weedy control, more than the 0 - 179 kg ha - 1 biomass in all other treatments except oilseed radish (214 kg ha - 1 ), crimso n clover (565 kg ha - 1 ), and winter pea (663 kg ha - 1 ). Differences in weed control across season and years were partly the result of varying cover crop biomass. Since different weeds react differently to cover crop residue (Teasdale, 1996) , differences betw een years could also have been due to the varying weed composition in each field - year. The presence of volunteer wheat may also have played a role. Wheat itself is frequently used as a fall cover crop and is known to be a vigorous and hardy species. Biomas s relative yields (RY) Fall biomass RYs were calculated for oilseed radish, each complementary species treatment*year interaction for any permutation of the biomass RYs (P = 0.75, 0.32, and 0.95, respectively). Since there were no treatment*year interactions, we averaged the oilseed radish and species B biomass RYs across the three years of this study and plotted them on the graph created by Williams and McCarthy (2001) (Fig ure 3.4). Oilseed radish biomass RY is on the y - axis and species B biomass RY is on the x - axis. All of the treatment points fell into the area bounded by the vertical line x = 0, the horizontal line y = 1, and the diagonal line biomass RY oilseed radish = b iomass RY species B . According to Williams and McCarthy (2001) , this positioning on the graph indicates that oilseed radish likely competed/interfered with the complementary species in each mixture. While the coordinates for the treatments all fell into the same space on the graph, those of oilseed radish + cereal rye and oilseed radish + oats fell closest to the line biomass RY oilseed radish = biomass RY species B . In these two treatments, the effects of the cover crops 95 overwintered in all three years (hairy vetch) had coordinates nearest the horizontal line above which oilseed radish would be said to suppress the complementary species (Figure 3.4). It is harder to draw con clusions about crimson clover and winter pea as those legumes were clearly heavily influenced by weather conditions and establishment issues as previously discussed. Another potentially confounding factor in this study was weeds, which were not controlled in treatment plots after cover crop planting. The weeds biomass data in Figure 3.2, however, indicate that oilseed radish was likely competing/interfering with weeds as well as the complementary species. Fall weed biomass production was smaller in the leg ume and oats bicultures than in the corresponding monoculture treatments. Since weed biomass was greater in monoculture treatments, the weeds were likely competing with each complementary cover crop and thus potentially decreasing monoculture treatment bio mass. Since the RY equation calls for this biomass number to be multiplied by the biculture planting rate (0.5) and then used as the denominator to determine biculture species RY, our biomass RY calculations were likely underestimates. Cover crop impact o n corn Corn grain yield. Corn grain yield data were not combined. In 2012, corn in the annual ryegrass treatment produced 27 - 48% less grain than that in all other treatments except cereal rye and oilseed radish + cereal rye (Table 3.5). In 2013 there were no differences in corn yi eld. In 2014, corn in the cereal rye treatment yielded less than that in all other treatments except oilseed radish + cereal rye and winter pea. A number of factors may have contributed to our inability to detect differences including variability in the da ta, soil type and fertility differences between field sites, and varying weather each year. The most differences were detected in 2012, which was a drought year (Figure 3.1). Cash crop yield possibly benefits most from cover crop influences in 96 years with a dverse weather conditions or sites with adverse soil conditions. In 2014, the field tested at 3.8% soil organic matter. Soil N reserves may have obscured differences between controls and cover crop treatments. Weeds in the weedy control could have acted si milarly to cover crops, also obscuring differences. The results of our study with regard to the neutral to positive impact of legume cover crop monocultures and bicultures on corn yield are generally in agreement with the literature. Our study as not desig ned to separate cover crop effects on N from the cover crop rotation effect, both of which could help explain our results. There have been other studies involving cover crops in which there was at least one no - fertilizer treatment. Most involved the use of legume cover crops. Legume cover crops have been found to increase crop yield. This effect is often attributed primarily to the N contributed to the system by the legumes (Kuo and Jellum, 2002; Torbert et al., 1996) . In a meta - analysis, Miguez and Bollero (2005) found winter legumes increased corn yield 37% in the absence of applied N fertili zer as compared to a bare ground control, while cover crop bicultures increased corn yield 21%. Smith et al. (2008) attributed the beneficial effect of a diverse crop rotation on corn yield in to the N contributed by legume cover crops; corn yield in the most diverse rotation was the same as the county average, despite the lack of conventional fertilizer. Contrary to our results, some studies have found grass cover crop s to have no impact, or a positive impact, on corn yield. Andraski and Bundy (2005) found that corn yield benefited from the use of an oat or cereal rye cover crop as compared to fallow, but attributed the impact to rotation effects rather than N contribut ion from the cover crops. In the Miguez and Bollero (2005) meta - analysis, grass winter cover crops had no impact on corn yield regardless of whether N fertilizer was applied. Kuo and Jellum (2002) and Isse (1999) 97 determined that annual ryegrass had no impa ct on corn yield. Kuo and Jellum (2002) , Crandall et al. (2005) , Yenish et al. (1996) , and Vyn et al. (2000) found the same of cereal rye. As in our study, other studies have also found annual ryegrass or cereal rye to negatively impact corn yield in at least some years (Hively and Cox, 2001; Johnson et al., 1998) . The negative effect of the grass cover crop treatments on corn in some years o f our study could be due to a number of factors, including corn stands up to 15% smaller in the annual ryegrass and cereal rye treat ments than in the bare ground control (Table F.4). Kaspar et al. (2007) postulated that a combination of large cereal rye bi omass production and insufficient time between cover crop termination and corn planting lead to the decreased yield seen in one year of that study. In this study in 2014, there likewise may not have been sufficient time between cover crop termination and c ash crop planting (Table 3.2) given the large amount of biomass produced by the cereal rye treatment (Figure 3.2 - 3.3). Other authors have noted the difficulty of terminating annual ryegrass (Creamer et al., 1996; Madden et al., 2004) . We had difficulty cre ating a uniform seedbed during tillage in annual ryegrass plots, which could have interfered with corn planting or germination and thus affected stand (Table 3.5). In some years annual ryegrass or cereal rye may have immobilized soil N to the detriment of the corn. In the unfertilized treatment of a study by Kuo and Jellum (2000) , corn N uptake was significantly smaller in annual ryegrass and cereal rye treatments than in a hairy vetch treatment, though not generally different than the weedy control. Other researchers have also suggested N immobilization as a mechanism by which cereal rye (Miguez and Bollero, 2006; Vaughan and Evanylo, 1999) and annual ryegrass (Vyn et al., 1999) decreased corn yields. Finally, in the two years of this study in which there w ere corn yield differences, there were also differences in corn tasseling in late July early August (Table F.6). Tasseling in the annual ryegrass and cereal rye treatments was delayed. This may have 98 resulted in poor pollination, or it could have decrease d the amount of time available for corn grain fill. All of these factors could have contributed to the smaller yields seen in the annual ryegrass and cereal rye treatments. CONCLUSIONS 99 There are several avenues of research leading from this work. It would be inte resting to repeat this experiment as a multifactorial plot design with the added factors of fall fertilizer application to the cover crops and weed control vs. no weed control in the cover crop plots. Cover crop effects on soil properties like bulk density could be measured. A quantification of the ecosystem services provided would be timely and relevant. A study using a substitutive seeding rate design could be performed to optimize the proportion of oilseed radish to other species in biculture mixtures. 100 APPENDICES 101 APPENDIX A CHAPTER 3 TABLES AND FIGURES 102 103 104 105 106 Figure 3.1. Monthly and 30 - yr mean air temperature and precipitation for Lansing, MI 2011 - 2014. Precipitation is represented on the left y - axis and by bars. Temperature is represented on the right y - axis and by lines. -15 -10 -5 0 5 10 15 20 25 30 0 50 100 150 200 250 Temperature ( C) Precipitation (mm) 2011-20122 2012-20132 2013-20142 30-yr mean precip 2011-2012 2012-2013 2013-2014 30-yr mean temp 107 Figure 3.2. Mean (±SE) dry aboveground fall bi omass of legume and grass cover crops in monoculture and in biculture with oilseed radish in Lansing, MI in 2011 - 2013. 0 1000 2000 3000 4000 5000 2011 0 1000 2000 3000 4000 5000 Dry aboveground biomass kg ha - 1 radish grass or legume weed v. wheat 2012 0 1000 2000 3000 4000 5000 2013 108 Figure 3.3. Mean (±SE) dry aboveground spring biomass of legume and grass cover crops in monoculture and in biculture with oi lseed radish, weeds, and volunteer wheat in Lansing, MI in 2012 - 2014. 0 1000 2000 3000 4000 5000 2012 0 1000 2000 3000 4000 5000 Dry aboveground biomass kg ha - 1 legume or grass weed v. wheat 2013 0 1000 2000 3000 4000 5000 2014 109 Figure 3.4. Fall relative yields of oilseed radish and complementary species (species B) in each cover crop biculture treatment. Bars represent one SE from the mean. To the left of the diagonal line RY oilseed radish =RY species B , oilseed radish had the competitive adv antage over species B while to the right of the diagonal line, the reverse was true. In the upper right quadrant, the species interaction was mutually beneficial. In the lower right quadrant, species B suppressed oilseed radish growth. The lower left quadr ant indicates competition or interference between the two species. In the upper left quadrant, oilseed radish suppressed the growth of species B. Figure adapted from Williams and McCarthy (2001) . 0.0 1.0 2.0 0.0 1.0 2.0 RY radish RY species B + crimson clover + hairy vetch + winter pea + annual ryegrass + cereal rye + oats RY radish =RY species B 110 AP PENDIX B CHAPTER 3 SOIL, WEED, AND CORN SUPPLEMENTARY DATA COLLECTION METHODS Soil sampling Pre - side dress nitrate - N testing (PSNT) was performed at corn stage V6 - V8 by collecting five soil cores per plot with a soil probe in an H - pattern to a depth of 15 - 20 cm. Soil samples were aggregated by treatment and ground to pass through a 2 mm sieve. In the 2012 and 2014, the samples were sent to the Michigan State University Soil and Plant Nutrient Laboratory for PSNT. In 2013 samples were extracted and analyzed at Michigan State University Kellogg Bi ological Station (Hickory Corners, MI) using a flow injector analyzer (NH 4 + - N via the diffusion colorimetry technique and NO 3 - N via cadmium reduction and colorimetry). PSNT data are presented in Table 4.3. After corn harvest, soil samples were collected t o determine residual soil nitrate levels. Five cores per plot were collected with a soil probe to a depth of 15 - 20 cm; soil was dried at 60 °C, ground, and passed through a 4 mm sieve. A KCl extraction was performed as per KBS021 (2015) . In 2012, samples w ere analyzed using cadmium reduction and colorimetry (KBS021, 2015) . In years 2013 and 2014, extracts were sent to the Michigan State University Soil and Plant Nutrient Laboratory for nitrate analysis. Weed counts In the first year of the study, weeds were counted in two 1 - m 2 quadrats per treatment plot in June at the time of the first herbicide application (Table F.1). Weed densities were so large in the second and third years of the study that weeds were counted in two 0.1 - m 2 quadrats per treatment plot. Weeds were identified to species. Corn measurements 111 Corn stand, chlorophyll content, tasseling, and leaf area index (LAI) data were collected on the dates listed in Table 3.6 . The height of 10 - 15 corn plants per plot was measured with a meter stick in June at corn stage V6 - V8. In late July to early August the percentage of plants in tassel in each plot was visually estimated by rating the plots as 0 - 25, 26 - 50, 51 - 75, or 76 - 100 percent in tassel. One non - destructive method of testing corn for relative N stat us is the use of a SPAD proxy for leaf chlorophyll content (Bullock and Anderson, 1998; Shapiro et al., 2006) . To assess corn N status, a Minolta SPAD - 502 chlorophy ll meter (Spectrum Technologies, Inc., Aurora, IL) was used to collect leaf chlorophyll content data at corn stages V6 and VT. The meter was placed on the newest fully mature leaf of 15 plants in each experimental unit. An AccuPAR LP - 80 photosynthetically - active radiation (PAR) sensor (Decagon Devices, Inc., Pullman, WA) was used to determine corn leaf area indices (LAI) at corn stage R6. Data were collected on days when the sky was clear. Statistical analysis Data were analyzed with SAS Enterprise Gu ide v. 6.1 (SAS Institute Inc., Cary, NC). Analysis of variance was carried out using PROC MIXED and the Tukey adjustment for all < 0.05). Data were tested for treatment*year interactions. Replication was treated as a random effect. Mean separation was conducted with the PDMIX 800 macro (Saxton, 1998) . An average soil bulk density of 1.6 g cm - 3 was used to convert soil nitrate values from ppm to kg ha - 1 using an average soil bulk density of 1.6 g cm - 3 (USDA - NRCS, 2015) . Corn height a s a percentage of the control was calculated by dividing the heights of the corn plants in each replicate by the average height of the corn in the bare ground control plots in each replicate and then multiplying by 100. Weed density data were transformed with log base 10 prior to 112 ANOVA and mean separation. Post - corn harvest soil NO 3 data were transformed using the natural log prior to analysis to improve normal ity. 113 APPENDIX C CHAPTER 3 SUPPLEMENTARY TABLES 114 115 < 0.05). ¶ Means were back - transformed from log base 10 after ANOVA and mean separation. § There was no treatment*year interaction (p = 0.37), so data were combined over years. 116 117 < 0.05). 118 LP - 80 photosynthetically active radiation (PAR) sensor (Decagon Devices, Inc., Pullman, WA). ¶ Leaf area index data were combined because there was no treatment*year interaction (P = 0.83). § Values in parentheses represent one standard error of the mean. 119 LITERATU RE CITED 120 LITERATURE CITED Akemo M.C., E.E. Regnier and M.A. Bennett. 2000. Weed suppression in spring - sown rye ( Secale cereale ): pea ( Pisum sativum ) cover crop mixes. Weed Technol 14: 545 - 549. Andraski, T.W. and L.G. Bundy. 2005. Cover crop effects on corn yield response to nitrogen on an irrigated sandy soil. Agron J 97: 1239 - 1244. Axelsen, J.A. and K.T. Kristensen. 2000. Collembola and mites in plots fertilised with different types of green ma nure. Pedobiologia 44: 556 - 56Bauer, P., J. Camberato and S. Roach. 1993. Cotton yield and fiber quality response to green manures and nitrogen. Agron J 85: 1019 - 1023. Bedoussac, L. and E. Justes. 2011. A comparison of commonly used indices for evaluating s pecies interactions and intercrop efficiency: Application to durum wheat winter pea intercrops. Field Crop Res 124: 25 - 36. Bensen, T. A., R.F. Smith, K.V. Subbarao, S.T. Koike, S.A. Fennimore, and S. Shem - Tov. 2009. Mustard and other cover crop effects var y on lettuce drop caused by Sclerotinia minor and on weeds. Plant Dis. 93: 1019 - 1027. Blanchart, E., C. Villenave, A. Viallatoux, B. Barthès, C. Girardin, A. Azontonde and C. Feller. 2006. Long - term effect of a legume cover crop ( Mucuna pruriens var. utilis ) on the communities of soil macrofauna and nematofauna, under maize cultivation, in southern Benin. Eur J Soil Biol 42: 136 - 144. Brainard D. C., Bellinder R. R., and V. Kumar. 2011. Grass - legume mixtures and soil fertility affect cover crop performa nce and weed seed production. Weed Technol 25 :473 - 479. Brennan, E.B. and R.F. Smith. 2009. Winter cover crop growth and weed suppression on the Central Coast of California. Weed Technol 19: 1017 - 1024. Bullock, D. and D. Anderson. 1998. Evaluation of the Minolta SPAD 502 chlorophyll meter for nitrogen management in corn. J Plant Nutr 21: 741 - 755. Carlson, S. and R. Stockwell. 2013. Research priorities for advancing adoption of cover crops in agriculture - intensive regions J Agric Food Syst Community Dev 3: 125 - 129. Cavadini, J.S. 2013. Cover crop bicultures and their effects on phosphorus cycling and soil Publishing # 1549309. Chen, G. and R.R. Weil. 2010. Penetration of cover crop roots through compacted soils. Plant Soil 331: 31 - 43. 121 Clark A.J. (ed.). 2008. Hairy vetch. In: Managing Cover Crops Profitably, 3 rd ed. Available at: htt p://www.sare.org/Learning - Center/Books/Managing - Cover - Crops - Profitably - 3rd - Edition Accessed 20 July 2015. Clark, A.J., A.M. Decker and J.J. Meisinger. 1994. Seeding rate and kill date effects on hairy vetch - cereal rye cover crop mixtures for corn production. Agron J 86: 1065 - 1070. Conklin, A.E., Erich, M.S., M. Liebman, D. Lambert, E.R. Gallandt and W.A. Halteman. 2002. Effects of red clover ( Trifolium pratense ) green manure and compost soil amendments on wild mustard ( Brassica kaber ) growth and in cidence of disease. Plant Soil 238: 245 - 256. Crandall, S., M. Ruffo and G. Bollero. 2005. Cropping system and nitrogen dynamics under a cereal winter cover crop preceding corn. Plant Soil 268: 209 - 219. Creamer, N.G., M.A. Bennett and B.R. Stinner. 1997. Ev aluation of cover crop mixtures for use in vegetable production systems. HortScience 32: 866 - 870. Creamer, N.G., M.A. Bennett, B.R. Stinner and J. Cardina. 1996. A comparison of four processing tomato production systems differing in cover crop and chemical inputs. J Am Soc Hortic Sci 121: 559 - 568. CTIC (Conservation Technology Information Center) and NCR - SARE. 2015. 2014 - 2015 annual report cover crop survey. Available at: http://www.sare.org/Learning - Center/From - the - Field/North - Central - SARE - From - the - Field/2015 - Cover - Crop - Survey - Analysis Accessed 20 July 2015. Dabney, S., J. Delgado and D. Reeves. 2001. Using winter cover crops to improve soil and water quality. Commun Soil Sci Plant Anal 32: 1221 - 1250. Daniel, J., A. Abaye, M. Alley, C. Adcock and J. Maitland. 1999. Winter annual cover crops in a Virginia no - till cotton production system: II. Cover crop and tillage effects on s oil moisture, cotton yield, and cotton quality. J. Cotton Sci 3: 84 - 91. Dapaah, H. and T. Vyn. 1998. Nitrogen fertilization and cover crop effects on soil structural stability and corn performance. Commun Soil Sci Plant Anal 29: 2557 - 2569. Dean, J.E. and R .R. Weil. 2009. Brassica cover crops for nitrogen retention in the Mid - Atlantic Coastal Plain. J Environ Qual 38: 520 - 528. De Bruin, J.L., P.M. Porter and N.R. Jordan. 2005. Use of a rye cover crop following corn in rotation with soybean in the upper Midwe st. Agron J 97: 587 - 598. Decker, A.M., A.J. Clark, J.J. Meisinger, F.R. Mulford and M.S. McIntosh. 1994. Legume cover crop contributions to no - tillage corn production. Agron J 86: 126 - 135. Decourtye, A., E. Mader and N. Desneux. 2010. Landscape enhancement of floral resources for honey bees in agro - ecosystems. Apidologie 41: 264 - 277. 122 Fowler, N. 1982. Competition and coexistence in a North Carolina grassland: III. Mixtures of component species. J Ecol 10 :77 - 92. Francis, G., K. Bartley and F. Tabley. 1998. T he effect of winter cover crop management on nitrate leaching losses and crop growth. J Agric Sci 131: 299 - 308. Groves, R., J. Walgenbach, J. Moyer and G. Kennedy. 2001. Overwintering of Frankliniella fusca (Thysanoptera: Thripidae) on winter annual weeds infected with tomato spotted wilt virus and patterns of virus movement between susceptible weed hosts. Phytopathology 91: 891 - 899. Haramoto, E.R. and E.R. Gallandt. 2005. Brassica cover cropping: I. Effects on weed and crop establishment. Weed Sci 53:695 7 01. Hauggaard - Nielsen, H., P. Ambus and E.S. Jensen. 2001. Interspecific competition, N use and interference with weeds in pea barley intercropping. Field Crop Res 70: 101 - 109. Hayden Z. D., Ngouajio M., and D.C. Brainard . 2014. Rye vetch mixture proportion tradeoffs: Cover crop productivity, nitrogen accumulation, and weed suppression. Agron J 106 :904 - 914. Hayden Z. D., Ngouajio M., and D.C. Brainard . 2015. Planting date and staggered seeding of rye vetch mixtures: Biomass, nitrogen, and legume winter survival. Agronomy Journal 107 : 33 - 40. Henderson, I., J. Vickery and N. Carter. 2004. The use of winter bird crops by farmland birds in lowland England. Biol Cons 118: 21 - 32. Hively, W. and W.J. Cox. 2001. Interseeding cover crops into soybean and subsequent corn yields. Agron J 93: 308 - 313. Isse, A., K. Stewart, D.L. Smith D, A.F. MacKenzie and D.C. Cloutier. 1999. Cover crops and nutrient retention for subsequent sweet corn production. Agron J 91: 934 - 939. Johnson, T., T. Kaspar, K. Kohler, S. Corak and S. Logsdon. 1998. Oat and rye overseeded into soybean as fall cover crops in the upper Midwest. J Soil Water Cons 53: 276 - 279. Kaspar, T., J. Radke and J . Laflen. 2001. Small grain cover crops and wheel traffic effects on infiltration, runoff, and erosion. J Soil Water Cons 56: 160 - 164. Kaspar, T., D. Jaynes, T. Parkin and T. Moorman. 2007. Rye cover crop and gamagrass strip effects on NO concentration and load in tile drainage. J Environ Qual 36: 1503 - 1511. Kuo, S. and E.J. Jellum. 2000. Long - term winter cover cropping effects on corn ( Zea mays L.) production and soil nitrogen availability. Biol Fertil Soil 31: 470 - 477. Kuo, S. and E.J. Jellum. 2002. Influ ence of winter cover crop and residue management on soil nitrogen availability and corn. Agron J 94: 501 - 508. 123 Kuo, S., U. Sainju U. and E.J. Jellum. 1997. Winter cover crop effects on soil organic carbon and carbohydrate in soil. Soil Sci Soc Am J 61: 145 - 152. Lawley, Y.E., R.R. Weil and J.R. Teasdale. 2011. Forage oilseed radish cover crop suppresses winter annual weeds in fall and before corn planting. Agron J 103: 137 - 144. Lehrsch, G. and J. Gallian. 2010. Oilseed oilseed radish effects on soil structure and soil water relations. J Sugar Beet Res 47: 1 - 21. Madden, N.M., J. Mitchell, W. Lanini, M. Cahn, E. Herrero, S. Park, S. Temple and M. Van Horn. 2004. Evaluation of conservation tillage and cover crop systems for organic processing tomato production. H ortTechnology 14: 243 - 250. Martins B.A.B., Sanchez - Olguin E., Perez - Jones A., and C. Mallory - Smith. 2012. ACCase cross - resistance in Italian ryegrass ( Lolium multiflorum ) sub - populations from Oregon, US. Julius - Kühn - Archiv 434:177 - 182. Meisinger, J., W. Ha rgrove, R. Mikkelsen, J. Williams and V. Benson. 1991. Effects of cover crops on groundwater quality, In: W.L. Hargrove (ed.) Cover Crops for Clean Water. Proceedings of an international conference. April 9 - 11, 1991., Soil and Water Conservation Society, A nkeny, IA., Jackson, TN. pp. 57 - 68. Melakeberhan, H., S. Mennan, M. Ngouajio and T. Dudek. 2008. Effect of Meloidogyne hapla on multi - purpose use of oilseed oilseed radish ( Raphanus sativus ). Nematology 10: 375 - 379. MCCC (Midwest Cover Crops Council). 2015 . Midwest Cover Crops Council cover crop decision tools. Available online at: http://mccc.msu.edu/selectorINTRO.html Accessed 10 July 2015. Miguez, F.E. and G.A. Bollero. 2005. Review of corn yield response under winter cover cropping systems using meta - analytic methods. Crop Sci 45:2318 - 2329. Miguez, F.E. and G.A. Bollero. 2006. Winter cover crops in Illinois. Crop Sci 46: 1536 - 1545. Miyazawa, K., M. Takeda, T. Murakami and T. Murayama. 2014. Dual a nd triple intercropping: Potential benefits for annual green manure production. Plant Prod Sci 17: 194 - 201. Möller, K. and H.J. Reents. 2009. Effects of various cover crops after peas on nitrate leaching and nitrogen supply to succeeding winter wheat or po tato crops. J Plant Nutr Soil Sci 172: 277 - 287. MSUEW ( Michigan State University Enviro - weather). 2015. Enviro - weather automated weather station network. Available at: http://www.agweather.geo.msu.edu/mawn/ Accessed 1 July 2015. 1 24 NOAA ( National Oceanic and Atmospheric Administration). 2015. Climate data online search. Available at: http://www.ncdc.noaa.gov/cdo - web/search?datasetid=GHCND Accessed 1 July 2015. Norsworthy, J.K., M. McClell and, G. Griffith, S.K. Bangarwa and J. Still. 2010. Evaluation of legume cover crops and weed control programs in conservation - tillage, enhanced glyphosate - resistant cotton. Weed Technol 24: 269 - 274. Odhiambo, J.J., and A.A. Bomke. 2001. Grass and legume cover crop effects on dry matter and nitrogen accumulation. Agron J 93: 299 - 307. Odhiambo, J.J., W.D. Temple and A.A. Bomke. 2012. Managing Cover Crops for Conservation Purposes in the Fraser River Delta, British Columbia. Available at: http://ainfo.cnptia.embrapa.br/digital/bitstream/item/59108/1/Crop - Management - Cases - and - Tools - for - Higher - Yield - a nd - Sustainability.pdf#page=49 Accessed 1 February 2015. O'Reilly, K.A., D.E. Robinson, R.J. Vyn and L.L. Van Eerd. 2011. Weed populations, sweet corn yield, and economics following fall cover crops. Weed Technol 25: 374 - 384. Parr, M., J. Grossman, S. Reber g - Horton, C. Brinton, C. Crozier. 2011. Nitrogen delivery from legume cover crops in no - till organic corn production. Agron J 103: 1578 - 1590. Perez - Jones A., Park K.W., Colquhoun J., Mallory - Smith C., and D. Shaner. 2005. Identification of glyphosate - resis tant Italian ryegrass ( Lolium multiflorum ) in Oregon. Weed Sci 53:775 - 779. Radersma, S. and A.L. Smit. 2011. Assessing denitrification and N leaching in a field with organic amendments. Wageningen J Life Sci 58 : 21 - 29. Ranells, N.N. and M.G. Wagger. 1996. Nitrogen release from grass and legume cover crop monocultures and bicultures. Agron J 88: 777 - 882. Ranells, N.N. and M.G. Wagger. 1997. Grass - legume bicultures as winter annual cover crops. Agron J 89: 659 - 665. Rosecrance, R., G. McCarty, D. Shelton , and J. Teasdale. 2000. Denitrification and N mineralization from hairy vetch ( Vicia villosa Roth) and rye ( Secale cereale L.) cover crop monocultures and bicultures. Plant Soil 227: 283 - 290. Ruffo, M.L., D.G. Bullock and G.A. Bollero. 2004. Soybean yield as affected by biomass and nitrogen uptake of cereal rye in winter cover crop rotations. Agron J 96: 800 - 805. Rusinamhodzi, L., M. Corbeels, J. Nyamangara and K.E. Giller. 2012. Maize grain legume intercropping is an attractive option for ecological inten sification that reduces climatic risk for smallholder farmers in central Mozambique. Field Crop Res 136: 12 - 22. Sainju, U.M., S. Rahman and B.P. Singh. 2001. Evaluating hairy vetch residue as nitrogen fertilizer for tomato in soilless medium. HortScience 3 6: 90 - 93. 125 Saxton, A.M. 1998. A macro for converting mean separation output to letter groupings in Proc Mixed. Proceedings of the 23rd SAS Users Group International, SAS Institute, Cary, NC pp. 1243 - 1246. Shapiro, C.A., J. Schepers, D. Francis and J. Shanahan. 2006. Using a chlorophyll meter to improve N management. University of Nebraska - Lincoln Extension Bulletin G 1632. Singer, J., S. Nusser and C. Alf. 2007. Are cover crops being used in the US corn belt? J Soil Water Cons 62: 353 - 358. Singogo, W. , W.J. Lamont and C.W. Marr. 1996. Fall - planted cover crops support good yields of muskmelons. HortScience 31: 62 - 64. Smith, R.G., K.L. Gross and G.P. Robertson. 2008. Effects of crop diversity on agroecosystem function: crop yield response. Ecosystems 11: 355 - 366. Smith, R.G., L.W. Atwood and N.D. Warren. 2014. Increased productivity of a cover crop mixture is not associated with enhanced agroecosystem services. PloS One doi: 10.1371/journal.pone.0097951. . Mustards a Brassica cover crop for Michigan. Michigan State Extension Bulletin E - 2956. Stivers - Young, L. 1998. Growth, nitrogen accumulation, and weed suppression by fall cover crops following early harvest of vegetables. HortScience 33: 60 - 63. Stroc k, J., P. Porter and M. Russelle. 2004. Cover cropping to reduce nitrate loss through subsurface drainage in the northern US Corn Belt. J Environ Qual 33: 1010 - 1016. Teasdale, J.R. 1996. Contribution of cover crops to weed management in sustainable agricul tural systems. J Prod Agric 9: 475 - 479. Teasdale, J.R. and A.A. Abdul - Baki. 1998. Comparison of mixtures vs. monocultures of cover crops for fresh - market tomato production with and without herbicide. HortScience 33: 1163 - 1166. Thorup - Kristensen, K. 2001. A re differences in root growth of nitrogen catch crops important for their ability to reduce soil nitrate - N content, and how can this be measured? Plant Soil 230: 185 - 195. Tonitto, C., M. David and L. Drinkwater. 2006. Replacing bare fallows with cover crop s in fertilizer - intensive cropping systems: A meta - analysis of crop yield and N dynamics. Agric Ecosys Environ 112: 58 - 72. Torbert, H.A., D.W. Reeves and R.L. Mulvaney. 1996. Winter legume cover crop benefits to corn: rotation vs. fixed - nitrogen effects. A gron J 88: 527 - 535. 126 Unger, P.W. and M.F. Vigil. 1998. Cover crop effects on soil water relationships. Journal of Soil and Water Conservation 53: 200 - 207. USDA - NRCS. 2015. Estimating moist bulk density by texture. Available at: http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/office/ssr10/tr/?cid=nrcs144 p2_074844 Accessed 20 February 2015. Vaughan, J.D. and G.K. Evanylo. 1999. Soil nitrogen dynamics in winter cover crop - corn systems. Commun Soil Sci Plant Anal 30: 31 - 52. Venkatesh, R., S.K. Harrison and R.M. Riedel. 2009. Weed hosts of soybean cyst nematode ( Heterodera glycines ) in Ohio. Weed Technol 14: 156 - 160. Villamil, M., G. Bol lero, R. Darmody, F. Simmons and D. Bullock. 2006. No - till corn/soybean systems including winter cover crops. Soil Sci Soc Am J 70: 1936 - 1944. Vyn, T.J., K.J. Janovicek, M.H. Miller and E.G. Beauchamp. 1999. Soil nitrate accumulation and corn response to p receding small - grain fertilization and cover crops. Agron J 91: 17 - 24. Vyn ,T.J., J.G. Faber, K.J. Janovicek and E.G. Beauchamp. 2000. Cover crop effects on nitrogen availability to corn following wheat. Agron J 92: 915 - 924. Wang, G., M. Ngouajio and D.D. Warncke. 2008. Nutrient cycling, weed suppression, and onion yield following brassica and sorghum sudangrass cover crops. HortTechnology 18: 68 - 74. Weil, R., C. White and Y. Lawley. 2009. Forage oilseed radish: A new multi - purpose cover crop for the Mid - At lantic. Maryland Cooperative Extension Fact Sheet 824. Weil, R.R. and J.E. Dean. 2009. Brassica cover crops for nitrogen retention in the Mid - Atlantic Coastal Plain. J Environ Qual 38: 520 - 528. White, C.M. and R.R. Weil. 2010. Forage oilseed radish and cer eal rye cover crop effects on mycorrhizal fungus colonization of maize roots. Plant Soil 328: 507 - 521. Williams, A.C. and B.C. McCarthy. 2001. A new index of interspecific competition for replacement and additive designs. Ecol Res 16: 29 - 40. Williams, S.M. and R.R. Weil. 2004. Crop cover root channels may alleviate soil compaction effects on soybean crop. Soil Science Society of America Journal 68: 1403 - 1409. Wortman, S., C. Francis and J. Lindquist. 2012. Cover crop mixtures for the western Corn Belt: Oppo rtunities for increased productivity and stability. Agron J 104: 699 - 705. Yenish, J.P., A.D. Worsham and A.C. York. 1996. Cover crops for herbicide replacement in no - tillage corn ( Zea mays ). Weed Technol 10: 815 - 821. 127 CHAPTER 4 IMPACT OF FALL - PLANTED COVER CROPS ON NITROUS OXIDE EMISSIONS ABSTRACT Cover crops may help retain N in agricultural systems and decrease emissions of the potent greenhouse gas (GHG) nitrous oxide (N 2 O). The purpose of this research was to investigate the impact annual ryegrass ( Lolium multiflorum Lam.) and oilseed radish ( Raphanus sativus L. var. oleiformis Pers.) on N 2 O emissions during the cover crop period and subsequent corn ( Zea mays L.) growing season. The experiment was conducted at W.K. Kellogg Biological Station (KBS) ( Hickory Corners, MI) and Michigan State University (MSU) (Lansing, MI) from fall 2012 - summer 2014 for a total of four site - years. Treatments included fall - planted oilseed radish, annual ryegrass, a mixture of oilseed radish and annual ryegrass, and a bar e ground control. Cover crops were planted after wheat ( Triticum aestivum L.) harvest and terminated with herbicides prior to planting field corn. Nitrous oxide emissions were measured three to five times during three periods: fall, spring, and summer. Max imum likely nitrogen inputs from cover crops ranged from 18 - 49 kg N ha - 1 at KBS and 35 - 83 kg N ha - 1 at MSU. After cover crop termination, spring - summer daily N 2 O - N emissions at KBS ranged from 1.3 - 1.9 and 1.6 - 2.6 g N ha - 1 in 2013 and 2014, respectively, an d from 1.8 - 4.2 and 8.1 - 13.7 g N ha - 1 at MSU in 2013 and 2014, respectively. Cumulative spring - summer N 2 O - N emissions at KBS were 222 - 325 and 266 - 429 g N ha - 1 in 2013 and 2014, respectively. At MSU, the 2013 and 2014 spring - summer cumulative N 2 O - N emissions were 322 - 769 and 1385 - 2361 g N ha - 1 , respectively. No differences were detected among treatments, including the bare ground control, for cumulative spring - summer N 2 O emissions, cumulative emissions scaled to total cover crop biomass, and cumulative emissi ons scaled to cover crop N. Nitrous oxide emissions did not represent a major pathway for 128 N loss in our study, and these results suggest that fall - planted non - legume cover crops do not increase N 2 O emissions in N - limited corn - based rotations. INTRODUCTION Crop production systems impact the global dynamics of three of the major greenhouse gases (GHGS): methane (CH 4 ), carbon dioxide (CO 2 ), and nitrous oxide (N 2 O) (Robertson et al., 2000) . Agriculture accounts for 8.1% of the total greenhouse gas emissions in the U.S., with cropland the source of 61% of direct N 2 O emissions and 79% of indirect N 2 O emissions (USEPA, 2014) . Nitrous oxide is the most potent of the agricultural GHGs. One molecule of N 2 O has about 300 times the capacity of one carbon dioxide molecul e to trap heat in the atmosphere (IPCC, 2007) , therefore even relatively small reductions in N 2 O emissions benefit the environment. Soil fertility and physical properties, weather, crop residue quality, and field operations can all influence N 2 O fluxes (Mi llar et al., 2010; Novoa and Tejeda, 2006). Nitrogen plays a key role in agricultural systems. Soil N availability is one of the major drivers of N 2 O emissions (Bouwman et al., 2002a; Bouwman et al., 2002b; Millar et al., 2010; Mosier et al., 1996; Robert son et al., 2000) . In particular, N fertilizer management, including amount, formulation, timing, and application method, affect N 2 O emissions ( Bouwman et al., 2002a; Bouwman et al., 2002b; Drury et al., 2012; Halvorson et al., 2010; Nash et al., 2012; Pel ster et al., 2011). From both an environmental and an economic perspective, it is beneficial to maximize nitrogen use efficiency (NUE) of cash crops and retain N in the cropping system between cash crops. Nitrogen which is not retained may be lost via leac hing, surface runoff, or gaseous emissions. It has been estimated that 50% of the N applied in agricultural systems is lost through these pathways (Tonitto et al., 2006) . Millar et al. (2010) state that a reduction in applied N fertilizer is the most relia ble way to reduce N 2 O emissions from agronomic systems. 129 Production practices which alter the amount of mineral N in the soil can affect N 2 O emissions by increasing or decreasing the amount of substrate available for the microbial processes of nitrification and denitrification responsible for N 2 O production (Robertson and Groffman, 2015) . Planting cover crops has been shown to increase, decrease, or have no impact on N 2 O emissions by affecting soil N availability (Baggs et al., 2000a; Basche et al., 2014; Wa gner - Riddle et al., 1997) . A number of factors may be responsible (Basche et al., 2014) . The C and N content and ratio in cover crop material helps determine the nature and quantity of substrate available for nitrification and denitrification, and impact t he synchrony between N release during cover crop decomposition and N uptake by the cash crop (Millar et al., 2004; Mitchell et al., 2013) . Due to its mobility, NO 3 that leaches through the soil profile and into waterways can later co ntribute to indirect N 2 O emissions in locations other than that to which the N was originally applied (Nevison, 2000) . Cover crops can reduce N leaching and runoff (Mitchell et al., 2013; Parkin et al., 2006; Syswerda et al., 2012). A meta - analysis by Toni tto et al. (2006) found that non - legume and legume cover crops decreased NO 3 leaching by 70% and 40%, respectively. Results from the limited literature on cover crops and their impact on greenhouse gas emissions vary. In 40% of the 26 studies selected for meta - analysis by Basche et al. (2014) , cover crops decreased N 2 O emissions. In the other 60% of the studies evaluated, cover crops increased N 2 O emissions. Legume cover crops generally had larger, positive response ratios (a quantification of the effect on N 2 O emissions) than grass cover crops, in the absence of applied fertilizer. As fertilizer application rates increased, the legume response ratios declined while the grass response ratios increased a small amount. In one of the few experiments examining g reenhouse gas emissions in wheat/corn/soybean [ Glycine max (L.) Merr.] rotations using cover crops, Parkin and Kaspar (2006) found that while corn plots emitted significantly more N 2 O than 130 soybean plots, the inclusion of a cereal rye ( Secale cereale L.) winter cover crop did not affect N 2 O emissions. Robertson et al. (2000) state that the primary driver of N 2 O fluxes in agricultural systems is the amount of N available in the soil; the more soil available N, the higher the flux. This is also supported by Gomes et al. (2009) , who found leguminous cover crops had larger cumulative N 2 O emissions during a 45 day period after cover crop residue management than a grass cover crop in no - till maize rotations in a subtropical climate. During this same period, Gomes et al. (2009) also found that N 2 O emissions were correlated with cover crop N added to the soil, after which total soil N content drove N 2 O emissions. Nitrous oxide emissions took longer to peak in a legume/grass cover crop mixture than in the legume alon e. The use of a cover crop mixture with a balanced C:N ratio may improve N synchrony by delaying N loss from the system (Aulakh et al., 1991; Baggs et al., 2000b; Gomes et al., 2009) . The objective of this study was to quantify the impact of the presence and absence of fall - planted cover crops with varying C:N ratios and biomass on N 2 O emissions. We hypothesize that N 2 O emissions will: a) be smaller in the cover crop treatment with a high C:N ratio (annual ryegrass) than the bare ground control, and b) be larger in the cover crop treatment with a low C:N ratio (oilseed radish) than the control. We further hypothesize that N 2 O emissions in the oilseed radish + annual ryegrass treatment will be intermediate between the annual ryegrass and oilseed radish monoc ulture treatments. MATERIALS AND METHODS Site characteristics and field operations The experiment was conducted from 2012 - 2014 at Michigan State University (MSU) (Lansing, MI, USA; latitude 42° 46' N, longitude 84° 34' W) and W.K. Kellogg Biological 131 Station (KBS) (Hickory Corners, MI, USA; latitude 42° 24' N, longitude 85° 24' W). The soil at KBS both years was a Kalamazoo loam (fine - loamy mixed mesic Typic Hapludalfs). In 2012 - 2013 the soil at MSU was a Riddles - Hillsdale sandy loam (fine - loamy, mixe d, mesic Typic Hapludalfs). In 2013 - 2014 the soil was a Capac loam (fine - loamy, mixed, mesic Aeric Ochraqualfs). Site characteristics are listed in Table 4.1. The 30 - yr mean annual temperatures at KBS and MSU, respectively, are 10.1 and 9.3° C (NOAA, 2015) . The 30 - yr average yearly precipitation is 1005 and 817 mm, respectively, at MSU and KBS. The experiment was structured as a randomized complete block design with four replicates per site - year. Experimental units were a minimum of 3 x 12 m in size. The fo ur OR) (11.2 kg ha - 1 ), annual ryegrass (44.8 kg ha - 1 ), and a mixture of oilseed radish and annual ryegrass (11.2 and 22.4 kg ha - 1 , respectively). Field operation dates are listed in Table 4.2. The preceding wheat crop was fertilized as per standard growing practices with 123 kg N ha - 1 at green - up. Cover crops were planted in 15 - cm rows in August - September following wheat harvest. At MSU the cover crops were planted with a no - till drill while at KBS the field was chisel plowed and cultivated prior to planting. The following spring, plots were sprayed with glyphosate ( N - (phosphonomethyl)glycine) and 2,4 - D ( 2,4 - dichlorophenoxyacetic acid) to terminate weeds and cover c rops which survived the winter. Because N fertilizer rate has been found to explain much of the variability found in other N 2 O emissions research (Hoben et al., 2011) , no fertilizer was applied to the cover crops in this study to avoid confounding or maski ng the effect of the cover crops on N 2 O emissions. Weeds were controlled in the bare ground treatment during the cover crop phase through the application of glyphosate once each fall. 132 During the corn phase, weeds were controlled in all plots with glyphosat e once or twice per summer as per standard practices. Data collection Temperature and precipitation data. Daily temperature and precipitation data for the KBS location were collected from the KBS Long Term Ecological Research (LTER) weather station ( http://lter.kbs.msu.edu/datatables/12 ). Temperature data and precipitation values for 1 May 31 October for the MSU location were collected from the Lansing/MSUHORT Enviro - weather station (MSUEW, 2015). Be cause the latter facility is not set up to measure solid precipitation, precipitation data for the MSU site for 1 November 30 April were collected from the National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS) website (NOA A, 2015). Cover crop data collection. Cover crop aboveground biomass was harvested from two randomly placed 0.25 m 2 quadrats per experimental unit each fall and again the following spring (Table 4.2). Biomass was separated into its component cover crop fractions before processing. Fall biomass data were collected at the point of peak biomass production before the onset of hard frosts. Spring biomass data were collected immediately prior to cover crop termination. Harvested plants were dried at 70° C for 5 - 10 d, weighed, and ground using a Wiley mill (Thomas Scientific, Swedesboro, NJ). The C:N ratio was determined via c ombustion at KBS in 2012 - 2013 using an Elemental Combustion System 4010 CHNS - O (Costech Analytical Technologies, Inc.; Valencia, CA, USA) and at Midwest Laboratories, Inc. (Omaha, NE, USA) in 2013 - 2014. 133 Soil sampling. To determine baseline site soil charac teristics, a 2 - cm diameter soil probe was used to collect 20 subsamples to a depth of 15 - 20 cm in each replicate of each field in the fall of each year. The samples were stored at 4° C, ground to pass through a 2 - mm mesh screen, and sent to the Michigan St ate University Soil and Plant Nutrient Laboratory to determine parameters including soil pH and organic matter. Values were averaged to create a composite sample for each field - year. Fall and spring soil NO 3 - values were determined from soil samples collec ted to a depth of 15 - 20 cm (Table 4.2) with a 2 - cm diameter soil probe. Five samples were collected in through a 2 - mm mesh screen. In 2012 - 2013, extractions were per formed and extracts were analyzed for NO 3 - - N and NH 4 + - N at KBS using a flow injector analyzer using the diffusion colorimetry technique for NH 4 + - N and cadmium reduction and colorimetry for NO 3 - - N. In 2013 - 2014, extracts were sent to the Michigan State Un iversity Soil and Plant Nutrient Laboratory for analysis. Using the same soil sample collection method as for the NO 3 - protocol above, subsamples were collected at corn V6 - V8 for pre - side dress NO 3 - - N testing (PSNT). Samples were aggregated by treatment, g round to pass through a 2 - mm mesh screen, and sent to the Michigan State University Soil and Plant Nutrient Laboratory for NO 3 - analysis. Nitrous oxide sampling and analysis. Cylindrical stainless steel chambers (Kahmark and Millar 2014; http://lter.kbs.msu.edu/citations/3418 ) were installed to a depth of 5 - cm in the soil after cover crop planting. Each chamber was centered over one cover crop row. Chambers were removed prior to tillage and corn plant ing, and immediately reinstalled between corn rows thereafter. A manual sampling chamber protocol was used to determine greenhouse gas fluxes (Holland et al., 1999). On each sampling date, chamber lids were installed and then headspace 134 gas was immediately extracted using a 10 - mL nylon syringe and a 23 - gauge needle. At 20 - min intervals over a 60 - min period, three more samples were collected from each chamber. Gas samples were placed in 5.9 mL Exetainer® vials (Labco Limited, UK), which had been previously fl ushed with 10 - mL of chamber air. Each vial was over - pressurized to 10 - mL to avoid contamination and facilitate analysis. Soil temperature near each experimental unit was collected, along with chamber height to soil surface measured at four points around th e circumference. Calculations to determine flux rates (µg N 2 O - N m - 2 hour - 1 ) were made using the following equation: N 2 A *60)/(A*MV corr ) 2 O concentrations during the period when the chamber is closed, V is the chamber headspace volume in liters, W A is the atomic mass of the N present in a molecule of N 2 O (28), 60 is a conversion factor from minutes to hours, A is the soil surface area covered by the chamber (m - 2 ), and MV corr corrects for temperature and pressure mole volume at sampling. N 2 O - N fluxes were then converted from µg N 2 O - N m - 2 hour - 1 to g N 2 O - N ha - 1 day - 1 for each sampling date. Samples were analyzed at the W.K. Kellogg Biological Station using an Agilent 7890A gas chromatograph (Agilent Indust ries, Inc.; Wilmington, DE, USA) fitted with a 63Ni electron capture detector and a Gerstel MPS2XL autosampler (Gerstel; Linthicum, MD, USA) (Kahmark and Millar, 2008) . There were three intensive gas sampling periods throughout the year. Sampling dates are listed in Table 4.2. Fall samples were collected three times over a 5 - wk period. A minimum of three sets of samples were collected over a 7 - 12 d period during oilseed radish decomposition after winter - kill in early spring (usually in late March). Samples were collected three times over a 4 - wk period to estimate spring gas fluxes near the time of corn planting, then once per month 135 during corn growth. Because of the length of time between fall and early spring gas sampling dates, cumulative N 2 O - N emissions w ere calculated separately for fall and spring - summer periods by interpolating the areas under the lines delimited by the N 2 O fluxes. Cumulative N 2 O emissions for each period were divided by the number of days in the respective period to derive daily emissi ons values. Data analysis Data were analyzed using analysis of variance (ANOVA) with PROC MIXED in SAS were transformed as necessary to achieve normality. Data for each site - year are presented individually, rather than combined across years, to allow for more nuanced interpretation of the N 2 O emissions data. When the F - test was significant, means were separated with the PDMIX 800 macro (Saxton, 1998) . Soil NO 3 v alues were converted from ppm to kg ha - 1 using an average soil bulk density of 1.6 g cm - 3 (Crum and Collins, 1995; USDA - NRCS, 2015) . The fall oilseed radish + annual ryegrass C:N ratio was calculated by taking a weighted average using the C:N ratios of the component cover crops and the percentage of dry aboveground biomass of each cover crop in each replicate. Because oilseed radish failed to survive the winter and annual ryegrass survival was low at the KBS site, in each site - year total dry aboveground bio mass was calculated using a combination of fall and spring data. In the annual ryegrass treatment, fall biomass values were considered as total biomass if spring biomass production was smaller than fall due to poor winter survival. If spring values were la rger than fall, then the spring biomass was taken as total biomass. In the oilseed radish treatment, fall biomass values were used as the total biomass because oilseed radish winter kills and no biomass was present in the spring. In the annual ryegrass + o ilseed radish treatment, oilseed radish fall biomass was added to the larger of 136 the fall or spring annual ryegrass biomass values. The total biomass was used to scale N 2 O spring summer emissions by dividing the total biomass by the cumulative spring su mmer N 2 O emissions. Spring summer N 2 O emissions were also scaled to cover crop N input using the same procedure to calculate maximum contributed N. Fall sampling cumulative emissions were not included in the calculations because all cover crop treatments were alive during the fall sampling period. RESULTS AND DISCUSSION Cover crop biomass Fall . At KBS, oilseed radish produced 42% more fall aboveground biomass than annual ryegrass in 2012; oilseed radish + annual ryegrass was intermediate between the two (Table 4.3). There were no differences in fall cover crop biomass at KBS in 2013 or at MSU in 2012. At MSU in fall 2013, the oilseed radish and oilseed radish + annual ryegrass treatments produced an average of 63% more biomass than the annual ryegrass treatment. In the oilseed radish + annual ryegrass treatment, 66% of the mixture biomass consisted of oilseed radish each year at KBS (data not shown). At MSU, oilseed radish comprised 70 and 90% of the mixture biomass in fall 2012 and 2013, respectively. Fall KBS cover crop biomass was 44 - 78% of that at MSU (Tale 4.3), probably as a result of differences in soil type and fertility. The differences were not likely to be the result of different planting dates, since KBS was planted earlier than MSU both yea rs (Table 4.2) and cover crop biomass production is usually larger with earlier fall planting dates. Fall oilseed radish biomass in this study was lower than that found in other studies. Fall - planted oilseed radish has been observed to produce 3,000 - 5,600 kg ha - 1 of aboveground biomass (Allison et al., 1998; Axelsen and Kristensen, 2000; Isse et al., 1999; Thorup - Kristensen, 2001; 137 Thorup - Kristensen, 2006) . On the other hand, annual ryegrass biomass production in this study was similar to that found in other studies. Annual ryegrass has been found to produce 1160 2290 kg ha - 1 of dry aboveground biomass in the fall (Dapaah and Vyn, 1998) and 2150 - 4310 kg ha - 1 by the end of March and April, respectively (Odhiambo and Bomke, 2001) . Spring. Across site - years, th e annual ryegrass treatment produced 1078 - 2145 kg ha - 1 biomass compared with 549 - 1601 kg ha - 1 biomass in the oilseed radish + annual ryegrass treatment (Table 4.3). The larger spring biomass production was not unexpected since the annual ryegrass seeding r ate in mixture was half that of the monoculture treatment and oilseed radish failed to overwinter. In contrast to the fall trend, in the spring the cover crops produced up to three times more biomass at KBS than at MSU. The larger winter precipitation at K BS (Figures 4.1 and 4.2) provided insulation that protected the annual ryegrass from freezing, decreasing the amount of winterkill. In addition, the larger amount of spring biomass at KBS in 2013 compared with 2014 may have been the result of the greater w inter precipitation in 2012 - 2013 (Figure 4.1). Total biomass. Fall and spring aboveground dry biomass data were used to calculate the total maximum likely amount of biomass contributed by each cover crop treatment from fall through spring, removing the co nfounding effects of the failure of oilseed radish to survive the winter and low annual ryegrass winter survival at the MSU site. Across fall and spring at KBS, oilseed radish + annual ryegrass produced 1827 - 2592 kg ha - 1 total biomass in 2012 - 2013 and 2013 - 2014, more than the 1234 - 1827 kg ha - 1 produced by oilseed radish (Table 4.3). No differences were detected in cover crop total biomass at MSU in 2012 - 2013. In 2013 - 2014, the oilseed radish + 138 annual ryegrass and oilseed radish total biomasses were 2642 and 2806 kg ha - 1 , respectively, compared with 1676 kg ha - 1 annual ryegrass total biomass. Cover crop C:N ratios, N and soil N Fall C:N ratios. No differences were detected in fall aboveground biomass C:N ratio in 2012 at KBS (Table 4.4). The C:N ratios ran ged from 17:1 - 20:1. In 2013 at KBS the C:N ratio of annual ryegrass was 13 - 29% higher than that of the oilseed radish + annual ryegrass and oilseed radish treatments, respectively. In fall 2012 at MSU the annual ryegrass treatment C:N ratio was 24:1, large r than the 16:1 ratio in the oilseed radish treatment. In fall 2013, the oilseed radish + annual ryegrass C:N ratio was 13:1, larger than the 9:1 - 12:1 observed in the other two cover crop treatments. The C:N ratios of the cover crops studied generally foll owed the expected pattern. Oilseed radish had the lowest C:N ratio, annual ryegrass the highest ratio, and the oilseed radish + annual ryegrass treatment was intermediate between the two. The exception was MSU in fall 2013, when the oilseed radish + annual ryegrass treatment had the highest C:N ratio. The C:N ratios in that site - year ranged from 9:1 13:1 (Table 4.4), unexpectedly low for all treatments. Given how low the C:N ratios were, it is unlikely that the differences were biologically significant. Spring C:N ratios. At KBS, the annual ryegrass C:N ratio was eight percent higher than the oilseed radish + annual ryegrass treatment in spring 2013 and 12% higher in 2014 (Table 4.4). No differences were detected at the MSU site, though in 2014 annual rye grass and oilseed radish + annual ryegrass had spring C:N ratios of 10:1 and 16:1, respectively. 139 The annual ryegrass in our study generally had higher C:N ratios in the fall than have been reported in the literature. For example, Thorup - Kristensen (1994) found annual ryegrass to have a C:N ratio of 12:1 in the fall. However, in that study fertilizer was applied to the cover crops, while in our study it was not. The spring annual ryegrass C:N ratios at the KBS location in our study were similar to those com monly found in the literature. Kuo and Sainju (1998) found fall - planted annual ryegrass to have a C:N ratio of 24:1 in the spring, while Baggs et al. (2000b) found it to have a C:N ratio of 25:1. As with annual ryegrass, oilseed radish C:N ratios in this s tudy were similar to those found in the literature, generally 13:1 to 18:1 (Baggs et al., 2000b; Thorup - Kristensen, 1994) . In our study, the C:N ratios for all treatments in both the fall and spring were higher at KBS than MSU (Table 4.3). The typical C:N ratio threshold above which N immobilization is expected to occur is 20 - 25:1 (Cochran et al., 1980; Kuo and Jellum, 2002) . The fall and spring C:N ratios in our study fell within a narrow range around this threshold. While it is possible that net N immobil ization occurred as a result of cover crop inclusion in this cropping system, it was unlikely to have been a major factor influencing N 2 O - N emissions. The question arises as to whether the statistical significances observed were also biologically significant, given both the generally low C:N ratios and the small percentage differences. Cover crop N input. To remove the differences caused by varying levels of winter survival, cover crop maximum N input was calculated in the same way as total biomas s. No differences were detected in cover crop N input at KBS in 2012 - 2013, probably because of the variability in the data (Table 4.4). In 2013 - 2014, the oilseed radish + annual ryegrass treatment had 43 - 67 % larger N input than the oilseed radish and annu al ryegrass treatments, respectively. At MSU, the cover crop N input was 1.4 - 1.7 times larger in the oilseed radish treatment than in the oilseed 140 radish + annual ryegrass and annual ryegrass treatments, respectively. No differences at MSU were detected in 2013 - 2014. The N input calculations suggest that at KBS, cover crops at most could have contributed 33 - 45 kg N ha - 1 in 2012 - 2013 and 18 - 30 kg N ha - 1 in 2012 - 2014. At MSU these values were, respectively, 35 - 60 and 68 - 83 kg N ha - 1 in 2012 - 2013 and 2013 - 2014. Given the generally greater biomass production (Table 4.3) and lower C:N ratios (Table 4.4) at MSU compared with KBS, it is unsurprising that the cover crops contributed more N to the cropping system at MSU than KBS. Soil NO 3 levels. Soil samples were co llected in the fall and spring during cover crop growth to test soil NO 3 levels. In fall 2013, NO 3 levels were 2.8 - 5.1 times greater in the control and oilseed radish treatments than in the annual ryegrass and oilseed radish + annual ryegrass treatments at KBS (Table 4.5), suggesting more soil N was available to undergo denitrification in these plots. At MSU, NO 3 levels were 128 - 249% greater in the control than in the cover crop treatments. At KBS in the spring, NO 3 levels were higher in the oilseed radish treatment than in the other treatments (8 kg N ha - 1 vs. 2.4 - 5.5 kg N ha - 1 ). Denitrification and subsequent N 2 O emissions would thus be expected to be higher in the oilseed radish plots than in other plots. At MSU in spring 2013 oilseed radish and control p lots had two times more kg N ha - 1 than the annual ryegrass and oilseed radish + annual ryegrass plots. In spring 2014 at MSU no differences were detected. Soil samples were collected twice during the corn phase of the experiment to test for soil N, once a t corn V6 - V8 for PSNT and again after corn harvest. PSNT data were pooled, so no statistical analysis was performed (Table 4.1). At KBS, control and oilseed radish plots tested at 20 30 kg N ha - 1 , while annual ryegrass and oilseed radish + annual ryegrass plots tested at 11 14 141 kg N ha - 1 . The range was narrower at MSU in 2013 in the spring, with 11 16 kg N ha - 1 . In 2014, soil NO 3 values were lower at KBS than MSU, ranging from 9 - 16 and 16 - 24 kg N ha - 1 , respectively. After corn harvest, no differences were ge nerally detected in soil NO 3 after corn harvest (Table 4.5). The finding in our study that annual ryegrass decreased soil NO 3 levels more than oilseed radish is in line with the results of other studies (Thorup - Kristensen, 1994; Vyn et al., 2000) . NO 3 res ults likely varied between locations at each sampling point due to a combination of different soil fertility levels (Table 4.1), weather (Figures 4.1 and 4.2), and cover crop biomass production and C:N ratios (Tables 4.3 and 4.4). Nitrous oxide emissions N 2 O fluxes. Nitrous oxide flux emissions were highly variable in our study between sites and years, and were an order of magnitude larger at MSU in spring - summer 2014 than 2013 (Figures 4.1 and 4.2). At both KBS and MSU, N 2 O emissions were smallest during t he fall sampling period of late October - December, peaked in March June, and then returned to near - fall levels in July September. The fluxes followed similar patterns at both locations over the course of each site - year. Cumulative N 2 O - N emissions were calculated for the fall and spring - summer sampling periods (Tables 4.6 and 4.7). N 2 O - N emissions were between 2 and 40 times higher at MSU than at KBS during both sampling periods each year. This is likely due to different soil types (Table 4.1), amount o f cover crop residue (Table 4.3), and cover crop C:N ratios and N inputs (Table 4.4) at the two locations. Soil organic matter (SOM) at MSU was more than twice that of KBS in 2013 - 2014. Lower SOM typically results in lower net N mineralization rates, leadi ng to less soil N available for nitrification and denitrification. Not only was SOM higher at MSU in 2014, but KBS cover crop total biomass was 60% lower than that of MSU while N 142 inputs at MSU were 3.2 times larger than at KBS (Tables 4.3 and 4.4). No diff erences were detected between treatments, including the bare ground control, in cumulative fall or spring - summer N 2 O - N emissions in any site - year. Cumulative N 2 O - N emissions in the fall were lower than in the spring - summer. During the 28 - d fall sampling pe riod in 2012 and the 35 - d period in 2013, N 2 O - N emissions ranged from 2 - 11 and 7 - 14 g N 2 O - N ha - 1 , respectively, at KBS (Table 4.6). At MSU the 2012 fall sampling period was 29 - d in length and cumulative emissions were 4 - 20 g N 2 O - N ha - 1 , while during the 39 - d 2013 fall sampling period cumulative emissions were 64 - 324 g N 2 O - N ha - 1 (Table 4.7). During spring - summer sampling period, cumulative emissions were 222 - 325 g N 2 O - N ha - 1 in 2013 at KBS and 266 - 429 g N 2 O - N ha - 1 in 2014 (Table 4.6). At MSU the spring - summ er 2013 cumulative emissions were 322 - 769 g N 2 O - N ha - 1 , while in 2014 they were 1385 - 2361 g N 2 O - N ha - 1 (Table 4.7). Spring - summer cumulative N 2 O emissions were scaled to total cover crop biomass and cover crop N inputs. There were no differences in N 2 O - N emissions per Mg cover crop biomass between any of the cover crop treatments and the bare ground control for any site - year (Tables 4.6 and 4.7). There were also no differences in N 2 O - N emissions per kg cover crop N input detected between treatments in any site - year (Tables 4.6 and 4.7). At the KBS location, 2014 N 2 O - N emissions per kg cover crop N were 1.5 - 2.7 times greater than 2013. At the MSU location, 2014 N 2 O - N emissions per kg cover crop N were 1.2 - 2.1 times greater than 2013. Interestingly, in spite of the differences in 2014 spring - summer N 2 O fluxes between the KBS and MSU sites (Figures 4.1 and 4.2), N 2 O - N emissions per kg cover crop N were of similar magnitude (Tables 4.6 and 4.7). As is typical with manual chamber sampling protocols, the discont inuous nature of the sampling may not have captured all large fluxes. However, our sampling method encompassed 143 periods of time during which a range of fluxes were encountered, enabling unbiased treatment comparisons to be made. Parkin and Kaspar (2006) found that including cover crops had no impact on N 2 O emissions when compared to their absence. Cover crop biomass production (Table 4.3) and C:N ratios (Table 4.4) may help explain the difference in the size of N 2 O emissions between the KBS and MSU sites . Spring biomass production and C:N ratios were both larger at KBS than MSU. Net immobilization of soil N could have decreased the amount of N available to the microbes that facilitate denitrification and thus also decreased N 2 O emissions. At MSU, spring c over crop biomass production was smaller and C:N ratios were under the 20:1 to 25:1 thresholds at which N immobilization becomes probable ( Cochran et al., 1980; Kuo and Jellum, 2002) . Given the availability of both C and N from the cover crops, and the ana erobic soil conditions typical of a wet Michigan spring, it is likely that most of the N 2 O emissions resulted from the denitrification process, corroborating isotopic studies by Ostrom et al. (2010) at nearby sites. Other research has also found N 2 O fluxes to be correlated with soil N in excess of crop uptake (McSwiney and Robertson, 2005; Van Groeningen et. al., 2010) . Based on meta - analysis and modeling done by Bouwman et al. (2002b) and Novoa and Tejeda (2006), about one percent of the N applied to a sys tem would be expected to be lost to N 2 O emissions. In a study by Gomes et al. (2009), one percent or less of the N in legume cover crop residue was lost to N 2 O emissions. We calculated that in our study, the probable maximum amount of N accumulated in cove r crop biomass tissue each year at KBS was 18 - 49 kg N ha - 1 (Table 4.4). Thus, a minimum estimate for N 2 O - N emissions would be 180 - 490 g N ha - 1 . Based on the probable maximum amount of N accumulated by cover crops at the MSU location (Table 4.4), we estimat e that N 2 O - N emissions would be in the range of 350 - 830 g N ha - 1 . At KBS, spring - summer cumulative N 2 O - N emissions were within the 180 - 490 g N ha - 1 estimate (Table 4.6). At 144 MSU, the 2013 spring - summer cumulative N 2 O - N emissions were also within the expecte d 350 - 830 g N ha - 1 range (Table 4.7). In 2014, however, N 2 O - N emissions were 1.7 - 2.8 times larger than the largest typical emissions based solely on cover crop N input. As has previously been discussed, the divergence from literature values was probably th e result of soil properties and surplus soil N. Overall, 1 - 2% of the N contributed to the soil by cover crop biomass was accounted for by N 2 O - N emissions; therefore, N 2 O emissions did not represent a major pathway for N loss from this cropping system. CO NCLUSIONS Contrary to our hypotheses and despite varying cover crop biomass and N inputs across site - years, the inclusion of fall - planted cover crops did not increase or decrease N 2 O emissions when compared to the bare ground control and each other. Overall, N 2 O emissions did not represent a major pathway for N loss. No differences were detected among treatments in terms of cumulative spring - summer N 2 O emissions or cumulative emissions scaled to total cover crop bio mass and N input. We may have failed to detect differences because leaching was a more prevalent pathway for N loss (leaving less N available for denitrification and emission), or because cover crop N inputs were too low to allow for differentiation betwee n treatments. These results suggest that while farmers may need to balance other advantages and disadvantages associated with fall - planted oilseed radish and annual ryegrass, impact on N 2 O emissions need not be a major concern. A future avenue of research to expand upon this work could include the repetition of this experiment with a split - plot design wherein N fertilizer application is a factor and gas and soil sampling are conducted more frequently to better quantify the movement of N 145 through the cropping system. It would be interesting to create an N budget by combining N 2 O and lysimeter sampling to quantify the partitioning between N 2 O emissions and NO 3 leaching. 146 APPENDICES 147 APPENDIX A CHAPTER 4 TABLES AND FIGURES 148 149 150 151 152 153 154 Figure 4.1. Temperature (°C), daily precipitation (mm), and nitrous oxide emissions (g N 2 O - N ha - 1 day) from 29 Oct. 2012 18 Sep. 2013 (year 1) and 28 Oct. 2013 17 Sep. 2014 (year 2) in Hickory Corners, MI (KBS). Each error bar represents one standard error from the mean. -30 -20 -10 0 10 20 30 40 0 10 20 30 40 50 60 70 80 90 Temperature ( C) Precipitation (mm) 0 5 10 15 20 25 g N 2 O - N ha - 1 oilseed radish annual ryegrass radish + ryegrass control Year 1 Year 2 155 Figure 4.2. Temperature (°C), daily precipitation (mm), and nitrous oxide emissions (g N 2 O - N ha - 1 ) from 11 Nov. 2012 27 Sep. 2013 and 24 Oct. 2013 19 Sep. 2014 in Lansing, MI (MSU). Note the right (2013 - 2014, year 2) y - axis is one order of magnitude larger than t he left y - axis. Each error bar represents one standard error from the mean. -30 -20 -10 0 10 20 30 40 0 10 20 30 40 50 60 70 80 90 Temperature ( C) Precipitation (mm) 0 50 100 150 200 250 300 350 0 5 10 15 20 25 30 35 2013 - 2014 g N 2 O - N ha - 1 2012 - 2013 g N 2 O - N ha - 1 oilseed radish annual ryegrass radish + ryegrass control Year 1 Year 2 156 APPENDIX B CORN GROWTH AND YIELD MATERIALS AND METHODS Field operations Field operation dates are listed in Table J.1. After cover crop termination as previously described, the field was prepared with a chisel plow and a soil finisher and 102 - 52 - - cm rows at a rate of 74,100 per ha - 1 using a four - row planter. Because fertilizer rate has been found to explain much of the variability in N 2 O emissions (Basche et al., 2014; Hoben et al., 2011) , no fertilizer was applied to the corn in this study to avoid confounding or masking the effect of the cover crops on N 2 O emissions. Weeds were controlled with glyphosate as needed at corn stages V4 - 6. Corn data collection Corn heights were measured when corn was at the V6 - V8 stage (Table J.1). A minimum of ten plants per experimental units were measured by holding the newest fully mature leaf up against a meter stick. Corn N status was determined after the onset of tasseling by collecting 25 corn ear leaves total (the leaf directly below the lowest ear of corn on each plant) from the two data rows in each experimental unit. Corn ear leaves were dried at 70 °C for 5 - 7 d and then grou nd with a Wiley mill (Thomas Scientific, Swedesboro, NJ). Two grams of this ground material were then sent to A&L Great Lakes Laboratories, Inc. (Fort Wayne, IN) for corn ear leaf N analysis. A Minolta SPAD - 502 chlorophyll meter (Spectrum Technologies, Inc ., Aurora, IL) was used to collect chlorophyll content data at corn stages V6 and VT. The meter was placed on the newest fully mature leaf of 15 plants in each experimental unit. An AccuPAR LP - 80 photosynthetically active radiation (PAR) sensor (Decagon De vices, Inc., Pullman, WA) was used to determine corn leaf area indices (LAI) at corn stage R6. Data were collected on days when the sky was clear. To determine corn grain yields, 9 m of each data row (outside of the area 157 in which corn ear leaves were colle cted) were harvested using a two - row research combine. Yield was adjusted to 15.5% moisture. Data analysis Data were analyzed as previously described in this chapter. Corn heights were converted into a percentage of the control by dividing the heights of the plants in each treatment in each replicate by the average height of the corn in the corresponding replic ate control treatment. 158 APPENDIX C CORN GROWTH AND YIELD TABLES 159 160 161 LITERATURE CITED 162 LITERATURE CITED Allison, M., M. Armstrong, K. Jaggard and A. Todd. 1998. Integration of nitrate cover crops into sugarbeet (Beta vulgaris) rotations. I. Management and effectiveness of nitrate cover crops. J Agr Sci 130: 53 - 60. Aulakh, M., D. Walters, J. Doran, D. Francis and A. Mosier. 1991. Crop residue type and placement effects on denitrification and mineralization. Soil Sci Soc Am J 55: 1020 - 1025. Axelsen, J.A. and K.T. Kristensen. 2000. Collembola and mites in plots fertilised with different types of green manure. Pedobiologia 44: 556 - 566. Baggs, E., R. Rees, K. Smith and A. Vinten. 2000a. Nitrous oxide emission from soils after incorporating crop residues. Soil Use Manage 16: 82 - 87. Baggs, E., C. Watson and R. Rees. 2000b. The fate of nitrogen from incorporated cover crop and green manure residues. Nut Cycl Agroecosys 56: 153 - 163. Basche, A .D., F.E. Miguez, T. Kaspar and M. Castellano. 2014. Do cover crops increase or decrease nitrous oxide emissions? A meta - analysis. J Soil Water Conserv 69: 471 - 482. Bouwman, A. F., L.J.M. Boumans and N.H. Batjes. 2002a. Emissions of N2O and NO from fertili zed fields: Summary of available measurement data. Glob Biogeochem Cycl : 16 (4): 1 - 13. Bouwman, A. F., L.J.M. Boumans and N.H. Batjes. 2002b. Modeling global annual N 2 O and NO emissions from fertilized fields. Glob Biogeochem Cycl 16(28) : 1 - 9. Cochran, V., L. Elliott and R. Papendick. 1980. Carbon and nitrogen movement from surface - applied wheat ( Triticum aestivum ) straw. Soil Sci Soc Am J 44: 978 - 982. Crum, J. and H. Collins. 1995. KBS soils. Available online at: http://lter.kbs.msu.edu/research/site - description - and - maps/soil - description/ Accessed 20 February 2015. Dapaah, H. and T. Vyn. 1998. Nitrogen fertilization and cover crop effects on soil structural stabilit y and corn performance. Commun Soil SciPlan 29: 2557 - 2569. Drury, C. F., W.D. Reynolds, X.M. Yang, N.B. McLaughlin, T.W. Welacky., W. Calder and C.A. Grant. 2012. Nitrogen source, application time, and tillage effects on soil nitrous oxide emissions and co rn grain yields. Soil Sci Soc Am J 76 : 1268 - 1279. 163 Gomes, J., C. Bayer, F. de Souza Costa, M. de Cássia Piccolo, J.A. Zanatta, F.C.B. Vieira and J. Six. 2009. Soil nitrous oxide emissions in long - term cover crops - based rotations under subtropical climate. S oil Till Res 106: 36 - 44. Halvorson, A. D., S.J. Del Grosso and F. Alluvione. 2010. Tillage and inorganic nitrogen source effects on nitrous oxide emissions from irrigated cropping systems. Soil Sci Soc Am J , 74 : 436 - 445. Hoben, J., R. Gehl, N. Millar, P. Grace and G. Robertson. 2011. Nonlinear nitrous oxide (N 2 O) response to nitrogen fertilizer in on farm corn crops of the U.S. Midwest. Global Change Biol 17: 1140 - 1152. Holland, E.A., G.P. Robertson, J. Greenberg, P.M. Groffman, R.D. Boone, and J.R. Gosz. 1999. Soil CO 2 , N 2 O, and CH 4 exchange. In: Standard Soil Methods for Long - Term Ecological Research. (eds Robertson GP, Bledsoe CS, Coleman DC, and P. Sollins) pp.185 - 201 Oxford, UK, Oxford University Press. IPCC. 2007. IPCC Fourth Assessment Report: The Ph ysical Science Basis. Available at: http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch2s2 - 10 - 2.html . Accessed 1 February 2015. Isse, A., K. Stewart, D.L. Smith, A.F. M acKenzie and D.C. Cloutier. 1999. Cover crops and nutrient retention for subsequent sweet corn production. Agron J 91: 934 - 939. Kahmark, K. and N. Millar. 2008. KBS LTER protocols: CH 4 , N 2 O, CO 2 fluxes (bucket method). Available at: http://lter.kbs.msu.ed u/protocols/113 Accessed 1 February 2015. Kuo, S. and E.J. Jellum. 2002. Influence of winter cover crop and residue management on soil nitrogen availability and corn. Agron J 94: 501 - 508. Kuo, S. and U. Sainju. 1998. Nitrogen mineralization and availabilit y of mixed leguminous and non - leguminous cover crop residues in soil. Biol Fertil Soil 26: 346 - 353. McSwiney, C.P. and G.P. Robertson. 2005. Nonlinear response of N 2 O flux to incremental fertilizer addition in a continuous maize ( Zea mays L.) cropping system. Global Change Biol 11: 1712 - 1719. Millar, N., J. Ndufa, G. Cadisch and E. Baggs. 2004. Nitrous oxide emissions following incorporation of improved fallow residues in the humid tropics. Global Biogeochem Cy doi 10.1029/2003GB002114 . Millar, N., G.P. Robertson, P.R. Grace, R.J. Gehl and J.P. Hoben. 2010. Nitrogen fertilizer management for nitrous oxide (N 2 O) mitigation in intensive corn (maize) production: an emissions reduction protocol for U.S. Midwest agriculture. Mitig Adapt Strategies Glob Ch ang 15: 185 - 204. Mitchell, D.C., M.J. Castellano, J.E. Sawyer and J. Pantoja. 2013. Cover crop effects on nitrous oxide emissions: Role of mineralizable carbon. Soil Sci Soc Am J 77: 1765 - 1773. 164 Mosier, A. R., J.M. Duxbury, J.R. Freney, O. Heinemeyer and K. Minami. 1996. Nitrous oxide emissions from agricultural fields: Assessment, measurement and mitigation. In Progress in nitrogen cycling studies (pp. 589 - 602). Springer Netherlands. MSUEW ( Michigan State University Enviro - weather). 2015. Enviro - weather au tomated weather station network. Available at: http://www.agweather.geo.msu.edu/mawn/ Accessed 1 July 2015. Nash, P. R., P.P. Motavalli and K.A. Nelson. 2012. Nitrous oxide emissions from claypan soils due to nitrogen fertilizer source and tillage/fertiliz er placement practices. Soil Sci Soc Am J 76 : 983 - 993. Nevison, C. 2000. Review of the IPCC methodology for estimating nitrous oxide emissions associated with agricultural leaching and runoff. Chemosphere Glob Chang Sci 2: 493 - 500. NOAA ( National Oceanic a nd Atmospheric Administration). 2015. Climate data online search. Available at: http://www.ncdc.noaa.gov/cdo - web/search?datasetid=GHCND Accessed 1 July 2015. Novoa, R. S. and H.R. Tejeda 2006. Evaluation of the N 2 O emissions from N in plant residues as affected by environmental and management factors. Nut Cycl Agroecosys 75 : 29 - 46. Odhiambo, J.J. and A.A. Bomke. 2001. Grass and legume cover crop effects on dry matter and nitrogen a ccumulation. Agron J 93: 299 - 307. Ostrom, N. E., P. H. Ostrom, H. Gandhi, N. Millar and G. P. Robertson. 2009. "Isotopologue data reveal denitrification as the primary source of nitrous oxide along a fertilization gradient in a temperature agricultural fie ld." In AGU Fall Meeting Abstracts , vol. 1, p. 0393. Parkin, T., T. Kaspar and J. Singer. 2006. Cover crop effects on the fate of N following soil application of swine manure. Plant Soil 289: 141 - 152. Parkin, T.B. and T.C. Kaspar. 2006. Nitrous oxide emis sions from corn soybean systems in the Midwest. J Environ Qual 35: 1496 - 1506. Pelster, D. E., F. Larouche, P. Rochette, M.H. Chantigny, S. Allaire and D.A. Angers. 2011. Nitrogen fertilization but not soil tillage affects nitrous oxide emissions from a cla y loam soil under a maize soybean rotation. Soil Till Res 115 : 16 - 26. Robertson, G. and P. Groffman. 2015. Nitrogen transformations. In: Soil Microbiology, Ecology, and Biochemistry. Academic Press, Burlington, MA. p. 421 - 446. Robertson, G.P., E.A. Paul an d R.R. Harwood. 2000. Greenhouse gases in intensive agriculture: contributions of individual gases to the radiative forcing of the atmosphere. Science 289: 1922 - 1925. 165 Saxton, A.M. 1998. A macro for converting mean separation output to letter groupings in P roc Mixed. In: Proceedings of the 23rd SAS Users Group International, SAS Institute, Cary, NC. pp. 1243 - 1246. Syswerda, S., B. Basso, S. Hamilton, J. Tausig and G. Robertson. 2012. Long - term nitrate loss along an agricultural intensity gradient in the Uppe r Midwest USA. Agric Ecosys Environ 149: 10 - 19. Thorup - Kristensen, K. 1994. The effect of nitrogen catch crop species on the nitrogen nutrition of succeeding crops. Nut Cycl Agroecosys 37: 227 - 234. Thorup - Kristensen, K. 2001. Are differences in root growth of nitrogen catch crops important for their ability to reduce soil nitrate - N content, and how can this be measured? Plant Soil 230: 185 - 195. Thorup - Kristensen, K. 2006. Effect of deep and shallow root systems on the dynamics of soil inorganic N during 3 - year crop rotations. Plant Soil 288: 233 - 248. Tonitto, C., M. David and L. Drinkwater. 2006. Replacing bare fallows with cover crops in fertilizer - intensive cropping systems: A meta - analysis of crop yield and N dynamics. Agric Ecosys Environ 112: 58 - 72. USDA - NRCS. 2015. Estimating moist bulk density by texture. Available at: http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/office/ ssr10/tr/?cid=nrcs144 p2_074844 Accessed 20 February 2015. USEPA. 2014. Inventory of US greenhouse gas emissions and sinks: 1990 - 2012. Available at: http://www.epa.gov/climatechange/Downloads/ghgemissions/US - GHG - Inventory - 2014 - Chapter - 6 - Agriculture.pdf. Accessed 1 February 2015 . Van Groenigen, J. W., G.L. Velthof, O. Oenema, K.J. Van Groenigen and C. Van Ke ssel. 2010. Towards an agronomic assessment of N 2 O emissions: a case study for arable crops. Eur J Soil Sci 61: 903 - 913. Vyn, T.J., J.G. Faber, K.J. Janovicek and E.G. Beauchamp. 2000. Cover crop effects on nitrogen availability to corn following wheat. Ag ron J 92: 915 - 924. Wagner - Riddle, C., G. Thurtell, G. Kidd, E. Beauchamp and R. Sweetman. 1997. Estimates of nitrous oxide emissions from agricultural fields over 28 months. Can J Soil Sci 77: 135 - 144.