|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 NITROGEN MASS BUDGET OF A SILAGE CORN FIELD AT THE UNIVERSITY OF FLORIDA DAIRY UNIT IN HAGUE, FL By REBECCA J HELLMUTH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 Rebecca J. Hellmuth
3 To my family
4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. George Hochmuth, for taking a chance on me and giving me the opportunity to pursue this degree. I also would like to thank him for all his help and direction throughout the process. I would like to thank my committe e members, Dr. Mark Clark, Dr. Adeg bola Adesogan, and Dr. Lynn Solle nberger for their guidance and input in my thesis work. I would like to thank Dawn Lucas who helped me with my field and lab work for always being patient and willing to answer all my questions. Your support and encouragement were greatly appreciated. thank my family for all their support and encouragement.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 13 Introduction ................................ ................................ ................................ ............. 13 Nutrient Mass Budgets ................................ ................................ ............................ 15 Components of Nutrient Mass Budgets ................................ ................................ .. 21 Soil Nitrogen ................................ ................................ ................................ ..... 22 Crop Uptake ................................ ................................ ................................ ..... 25 Leaching ................................ ................................ ................................ ........... 28 Volatilization ................................ ................................ ................................ ..... 31 Mineralization ................................ ................................ ................................ .......... 34 Objectives ................................ ................................ ................................ ............... 37 2 MATERIALS AND METHODS ................................ ................................ ................ 38 Dairy Unit Site Description ................................ ................................ ...................... 38 Nitrogen Mass Budget ................................ ................................ ............................ 40 Sampling Locations in the Fiel d ................................ ................................ ........ 41 Directly Measured Components of Nitrogen Mass Budget ............................... 42 Soil sampling and analysis ................................ ................................ ......... 42 Crop s ampling and analysis ................................ ................................ ....... 45 Leachate sampling and analysis ................................ ................................ 48 Mineralization Experiment ................................ ................................ ....................... 51 Data Analysis ................................ ................................ ................................ .......... 53 3 RESULTS AND DISC USSION ................................ ................................ ............... 58 Site Characterization ................................ ................................ ............................... 58 Spring Season ................................ ................................ ................................ ........ 59 N Inputs ................................ ................................ ................................ ............ 59 Initial soil N content ................................ ................................ .................... 59 Manure effluent and mineralized N ................................ ............................ 62 Inorganic N fertilizer ................................ ................................ ................... 65 Atmospheric deposition ................................ ................................ .............. 66 N Outputs ................................ ................................ ................................ ......... 66
6 Final soil N content ................................ ................................ .................... 66 Crop uptake ................................ ................................ ............................... 68 Leaching ................................ ................................ ................................ .... 72 Unaccounted for N ................................ ................................ ..................... 74 Summer Season ................................ ................................ ................................ ..... 76 N Inputs ................................ ................................ ................................ ............ 76 Initial soil N content ................................ ................................ .................... 77 Manure eff luent and mineralized N ................................ ............................ 78 Inorganic N fertilizer ................................ ................................ ................... 83 Atmospheric deposition ................................ ................................ .............. 84 N Outputs ................................ ................................ ................................ ......... 84 Final soil N content ................................ ................................ .................... 84 Crop uptake ................................ ................................ ............................... 85 Leaching ................................ ................................ ................................ .... 89 Unaccounted for N ................................ ................................ ..................... 92 Winter Season ................................ ................................ ................................ ........ 93 N Inputs ................................ ................................ ................................ ............ 93 Initial soil N content ................................ ................................ .................... 94 Manure effluent and mineralized N ................................ ............................ 94 Inorganic N fertilizer ................................ ................................ ................... 98 Atmospheric deposition ................................ ................................ .............. 99 N Outputs ................................ ................................ ................................ ......... 99 Final soil N content ................................ ................................ .................... 99 Crop uptake ................................ ................................ ............................. 101 Leaching ................................ ................................ ................................ .. 102 Unaccounted for N ................................ ................................ ................... 104 2011 2012 Cropping System Mass Balance ................................ ........................ 106 N Inputs ................................ ................................ ................................ .......... 106 N Outputs ................................ ................................ ................................ ....... 108 4 CONCLUSIONS ................................ ................................ ................................ ... 128 LIST OF REFERENCES ................................ ................................ ............................. 132 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 138
7 LIST OF TABLES Table page 2 1 Soil particle distribution by soil series (NRCS, 2010) ................................ .......... 55 2 2 Inputs and outputs of N mass budget ................................ ................................ 55 2 3 Installation and removal schedule for PVC N mineralization tubes for the summer silage corn crop and for the winter cover crop ................................ ...... 55 3 1 Physical and chemical properties of topsoil at lysimeter locations on 14 June 2012 ................................ ................................ ................................ ................. 112 3 2 Bulk density of soil samples (0 60 cm) on four sampling dates ........................ 112 3 3 Bulk density of soil samples by depth ................................ ............................... 112 3 4 Soil mineral N (nitrate N plus ammonium N) content in the 0 to 60 cm soil profile by s eason ................................ ................................ .............................. 112 3 5 Soil mineral N (nitrate N plus ammonium N) content divided into 30 cm increments by season ................................ ................................ ....................... 113 3 6 Soil nitrate N content in the 0 to 60 cm soil profile by season .......................... 113 3 7 Soil ammonium N content in the 0 to 60 cm soil profile by season ................... 113 3 8 Soil TKN content in the 0 to 60 cm soil profile by season ................................ 113 3 9 Soil nitrate N content i n 0 to 30 cm and 30 to 60 cm soil profile by sampling date ................................ ................................ ................................ .................. 114 3 10 Soil ammonium N content in 0 to 30 cm and 30 to 60 cm soil profile by sampling date ................................ ................................ ................................ ... 114 3 11 Soil TKN content in 0 to 30 cm and 30 to 60 cm soil profile by sampling date 114 3 12 Manure effluent N (TKN) application by month ................................ ................. 114 3 13 Inorganic N fertilizer (Total N) application by month ................................ ......... 115 3 14 Wet ion deposition by season (NADP, 2011) ................................ ................... 115 3 15 Crop dry weight yield by season ................................ ................................ ....... 115 3 16 University of Florida corn silage variety trial dry matter yields and crude protein concentration in 2011 (UF DAS, 2011) ................................ ................. 115
8 3 17 Mean N content of dry silage corn plant parts by season ................................ 116 3 18 Mean N concentration of silage corn plant parts by season ............................. 116 3 19 Harvested crop uptake and root and stubble uptake (N kg ha 1 ) on Field J at the Dairy Unit by season ................................ ................................ .................. 116 3 20 Sum of leachate loads (N kg ha 1 ) over 2011 2012 season .............................. 116 3 21 Nitrate N concentration (mg L 1 ) of leaching events by season ........................ 116 3 22 N mass balance of spring silage corn crop at the Dairy Unit ............................ 117 3 23 Soil mineral N (nitrate N plus ammonium N) content of each location over three time periods for the summer mineralization experiment on Field J at the Dairy Unit ................................ ................................ ................................ .......... 117 3 24 Change in soil mineral N content over three time periods for the summer mineralization experiment on Field J at the Dairy Unit ................................ ...... 118 3 25 N mass balance of summer silage corn crop at the Dairy Unit ......................... 119 3 26 Soil mineral N content of each location over three time periods for the winter mineralization experiment on Field J at the Dairy Unit ................................ ...... 119 3 27 Change in the soil mineral N content over three time periods for the winter mineralization experiment on Field J at the Dairy Unit ................................ ...... 120 3 28 N content of rye/ryegrass plant parts ................................ ................................ 121 3 29 Mean percent N concentration of rye ryegrass plant parts ............................... 121 3 30 N mass balance of winter rye/ryegrass crop at the Dairy Unit .......................... 121 3 31 Overall N mass balance of 2011 2012 corn corn rye/ryegrass cropping system at the Dairy Unit ................................ ................................ ................... 121
9 LIST OF FIGURES Figure page 2 1 Pre selected sampling locations in F ield J ................................ ........................... 56 2 2 Selection method of six additional sampling locations ................................ ........ 56 2 3 Drainage lysimeter design ................................ ................................ .................. 57 3 1 Manure effluent daily N application ................................ ................................ ... 122 3 2 Historical rainfall by year (2007 2011) from March 15 to June 25 in Alachua county ................................ ................................ ................................ ............... 123 3 3 N leaching versus rainfall, fresh water irrigation, and manu re effluent application (kg ha 1) ................................ ................................ ......................... 124 3 4 Historical rainfall by year (2007 2011) from June 2 to September 21 in Alachua county ................................ ................................ ................................ 12 5 3 5 Net mineralization/immobilization of summer mineralization experiment .......... 125 3 6 Mineral N content of summer mineralization experiment PVC pipes installed on 24 July 2011 ................................ ................................ ................................ 126 3 7 Net mineralization/immobilization of winter mineralization experiment ............. 126 3 8 Mineral N concentrations of winter mineralization experiment PVC pipes installed at 8 Nov. 2011 ................................ ................................ .................... 127 3 9 Historical rainfall by year (2007 2012) from September 21 to March 12 in Alachu a county ................................ ................................ ................................ 127
10 LIST OF ABBREVIATION S A RL Analytical Research Laboratory B MP Best Management Practices C AFO Concentrated Animal Feeding Operation E PA Environmental Protection Agency E STL Extension Soil Testing Laboratory F DEP Florida Department of Environmental Protection F NUE Fertilizer N itrogen Use Efficiency G PS Global Positioning System I FAS Institute of Food and Agricultural Sciences K Potassium M CL Maximum Contaminant Load N Nitrogen N ADP National Atmospheric Deposition Program N BS National Bureau of Standards N ELAC National Environmental Laboratory Accreditation Conference N MP Nutrient Management Plan N RCS USDA Natural Resource Conservative Serv ice P Phosphorus P VC Polyvinyl Chloride T KN Total Kjeldahl Nitrogen U F University of Florida
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science NITROGEN MASS BUDGET OF A SILAGE CORN FIELD AT THE UNIVERSITY OF FLORIDA DAIRY UNIT IN HAGUE, FL By Rebecca J. Hellmuth May 2013 Chair: George Hochmuth Major: Soil and Water Science This study was conducted to identify potential environmental impacts of silage production on a northeast Florida dairy farm and provide recommendations to minimize excess nitrogen pollution. Agriculture is a contributor of non point source nutrient polluti quantified for a production field during the 2011 2012 crop year using farm records, estimates from the NADP, and direct measurement of soils, crops, and leaching. Silage corn was grown in the spri ng and summer seasons and rye ryegrass was grown in the winter season All crops were fertilized by inorganic fertilizer and manure effluent available to crops as mineralized N The N balance inputs were soil mineral N content (286 kg ha 1 N), mineralized N (407 kg ha 1 N), inorganic fertilizer (139 kg ha 1 N), and atmospheric deposition (2 kg ha 1 N). The N outputs were soil mineral N content (28 kg ha 1 N), harvested crop uptake (423 kg ha 1 N), root and stubble uptake (131 kg ha 1 N) and leaching (23 kg ha 1 N). Unaccounted for losses, consisting of gaseous losses, (229 kg ha 1 N) were calculated as the difference in inputs and outputs. Leaching (3% of N outputs) was not a substantial N loss impacting the environment G aseous losses were 27% of N outputs due to volatilization of excessive manure effluent applications during
12 the spring season Therefore, more even fertilizer distribution throughout the crop year was recommended. Even with poor manure effluent management, in low rainfall years leaching may be minimal.
13 CHAPTER 1 LITERATURE REVIEW Introduction Dairies are an important agricultural industry in Florida producing 2.1 billion pounds of milk per year statewide (De Vries and Giesy, 2009). Balancing milk production with environment al concerns is a major issue for dairy fa rms. Dairy farms (US EPA, 2005 ). Non point source nutrient pollution is the main cause of impaired water quality in the United States (US EPA, 2005 ). Agric ulture is the greatest contributor of non point source pollution contributing excess nitrogen ( N ) and phosphorus (P) from commercial fertilizers and livestock manure (US EPA, 2005 ). The excess nutrients may increase algae growth (eutrophication) in water b odies which can eventually create hypoxic conditions detrimental to the environment (Shepard, 2005). In Florida, eutrophication is a concern in Lake Okeechobee where excess P contamination runs off from dairies during the rainy season ( Zhang et al., 2007 ). The agriculturally based Suwannee River basin watershed in Florida is vulnerable to contamination due to losses of fertilizer and distribution of manure waste as organic amendments applied by farmers to produce crops in the watershed. The low water hold ing capacity of the sandy soil in the region puts excess nitrate and P at a high risk for leaching into the groundwater spring system after rainfall or excessive irrigation (Mylavarapu, 2003). Over the past 20 years, nitrate levels in the groundwater have increased due to nonpoint nutrient sources such as fertilizers and animal wastes. Extensive efforts have been made by the Suwannee River Partnership (suwannee.org) to reduce nitrate levels in surface waters and groundwater. The Partnership helps
14 farmers im plement voluntary or incentive based programs to protect and conserve water resources using research based best management practices. In accordance with the United States Clean Water Act, the National Polluta nt Discharge Elimination System regulates point sources such as dairies by requiring permits in order to control pollution of the waters o f the United States (NPDES, 2011) Dairies are considered an Animal Feeding Operation where animals are raised in confined conditions; instead of grazing, feed is bro ught to the animals (NPDES 2011). In Florida, dairy farms are regulated by the Florida Department of Environmental Protection ( F DEP). Dairies are required by F DEP to maintain minimum ground water quality standards by installation of water monitoring wells and requiring nutrient management plans including m anure disposal plans ( FDEP, 2010 ). N pollution is the main contributor to all water quality problems in northern Florida. plu s clay) meaning the soils have the ability to retain large amounts of P (NRCS, 2010). If soils are coated and the water table is not classified as high in geographic elevation, then fertilizer recommendations are designed for N rather than for P loadings. Therefore, N is the most common mineral fertilizer applied to agricultural lands, because N is generally the most limiting nutrient to optimum crop production (Follett and Delgado, 2002). Excess N can cause environmental problems as well as harmful effects to infants if high levels of nitrates are consumed through drinking water (EPA, 2010). Methemoglobinemia, can result from ingesting drinking water containing high levels of nitrates, which inhibit transport of oxygen by the blood. This syndrome has been
15 d ocumented in infants younger than three months and is therefore also known as blue baby syndr ome (Follett and Walker, 1989). In the United States, t he current national standard for the maximum contaminant level (MCL) for nitrate N in drinking water is 10 mg L 1 Dairy farmers must prove that their water monitoring wells adhere to these standards for groundwater Nutrient management plans d eveloped for each dairy by the USDA Natural Resource Conservation Service (NRCS) are required to control nutrien t pollution on dairy farms in order to meet thes e standards. These plans provide guidelines to correctly deal with sources of nutrients such as manures and fertilizers in order to avoid nutrient pollution. They are specific to each farm and take into accou nt factors impacting the flow of nutrients on the farm such as the number of cows, type of waste management system, feed, typical farming practic es, soil type, and fertilizers. The nutrient management plan must be approved by the F DEP and is a contract bet ween the dairy operation and the F sources of nutrients on the farm. Shepard (2005) conducted a survey of farmers in two Wisconsin watersheds and showed that farmers with nutri ent management plans applied less total N and P than farmers without nutrient management plans. Nutrient Mass Budgets Nutrient mass budgets account for the inputs and outputs of nutrients in a chos en area such as a dai ry farm. S ome outputs are potential lo sses to the environment and therefore may constitute nutrient pollution. Typical inputs for an agricultural area are fertilizer and feeds bought off the farm; typical outputs include runoff, leaching, atmospheric losses, and agricultural products sold off the farm. Nutrient mas s budgets are used to evaluate the efficiency of the nutrient management plan determine
16 environmental concerns and provide economic evaluation of the flow of nutrients on a farm. Kuipers et al. (1999) conducted a study in the Nether lands where the government developed targets to reduce N losses from a griculture starting in the 1990 s. Standards were set for groundwater that it should contain less than 11.3 mg L 1 of nitrate N Kuipers et al. (1999) describe d the Netherlands use of gov ernment imposed nutrient balance sheets to control nutrient losses of N and P. Surpluses of N and P would result in a tax being imposed on a dairy farmer. Dairy farmers were expected to adjust farming practices to lessen nitrate leaching to groundwater and lessen P accumulation in the soil (Kuipers et al., 1999). Korevaar (1992) describe d several farming practices used in th e Netherlands to reduce N losses such as manure slurry injection, covering slurry storage, growing catch crops, increasing maize silage feeding while reducing grazing, an d reducing N application rate on grazing fields Nutrient mass budgets can be constructed for single fields, whole farms, water sheds, or regions by identify ing the nutrients cycling through a given area. Because nutrient cycling within a given area is never a fully closed system, estimates or modeling of potential inputs and outputs are sometimes necessary. Many times, certain pools of nutrients are very difficult to qua ntify. In the case of a N mass budget, volatilization denitrification, and atmospheric deposition are the most difficult to quantify because they deal with N in the atmosphere Rotz et al. (2005) used measured or estimated inputs, outputs, and flows to construct a complete farm balance with nutrient flow s o f N and P over two grassland farming systems in Germany and the Netherlands. These farm balances were used to evaluate nutrient management relative to average
17 commercial farms in terms of economics and environmental impacts. Models such as the Integrated F arm System Model (Rotz et al. 1999 ; Rotz et al. 2005) and the Dynamic North Florida Dairy Farm Model (Cabrera et al. 2005 a ) are used to understand how chang es in management decisions change the nutrient mass budget and in turn affect economics and envir onmental concerns. Kohn et al. (1997) modeled a 35 ha dairy farm using herd efficiency, crop production and feed purchase coefficients determined by Dou et al. (1996) and simulated four management scenarios of feed intensive, legume intensive, fertilizer i ntensive, and manure importer. Kohn et al. (1997) used N balance equations for the overall model farm to perform sensitivity analysis of herd nutrition, manure management, and crop selection on nutrient losses. In the model, Kohn et al. (1997) characterize d farm inputs as legume fixation, imported feeds, and imported fertilizers; outputs as crop production; and losses as leaching, runoff, volatilization, and denitrification. The researchers found that importing feed resulted in the lowest N losses relative to production. Farms using legumes for N fixation had more efficient N utilization and fewer N losses from leaching, runoff, or denitrification than farms applying inorganic N fertilizers. Improvement of N uptake of available soil N had a greater impact in farm efficiency and therefore fewer N losses than a similar improvement in manure availability. Historical records from farms coupled with research sampling are often used to constr uct nutrient mass budgets (Bacon et al., 1990; Klausner et al., 1998; Powe ll e t al., 2007). Typically, nutrient outputs are subtracted from nutrient inp uts to obtain a nutrient mass balance (Bacon et al. 1990; Rotz et al. 2005). Bacon et al. (1990) conducted a nutrient cycling study on a Pennsylvania dairy farm using farm reco rds and samples of
18 crops and purchased inputs for measurements of moisture and nutrient content. Nutrient mass balances were obtained by subtracting nutrient outputs from nutrient inputs; factors such as volatilization losses, residual manure decomposition leaching, denitrification, and biological N fixation were estimated due to the difficulty in direct measurement. A large net accumulation of N, P, and K occurred in the soil and one source was the nutrients in imported feedstuffs. A reduction in importe d feedstuffs was recommended. Once implemented, the reduction in importing feedstuffs did not significantly affect milk production but it provided an e conomic benefit to the farm and decrease d nutrients in the soil provid ing a potential benefit to the env ironment (Bacon et al., 1990) Powell et al. (2007) surveyed 54 dairy farms in Wisconsin to evaluate whether actual farmer nutrient management behaviors conformed to their farm specific nutrient management plans to minimize negative effects to water quality Manure collection m ethods of the dairies were examined first t hen feed and manure data were validated and the accuracy of farm records was analyzed over a whole farm basis. Finally, on a field by field basis, nutrient application was studied to d etermine if it conformed to state standards and how application corresponded to farm characteristics including livestock inventories, cropping practices, and field histories Manure collection and management practices were recorded by farmers and included type of manure, application method, quantity spread, and fields receiving the manure. M anure samples were also collected by farmers periodically and analyzed. Powell et al. (2007) found that N applications were approximately 40% fertilizer, 30% manure, and 30% from previous legume crops Fields which received low (1 80 kg
19 ha 1 ) available N application (in the form of fertilizer, manure, and previous legume) on their total corn land represented 38, 51, and 34% of the studied farms in the Northeast, South Cen tral, and Southwest regions of Wisconsin respectively. The large majority of farms had applied N within the agronomic N recommendations for corn (81 to 240 kg ha 1 ). Excessive N inputs (>240 kg ha 1 ) of fertilizer, manure, and previous legumes occurred on 6 of 12 farms in the Northeast, 5 of 12 farms in the South Central, and 6 of 9 f arms in the Southwest region, representing 40, 31, and 34%, respectively, of the total corn area on those farms. Powell et al. (2007) found that that few Wisconsin dairy farms use d nutr ient management practices that we re detrimental to surface water quality management areas An imp ortant factor to optimize nutrient management i s having a n a dequate amount of cropland to spread manure. Klau sner et al. (1998) calculated an overall farm nutrient mass balance of N, P, and K for a dairy farm in central New York. Farm records, legume acreage, and percentage legume of the forage crop were used to obtain nutrients in feeds, fertilizers, cattle, milk, and biological N fixation. Biological fixation of atmospheric N was estimated to be 40% of the legume N content at harvest. Forage and milk analyses were performed by a lab and nutrient content of purchased feed was obtained from the supplier. Nutrient composition of purchased and sold cattle was estimated. Soil testing, manure analysis, crop analysis, and feed analysis were performed to assess nutrient contents of nutrient imports and exports. Hutson et al. (1998) determined N leaching by background modeling, soil tests, and records using soi l hydraulic conduc tivity, cropping patterns, and conditions of the soil surface due to rainfall, temperature, eva poration, and nutrient addition
20 Klausner et al. (1998) found that 60 to 72% of imported N, P, and K were in excess of nutrient exports and the majority of the excess nutrients were from purchased feedstuffs. After the nutrient mass budget was evaluated, a reformulation of feed was recommended an d it resulted in reduction of total N excretion and increased net farm income by $40,200. N w as defined as the most limiting crop nutrient. C rop nutrient management recommendations were based on crop nutrient requirements taking into account soil tests, st arter fertilizer, and residual manure N available Decreasing the use of commercial synthetic fertilizer produced an additional $1 350 in net farm income, but manure storage capacity constraints hindered further economic benefit. Construction of a manure s torage pond to provide more manure based nutrient capacity was considered, but costs associated with the pond would have resulted in a net financial loss. Environmental implications of excess nutrient imports were not discussed. A difference in whole farm imports and exports of 46.7 Mg year 1 of N was reported, but estimations of possible exports to account for the difference such as volatilization, denitrification, leaching, runoff, etc. were not reported. Hall and Ri sser (1993) conducted a N budget on a Pennsylvania dairy farm to determine the losses of N to the environment. It was found that 37% of mean annual N outputs were from harvested crops, 25% of N losses from volatilization of N in applied manure, 38% of the losses were from N leaching to ground water, and less than 1% from surface runoff. Wang et al. (2000) used a whole herd optimization model developed for the Cornell Net Carbohydrate and Protein System model to evaluate the nutrient mass balance of a dairy farm and how changes in feed would aff ect economics and
21 environmental impacts. Dividing lactating dairy cows feed ration based on their level of milk production decre ased the remaining N (Imports minus Exports) in the N mass balance (Mg year 1 ) from 51.7 to 44. 7 Mg year 1 Increasin g forage q uality (lower neutral detergent fiber and higher crude protein) did not improve the N mass balance due to increased N fixation by the forage crop instead of greater crop uptake of applied N but improving overall yields to the maximum potential reduced the remaining N in the N mass balance by 29 Mg ye ar 1 Van Horn et al. (1996) developed a whole farm nutrient budget on a manure irrigated field in Tifton, Georgia using inputs of nutrients from animal manures and outputs of potential plant removal and losses due to manure management and fertilizer management of crop production. Exporting nutrients off farm, if necessary, was considered as an alternative output of nutrients. Van Horn et al. (1996) determined the t otal manure nutrient excretion, estimated volat ilization of N in manure before flushing and during holding and irrigation soil nutrient content, nutrients in rainfall, nutrients lost to surface runoff, nutrients lost to groundwater from leaching, nutrients in harvested crop uptake r ecycled feed nutri ent content, and purchased feed stuffs should all be included in the nutrient mass budget. Van Horn et al. (1996) highlighted the importance of nutrient budgets to document nutrient accountability and their use in optim izing the allocation of manure resources. Components of Nutrient Mass Budgets Th e research cited above describes the importance of nutrient mass budgets in assessing nutrient flows on a farm. The most important nutrient s studied are N and P and their quantification involves measuring N and P in various pools For example when conducting a N mass budget for a single agricultural field, the main N pools are in the
22 soil, crops, runoff, leaching, rain water, volatilization, denitrification, and atmospheric deposition. The pools are then separated into inputs and outputs to build the N mass budget. The following discussion describes the major N pools on farms and how the pools are quantified. Soil Nitrogen The N content of the soi l consists of ino rganic N (nitrate N and ammonium N ) and organic N. The amounts of inorganic N in the soil and potentially mineralized organic N are the forms available for crop uptake and therefore should be taken into account when determining the amount of fertilizer and manures to add for crop N needs. Soil nitrate N is important to consider because it is highly susceptible to leaching into groundwater and to denitrification losses. A pplication of excessive N for pl ant needs may result in N accumu latio n in the soil or to leaching below the root zone. Available soil N can be measured by soil testing or indirectly measured by growing non legume crops on unfertilized plots and measuring the resulting crop biomass for N content. This indirect method results in a slight undere stimation of the total N content of the soil (Haefele et al., 2002). Direct s oil testing for N content is the most common method used. In Swanton, Vermont, Jokela (1992) measured N fertilizer and manure application effects on soil nitrat e N and corn yield by sampling the soil profile from 0 to 150 cm for nitrate N content twice a year in May and October /Nov ember from Nov ember of 1986 to May of 1989. Jokela (1992) found that below 90 cm very little nitrate N was present and most nitrate N was f ound in the upper 60 cm. He found that soil nitrate N concentration was higher in Oct ober /Nov ember than in May for most treatments due to losses from leaching, denitrification, or immobilization during the previous fall and early spring. Application of manures at diff erent treatment levels
23 resulted in similar or only slightly higher soil nitrate N levels indicating the additional N with the higher rates of N was taken up by the crop. Van Horn et al. (1996) conducted soil testing on the sur face 30 cm of the soil profile on a manure irrigated field in Tifton, Georgia They found that soil inorga nic N in the upper 30 cm of soil was 29, 41, 57, and 54 kg ha 1 for manure application rates per year of 200, 400, 600 and 800 kg ha 1 total N respectively. All of the soils had mean nitrate N concentrations less than the critical pre sidedress nitrate N soil level for corn of 21 mg kg 1 When the nitrate N soil level is below the critical pre sidedress nitrate N soil level, a yield response to application of additional N is expec ted. Van Horn et al. (1996) found tha t a yield response to additional manure was expected in all soils due to soil nitrate N concentrations less than 10 mg kg 1 Van Horn et al. (1996) also found a pplying small amounts of manure over the crop growing seaso n did not result in soil N accumulation. Bacon et al. (1990) used a nutrient b alance to estimate soil N content. Calcula ting N inputs such as manures, fertilizer and residual N and N outputs in crops harve sted left a positive balance of total N which was account ed for by biological N fixation or soil N ac cumulation. Amounts of N found in precipitation, seeds, nonsymbiotic N fixation, and residual manure decomposition, and nutrient output by leaching or denitrification were not included in the nutrient balance calculations They found an accumulation of P and total N and a net depletion of available N i nutrient balances varied widely. They concluded there were advantages to record keepin g on individual field s as opposed to the whole farm.
24 Dou et al. (1996) used an equation to estimate the amount of residual N in the soil: N t = N f + (N 1 or N m ) where N t wa s the total soil residual N, N f wa s the residua l N credit from manure that was applied during the previous fall or winter, N 1 wa s the preexisting legume residues, and N m wa s the manure applied in the pa st excluding N f A sample dairy farm was selected for modeling with manure from 101 lactating cows and 89 heifers and 5500 kg of total N as manure from a contracted poultry section on the farm. Agricultural production included 24.3 ha of corn silage, 13.7 ha of sorghum sudangrass silage, 18.5 ha of alfalfa and grass hay, and 27.3 ha of ryelage planted as a winter cover crop. Using the Cornell Net Carbohydrate and Protein System, Dou et al. (1996) predicted total soil N reserves for a Pennsylvania dairy farm to be 424 kg year 1 based on the residual fecal N expected to be available in the subsequent three years. In New York, Klausner et al. (199 8) used an estimate of residual N from crop residues (alfalfa) and a pre sidedress nitrate N soil test for cor n to determine the soil available N content and predicted an additional 43 kg ha 1 N fertilizer needed for corn crops. Constantin et al. (2010 ) co llec ted soil samples three times during the year to determine soil inorganic N content change due to the catch crop treatments: white mustard, Italian ryegrass, and radish at Boig neville, Kerlavic, and Thibie, France, respectively Samples were col lected t o 90 cm at Boigneville and Kerlavic At Thibie, data from previous soil samples collected in 2003 to 110 cm was used Organic soil N was measured from soil samples to a depth of 60 cm and divided into 0 15 cm, 15 30 cm, and 30 60 cm sections. S amples for organic N determinations were taken less frequently than for inorganic N only 5 times over 2 years. Soil mineral N was found to be dependent on the previous crop grown, the climate, and the dates of sampling. At
25 harvest of the main crop the soil nitr ate N ranged from 29 to 58 kg ha 1 Catch crop treatments decreased the soil mineral N content in late autumn and mid winter. Catch crops had a significant effect on soil organic N with a mean annual N increase of 11.9, 24.2, and 22.2 kg ha 1 year 1 at Boig neville, Kerlavic, and Thibie, respectively. The soil organic N in the 30 60 cm layer d id not differ significantly between treatments (including catch crop treatments and no catch crop ) Soil organic N in the lower 30 60 cm layer contained a small portion of the N in the sampled soil profile. Soil mineral N content is a source of N available to agricultural crops and therefore should be measured and included as an input in the nutrient mass budget. Van Horn et al. (1996) and Klausner et al. (1998) used pre sidedress nitrate N soil testing to measure soil nitrate N, one component of soil mineral N. Both studies found soil nitrate N in the range of 29 to 58 kg ha 1 which indicated that additional N fertilization was needed for corn crops. Dou et al. (1996) sho wed manure applications to crops make estimating the soil mineral N input more difficult due to mineralization of organic N over the crop season. Soil mineral N content fluctuates over the crop year due to mineralization, immobilization, N fertilization, N leaching, and crop uptake. Crop Uptake The N content of the crop is an imp ortant component of the N mass budget as it is preferred output of N from the field. The N content of the crop is typically related to crop yields and it is used to for m the proteins and amino acids that will affect the nutrition al value of forage crop s as animal feed. Some w hole farm nutrien t mass budgets do not account for crop uptake o f nutrients because crops are used as animal feed. In these cases, t he nutrients ar e cycled within the farm, because crops are not sold off farm.
26 Jokela (1992) sampled grain and total silage yield for corn by harvesting 6 m of row (3 m each from the center 2 rows) at a nearly mature stage (milk line advanced one third to two thirds down kernel) and analyzing the plants using standard Kjeldahl methods (Bremner and Mulvaney, 1982) He used a split plot experimental design with manure effluent as the main plot (0 and 240 kg ha 1 N) and N ferti lizer treatments as the subplots (0, 56, 112, and 168 kg ha 1 N). Jokela (1992) found that total dry m atter yie ld ranged from 7.8 Mg ha 1 with the control treatment receiving no manure or fertilizer to 16.6 Mg ha 1 with the treatment combination of 112 kg ha 1 N synthetic fertilizer plus 240 kg ha 1 N from manure. In Tifton, Georgia, Van Horn et a l. (1996) studied spring corn silage N uptake using a variety of flushed manure N applications through a center pivot Annual N appli cation rates ranged from 240 t o 986 kg ha 1 Corn silage N uptake ranged from 49 kg ha 1 N harvested from 240 kg ha 1 N application to 265 kg ha 1 N harvested from 798 kg ha 1 A ha 1 N with 240 kg ha 1 N application to 249 kg ha 1 N with 739 kg ha 1 N application Mean N concentrations as a percent of dry matter ranged from 0 .90 to 1.20 % for silage corn vegetative forage, 1.57 to 2.0 % for corn grain, and 2.25 to 3.01 % for rye with annual dairy manure applications of 229 to 751 kg ha 1 N respectively The greatest N application resulted in the highest N concentration for each plant type. Eghball and Power (1999) measured N uptake of summer corn crops in Nebraska in a split plot experimental desig n with two tillage systems an d three fertilizer tre atments plus a control with no fertilizer application The three fertilizer treatments were composted or non composted beef cattle feedlot manures and commercial fertilizer.
27 Composted manure averaged 8.5 g kg 1 total N and non compost ed manure averaged 11.7 g kg 1 total N over the four year study period (1992 1995). The control plants receiving no fertilizer had total N uptake of 50 kg ha 1 N. The commercially fertilized plants resulted in the maximum total N uptake of 118 kg ha 1 N. The manure and composted manure treatment plants resulted in N uptakes of 89 and 87 kg ha 1 N, respectively. Plants from composted manure applications had similar N uptake to plants from non composted manured plots. Plants from fertilized plots had the greatest N uptake compared to plants from the composted and non composted manure plots due to greater N availability to plants. Bacon et al. (1990) studied nutrient flow on a whole farm basis as well as on an individual field boundary. For all fields, the authors found that for corn with inputs of 261 kg ha 1 year 1 t otal N the corn crop uptake output was 235 kg ha 1 y ear 1 total N for 1985. In 1986, t he corn input was 157 kg ha 1 y ear 1 total N an d the corn crop uptake output was 131 kg ha 1 year 1 total N Differ ing rates of manure application were determined by farm management and related more to livestock production and the type and capacity of manure storage than crop requirements. The differences in rates of manure application led to differences in r ates in N application on a field by field basis. This resulted in some crops being N deficient and some crops having high positive N balances and so possible N leaching. Crop uptake of N was shown to be dependent on N application by Bacon et al. (1990), Jo kela (1992), Van Horn et al. (1996), and Eghball and Power (1999) Corn crops were evaluated by several methods of dry matter yield, crop N uptake, and N
28 concentration of the corn plant. Past researchers found c orn silage N crop uptake varied from 49 to 26 5 kg ha 1 N depending on N application amount. Leaching W ater moving down through the soil profile carries soluble nutrients. Nutrients leached out of the root zone are no longer available to plants and p ose possible environmental risk (Van Horn et al., 19 96). N leaching to ground water is an output of the N mass budget that is undesirable to the dairy farmer. Not only is the N leached no longer available to the crop and represents an economic loss but it also is a pollutant to ground water. FDEP permitted dairy farms must implement farming practices in order to maintain a concentration of nitrate N in groundwater below the MCL of 10 mg L 1 nitrate N Leaching is difficult to measure since it is dependent on the soil hydraulic conductivity as well as cropping p atterns and soil moisture conditions due to rainfall, temperature, evaporation, and nutrient additions (Hutson et al., 1998). When the water holding capacity of the soil is exceeded, the soil water percolates and nutrients in the soil water are s ubject to leaching, but the water content of the soil is highly variable and continuously changing and so it is difficult to measure over time When the soil water moves it carries with it any N it contains in organic and inorganic forms deeper into the s oil profile and eventually to the ground water. A tool for determining leaching under fi eld conditions in situ is the drainage lysimeter The drainage lysimeter has a basin buried below the root zone within the soil profile which collects soil water moving down through the soil profile. The soil water carrying nutrients (leachate) drains into a reservoir where the leachate can then be sampled (Gazula et al., 2006). N loads for the lysimeter can be calculated by multiplying the N concentration by the volume for each sampling of leachate. The load can be
29 expressed on the basis of the cropped area using the load calculated from the soil area above the lysimeter. Duan et al. (2010) installed lysimeters with a 0.2 m diameter and a depth of 0.46 m, flush with the soil surface The lysimeters were filled with coarse gravel, fine gravel, and sand ( to ensure the ability to pump leachate ) and 0.3 m of undisturbed soil to mimic the natural setting. The lysimeters were sampled monthly for leachate using a PVC pipe instal led along the inside wall of the lysimeter. Leachate was analyzed for total N nitrate N, and ammo nium N within 24 hours of collection. Their lysimeters were relatively small, using 19 L buckets as the basin to catch leachate. Other researcher s have used l arger cylindrical lysimeters 49 cm in diameter and 70 cm deep, with the top edge placed at the soil surface (Silva et al., 2005) For long term studies of field conditions, lysimeters are buried below the root zone and deep enough to provide for normal til lage practices. Drainage lysimeters must be carefully installed so that the soil profile is returned with minimal mixing of the soil horizons and so that normal crop production can proceed above the lysimeter. Van Es et al. (2006) installed lysimeters 1.8 m deep with a central drain line and an acces s hole to allow for long term sampling of drainage water. Constantin et al. (2010) studied leaching and soil N contents over 13 to 17 y ears at Boigneville, Kerlavic, and Thibie in n orthern France. Suction lysime ters and drainage lysimeters we re installed to measure N concentration and the amount of nitrate N in the percolating water (load), respectively due to catch crop treatments Nitrate N measurements were made infrequently only 4, 9, and 4 times per year o n average at the three sites. The N leached per year was calculated based on nitrate N concentrations from the suction lysimeters and the drainage amounts from drainage lysimeters. N l eaching varied from
30 0 to 138 kg ha 1 year 1 depending on climate, site, crop type, drainage intensity and management practices. Catch crops reduced N leaching by 9, 32, and 19 kg ha 1 year 1 N, at the three respective experimental locations and their use appeared to be a very efficient way to decrease leaching losses of N ( Constantin et al., 2010). Leaching is often not directly measured and instead is estimated based on the difference in inputs and measured outputs. Jokela (1992) measured nitrate N concentrations in the soil solution, but not leaching load. Nitrate N concen trations in the soil pore water below 0 9 m were less than the national MCL for drinking water of 10 mg L 1 nitrate N and therefore significant nitrate N loading of ground water most likely did not occur. Hall and Risser (1993) estimated leaching as 38% of the total outputs of their N mass budget. Wang et al. (2000) conducted a whole farm mass bala nce and found remaining N (the difference in exports and inputs) to be 61.5% of the total budget. N leached to ground water would be included in thi s estimate of remaining N Haas et al. (2007) included losses through milk cash crops and animals as outputs for the whole farm nutrient mass budget but did not include leaching Dou et al. (1996) used the Cornell Net Carbohydrate and Protein System model to create a w hole farm balance for a sample dairy farm with 101 lactating dairy cows and 89 heifers. T otal inputs of 19,8 80 kg of N, total outputs of 8,44 0 kg of N, resulted in a difference of 11,4 40 kg of N which included various N losses such as leaching, run off, de nitrification, and volatilization. Although dairy farmers should strive to minimize leaching, Magette et al. (1989) confirmed the difficulty of managing leaching even when recommended nutrient management practices are followed. Weather conditions and annua l variations in crop production resulted in frequent discharges which exceeded the national standard of 10
31 mg L 1 of nitrate N to groundwater. Some N loading of ground water is inevitable, but farm management of irrigation in conjunction with rainfall and manure application can minimize s pikes in nitrate N concentration reaching the ground water. Direct measurement of leaching losses helps improve the accuracy of N mass balances. Volatilization The use of manures as fertilizer provides N in the two primary forms of am monium N and organic N. The amine groups fro easily converted to a gaseous NH 3 Most manures, lagoons and feed lot soil surfaces have a pH greater than 7 which results in a scarcity of hydrogen ions and prevents the conversi on of ammonia to NH 4 (Van Horn et al., 1994). Volatilization is highly variable and is affected by temperature, moisture content, pH, air movement, and possibly other factors (Van Horn et al., 1996) Organic constituents of manure degrade leading to further volatilization of ammonia thus enhancing the loss of N as ammonia gas (Van Horn et al., 1996). N losses from animal manures through volatilizatio n can reach 50 to 75%, most often from NH 3 lost to the atmosphere (Van Horn et al., 1994). Because volatilization is very difficult to measure, many researchers do not attempt to quantify volatilization directly. Instead volatilization can be treated as one possible pathway for N losses when a positive bal ance of nutrients is found (Bacon et al., 1990). Volatilization estimates are normally made to quantify potential losses. Bacon et al. (1990) estimated a N availability factor of 20 to 50% for the total N in manure applied to fields taking into account vol atilization losses and N unavailable to plants T hese estimates a re from previous research cited by Bacon et al. (1990). Klausner et al. (1998) did not measure volatilization directly during their research o n a dairy farm in New York, but co ncluded major l osses of N to volatilization due to a N
32 deficit in crops even though there was an apparent surplus of total N in manure applied. There was a long delay between application of manure and its incorporation in the field and so all ammoniacal N was assumed to be lost to volatilization. Haas et al. (2007) did not calculate volatilization for their farm nutrient budget, but instead made assumptions of volatilization outputs on single farm balances. Mean N surplus of 43 kg ha 1 N included denitrification and nitra te leaching which were not measured by Haas et al. (2007). Hall and Risser (1993) estimated volatilization to be 25% of outputs of N from their 55 acre site in Pennsylvania. The Cornell Net Carbohydrate and Protein System model (Dou et al., 1996) considers two major factors for manure transformations: mineralization of organic N and volatilization of ammonia N. The model assumes that the urine component of manure becomes inorg anic during manure collection and storage. Conditions to model volatilization during manure storage vary so widely due to temperature, pH, and loading rate that only the type of manure management facility (in ground pit, aboveground metal tank, earthen mou nds, etc.) is considered in the model. Several published papers reported estimate d losses via ammonia volatilization in dairy manure as affected by manure management system. For anaerobic lagoons, Gilbertson et al. (1979) estimated 75% N los s, whereas in a erobic lagoons or oxidation ditches, losses were estimated at 44%. Pennsylvania Department of Environmental Resources ( PSAG, 1986) estimated liquid and solid manure held in uncovered, watertight structure s to have losses of 30 40%; liquid and solid manure in a pond, agitated before spreading was estimated to have similar losses. Sever al other management facilities we re also used in the model but not cited. T he Cornell model uses the amount of available N after mineralization, the
33 expected volatilization ba sed on the manure management facility, the predicted ratio of organic and inorganic N of the stored manure and application method to determine N availability in the manure w hen making fertilizer recommend ations for manure application to plants (Dou et al. 1996). Kuipers et al. (1999) outlined the European guidelines for reducing ammonia volatilization 50 to 70%. One recommendation was for the application of manure slurry by injection into the soil which can reduce volatilization at the farm level by almost 50% compared with surface application (Kuipers et al., 1999) European guidelines recommended restricting the application of slurry to during the growing season when potential for volatilization would be diminished. Slurry storage facilities buil t after 1987 are required by Dutch policy to be covered (Kuipers et al., 1999). Eghball and Power (1999) estimated 50% total N losses mostly from volatilization from beef cattle feedlot manure by the time the manure is collected. There was no significant d ifference in N volatilization losses between surface application and incorporation of beef feedlot manure by conventional tillage when applied to fields (Eghball and Power, 1999). Beef cattle feedlot manure contained mostly organic forms of N and only smal l concentrations of ammonium N which is subject to volatilization. Van Horn et al. (1994) estimated that gaseous losses of volatilization and denitrification should be estimated at greater than 50% of the N in original manure excretions; less than 50% of N should be assumed availab le for crop uptake once manure wa s applied to fields. Overcash et al. (1983) found that in dairy, swine, or poultry ma nure, surface applied manure lost up to 50% of the total N in the ammonium N form
34 In constructing nutrient mass budgets, Van Horn et al. (1994) Bacon et al. (1990) Klausner et al. (1998) Haas et al. (2007) Hall and Risser (1993) and Eghball and Power (1999) estimated volatilization instead of directly measuring the losses. Most estimates of volatilization loss es were about 50 to 75% of N manure application. Mineralization Within the N mass budget, transformations of N over time make quantifying some sources of N difficult. Volatilization and mineralization are the two major transformations of N in manure. When manure is applied to crops as a fertilizer, these transformations play a major role in N crop availability. The inorganic ammonium N fraction of N in manure is subject to losses through volatilization. The organic fraction of N in manure is subject to mine ralization which is the process by which microbial populations in the soil break down organic N into inorganic N. Mineralization allows more inorganic N to become available to crops, but also leaves it subject t o losses (Dou et al., 1996). Mineralization n ot only occurs with organic N in manure, but also organic N of plant residues such as stubble and roots left behind from harvested crops. Mineralization occurs over time and is dependent on many factors such as the ditions, and management practices. The initial nutrient composition of manure varies with factors affecting the cow such as feed, water (Azeez and Van Averbeke, 2010). Different types of storage and application practices also affect the nutrient content of the manure before it is applied to the soil (Azeez and Van Averbeke, 2010). It is therefore very difficult to make recommendations for manure application to fulfill crop requirements. Measuring mineralization in field condit ions is the best way to make predictions for N availability from manure applications.
35 Mineralization rate can be measured in situ by many methods including the litterbag (or buried bag) method, the open ended Polyvinyl Chloride (PVC) tube method, and the s mall sheltered soil method. The litterbag method uses nylon mesh bags filled with manure buried in the soil. O ver time, the bags are removed for tests for organic matter decomposition and N release (Cusick et al., 2006). The PVC tube method uses pipes driven into the ground to trap a given amount of soil which is then incubate d in situ. Tubes are removed from the soil at intervals over the growing season for N determinations (Wienhold, 2007). The small shel tered soil method uses a shelter around an area of soil that experiences the same temperature, moisture and weather conditions as the bulk field soil. Soil is then sampled from under the shelter and evaluated for mineralization of N by analyzing the soil f or mineral and organic N constituents (Mikha et al., 2006). Mikha et al. (2006) found that no tillage management of fields resulted in conservation of added organic material which resulted in organic N released by mineralization in later years. Mikha et al (2006) used modeling to estimate N mineralization in addition to in situ sampling to measure actual N mineralized. T he model generally overestimate d N mineralization. It also did not include environmental factors such as soil water content and soil tempe rature. Azeez and Van Averbeke (2010) used a lab incubation study to measure ammonium, nitrate, and total mineral N over time. Soil was mixed with manure in 400 mL plastic containers and kept the soil moisture at field capacity in a dark cupboard at 23C f or 120 days. They found that there was a gradual buildup of ammonium N after 20 days, but then it declined markedly during the remaining 100 days of the experiment.
36 Total mineral N content decreased due to net immobilization after 10 days, then increased t o peak at 358 mg kg 1 total N a fter 55 days. Net immobilization due to an increase in reproductive rates of the microbes and therefore high competition for N led to a decrease in mineral N at 70 to 90 days of incubation. As nutrient release decreases, micr obes die off and decomposition of the microbial mass led to a soil mineral N net increase peaking at 19 1 mg kg 1 total N after 120 days. Immobilization of N by microbial populations makes the N unavailable to crops, but fluxes in microbial population tend to happen rapidly (Azeez and Averbeke, 2010). Cusick et al. (2006) found results similar to those of Azeez and Averbeke. In an incubation trial using dairy manure, during the first 21 days there was a net mineralization of N, followed by an immobilization from day 21 84, then a mineralization from day 84 to the end of the trial at 168 days. Overall, a low net mineralization from manure was found, but this was most likely due to the initial high level of total N present in the soil (Cusick et al., 2006). In the majority of mineralizable N was released from the litterbags. In the first year of the experiment by day 21, 52.6% of the mineralizable pool had been converted; in the second year, 85.7% of the mineralizable po ol of N had been mineralized by day 21. Over the 148 day experiment, an average of 67% of the total manure N was mineralized. The literature indicates there is a lack of studies compiling N mass budget s by direct measurement of all N inputs and outputs. Es timations are often used to account for leaching instead of direct measurement. This study aims to directly measure soil N inorganic fertilizer, crop uptake, and leaching while performing mineralization
37 experiments to estimate mineralized N Direct measur ement of N pools yields a greater understanding of the N mass budget and potential N losses. Objectives The obje ctives of the study were: 1) q uantify a N mass budget in order to evaluate the N balance of Field J at the University of Florida Dairy Unit so t hat determinations can be made of the most likely sources of N losses and 2) m ake recommendation s to the University of Florida Dairy Unit to amend their management practices that will in turn reduce N losses especially nitrate losses to leaching. The hypotheses of this study were: 1) f ertilizer N application to the fields is greater than the IFAS reco mmendation due to overestimated gaseous losses. Farm management assumes 40% losses (Martin, 2000) This leads to an over fertilization of the crops and leaching of N to groundwater ; 2) l eaching losses of N are mostly associated with rainfall after the crop is harvested and not to irrigation during the growing season ; 3) m anure application late in the growing season increases the likelihood of leaching losses in the fallow season ; and 4) m ineralization of soil organic matter accounts for less than 20% of the N in the total crop N inputs.
38 CHAPTER 2 MATERIALS AND METHOD S Dairy Unit Site Description A N mass budget study for a silage corn production system was conducted during the 2011 2012 crop year n Hague, FL. A 14.16 ha field, Field J (Figure 2 1), with a center pivot irrigation system, was selected for the study. Planting schedules and irrigation system were typical of farming practices for cultivated fields at the Dairy Unit and for commercial da iries in Florida. Field J was planted with two crops of silage corn ( Zea mays L. ) in the spring and summer of 2011, and a cereal rye ( Secale cereale L. ) and ryegrass ( Lolium multiflorum Lam. ) mixture during the winter of 2011 2012. The crops received manur e effluent and fresh water through the center pivot irrigation system. The soil in Field J is predominately Chipley fine sand (thermic, coated, Aquic Quartzipsamments) with less than 5% of the field Tavares fine sand (hyperthermic, uncoated, Typic Quartzip samments) in the northwest corner (NRCS, 2010) (Table 2 1). The soil is somewhat poorly drained with a slope of 0 5% (NRCS, 2010) and 1.9% organic matter. The mean soil pH (1:2 soil:deionized H 2 O ratio) was measured to be 7.5. All farming practices includi ng crop varieties, plant spacing, pest control, and application of inorganic fertilizers and manure effluent were determined by farm management according the IFAS guidelines for silage corn production (Mylavarapu et al., 2009; Wright et al., 2011 a ). Corn v for the summer season. Silage corn was planted in row s 76 cm apart. For the winter, a mixture
39 ryegrass (Ragan and Massey, Inc., Ponchatoula, LA, USA) was planted on 28 Oct. 2011. The planting was by grain drill with a seeding ra te of 90 kg ha 1 for rye and 17 kg ha 1 for annual ryegrass. The Dairy Unit at the University of Florida is classified as a m edium Concentrated Animal Feeding Operation (CAFO) because animals, feed, manure and urine, dead animals, and production operations are on a small land area, the feed is brought to the cows at the Dairy Unit rather than animals grazing in pastures, and t he dairy maintains an annual average of 200 to 699 mature dairy cows. Because the Dairy Unit is classified a s a CAFO, it is required by the EPA to submit a Nutrient Management Plan ( NMP). The NMP includes best management practices (BMPs), conservation practices, and management activities to ensure agricultural production goals are met as well as defining soil an d water conservation goals to reduce threats to water quality and public safety. The NMP must address manure storage, animal mortality management, clean water diversions, prevention of direct animal contact with water, chemical handling, conservation pract ices to control runoff, manure and soil testing protocols, land applicati on protocols, and record keeping requirements. The NMP for the Dairy Unit selected N as the controlling nutrient based on the on (Martin, 2000) Therefore, P would be less of a concern than N. The manure produced by the approximately 575 cow milking herd (approximately 186 kg day 1 N) travels through a manure solids separator and then a three stage waste storage pond system. The NMP assumed that approximately 72 million liters of manure effluent from the final stage waste storage
40 pond is applied to Field J through the wastewater application system each year and that 40% of the N in the applied wastewater will be lost due to volatiliz ation, denitrification, and leaching. Crop land requirements for the application of manure effluent (207 million liters applied annually) are approximately 75 hectares of the 136 hectares available at the Dairy Unit for application. The NMP also stipulates that neither manure effluent nor fresh water will be applied to fields unless a crop is actively growing (Martin, 2000) Nitrogen Mass Budget The following equation was used to compute the N mass budget for the study: (Soil Initial + Mineralized N + Inorg anic N Fertilizer + Atmospheric Deposition) (Soil Final + Harvested Crop Uptake + Root and Stubble Uptake + Leaching) = Unaccounted for N Soil Initial and Soil Final were the soil mineral N content (nitrate N plus ammonium N) of the soil at the beginning and end of the growing season for each crop. The inputs, outputs, and balance of the N mass budget are shown in Table 2 2. The inputs of the N mass budget were Soil Initial mineral ized N, inorganic N fertilizer, and atmospheric deposition. The outputs of the N mass budget were Soil Final harvested crop uptake, root and stubble uptake, and leaching. The balance of the N mass budget was unaccounted for N Runoff was assumed to be zero, because the slope of the field was less than 5% (NRCS, 2010) and no signs of runoff were observed in the field during any season. Amounts of manure effluent from the lagoon and inorganic fertilizer applications for each season were determined from farm records. The National Atmospheric Deposition data from the Bradford
41 due to its proximity to the Dairy Unit. Atmospheric N deposition was calculated using the season. For th e spring season, mineralized N was estimated from the crop available ammonium N content of manure effluent samples taken during the spring season For the summer and winter seasons, m in eralized N was estimated from N mineralization experiment s conducted du ring the summer 2011 and winter 2011 2012 crop seasons. The N mass balance was calculated at the end of each season by subtracting the outputs from the inputs. The unaccounted for N in the N mass balance (inputs minus o utputs) was assumed to consist of gas eous losses from denitrification and volatilization. Sampling Locations in the Field Six locations, lettered A through F, were selected in Field J as shown in Figure 2 1 for installation of lysimeters to measure leaching. The six lysimeter locations were c hosen to representative ly sample leaching in the field and to select sites which varied in the spatial distance from the center of the pivo t. It was assumed that there would be variation in the spray field distribution of the pivot and so lysimeter locatio ns were distributed along the three pivot wheel tracks. The lysimeter locations were also selected with the criteria that they be spread evenly throughout the field with three sites in the north half of the field, three sites in the south half, equally dis tributed from east to west. Once the general locations of the lysimeters were determined, the exact locations of the lysimeters were chosen randomly during installation and adjusted to avoid buried irrigation piping and electrical lines.
42 For soil and crop sampling, six additional locations were randomly selected for sampling on each sampling date in addition to sampling at the six lysimeter locations. A diagram of the site selection is shown in Figure 2 2. The six additional locations were selected by divi an area was selected by dividing the parcel in half, flipping a coin to select either the hal f of the parcel was selected, the area was then divided into two equal portions and a for soil and crop sampling, this procedure was repeated for each sixth of the field to select six additional locations to sample plus the lysimeter locations. GPS coordinates were taken for all locations. Directly Measured Components of Nitrogen Mass Budge t Soil sampling and analysis Soil cores were taken on 14 Apr. 2011, 1 July 2011, 29 Sept. 2011, and 16 Mar. 2012 to determine the Soil Initial and Soil Final values for the N mass budget. Soil was sampled using a bucket auger to a depth of 60 cm in two inc rements, 0 to 30 cm and 30 to 60 cm. Soil sampling depth extended below the silage corn root depth of 40 cm (rye ryegrass root depth was 25 cm) in order to include all soil N available for crop uptake in the root zone. The crop root depth s were determined by excavating soil to a depth below which no roots were observed in the soi l excavation face, washing the soil from the face and o bserving where rooting extended On 14 Apr. 2011, one soil core was collected at each of the lysimeter locations A through F
43 method referenced above, for a total of 24 soil samples once divided into 0 to 30 cm and 30 to 60 cm depths. On the remaining sampling dates, three soil cores were collected at each soil cores and 72 soil samples. Because it was difficult to precisely excavate 30 cm increments with the bucket auger, the soil sample depths were measured with a tape measure afte r each sample was taken and the actual soil sample depth was recorded for use in calculating soil bulk density and mass of soil for the sampling depth on a h ectare basis. The entire mass of soil was collected, the wet weight was measured, and soil was refr igerated at 4 C during storage. Soil samples were typically stored for less than one week. A known volume of each soil sample was taken for bulk density measurements and a second subsample taken for analytical N measurements. The bulk density samples were weighed and oven dried at 105C for at least 24 hours or until a constant weight was reached. The oven dried soil was weighed and bulk density was calculated by dividing the total oven dried soil sample weight by the calculated volume of the hole dug by t he bucket auger using the noted actual depth. The subsamples taken for N analytical measurements were air dried at 38C and sieved through a 2 mm screen. The air dried soil samples were analyzed for nitrate N, ammonium N, and total Kjeldahl N (TKN) using a utomated colorimetric analysis at the Analytical Research Laboratory at the University of Florida. Nitrate N and ammonium N concentrations were determined following KCl extraction. Fifty mL of 1 M KCl was added to 5 g of soil in a 125 mL polypropylene bott le, shaken for 30 minutes on a reciprocating shaker set at 180 reciprocations per minute. The solution was then filtered through a 15 cm Whatman
44 No. 42 filter paper into a 150 mL plastic cup. The solution was then poured into a 20 mL plastic scintillation vial; the remainder was discarded (Hanlon et al., 1996). The vials were refrigerated at 4C until analyzed by automated colorimetric analysis ( EPA Method 353.2. and EPA Method 350.1 (modified)) using the Alpkem Flow Solution I V (OI Analytical, College Stat ion, TX, USA). Soil analysis of TKN was determined by digesting 1 g of soil, 2 g of Kjeldahl mixture (Pope Kjeldahl Mixture, Inc., Dallas, TX, USA) and 5 mL of 18 M sulfuric acid in a 50 mL digestion tube on a preheated 250C aluminum block digester for 1 hour. After adding glass funnels on all digestion tubes, the temperature of the block digester was increased to 325C for an additional 2.5 to 3 hours. The sides of each digestion tube were washed with 5 to 10 mL of deionized water, tubes were mixed on a v ortex shaker, and then transferred to a 100 mL volumetric flask. After flasks were cool, the solution was diluted to volume, covered with parafilm, and mixed thoroughly. The solution was then filtered using a 15 cm Whatman No. 41 filter paper into a 20 mL scintillation vial; the remainder of the solution was discarded (Hanlon et al., 1996). The solution was analyzed for N by automated colorimetric analysis ( EPA Method 351.2) using the Alpkem Flow Solution I V (OI Analytical, College Station, TX, USA). Topsoil samples were collected on 14 June 2012 at each of the lysimeter locations. A soil probe was used to collect soil to a depth of 15 cm. At each lysimeter location, the soil probe was used to randomly collect 5 soil samples within 10 m of the lysimete r location. These soil samples were mixed thoroughly in a plastic bucket then sub sampled (Shober and Mylavarapu, 2009). The soil sub sample for each lysimeter location was analyzed for pH, organic matter, electrical conductivity, lime requirement,
45 and Meh lich 1 macro and micro nutrients including phosphorus, potassium, magnesium, calcium, manganese, zinc, and copper. The soil sub samples were analyzed at the Analytical Research Laboratory (ARL) at the University of Florida according to the methods describe d in the UF/IFAS Extension Soil Testing Laboratory (ESTL) Analytical Procedures and Training Manu al (Mylavarapu, 2002). Crop sampling and analysis Corn plant samples were taken on 21 June 2011 and 20 Sept. 2011 to determine crop N removal and root and stub ble N uptake of crops in the spring and summer seasons. Rye/ryegrass samples were taken on 17 and 20 Feb. 2012 to determine crop N removal and root and stubble N uptake of the winter season. The short battery life of battery powered clippers used for the r ye/ryegrass sampling prevented finishing sampling on a single day. Farm management indicated to the researcher the planned commercial harvest date and p lant sampl es were taken as close to the planned commercial harvest date as possible. Rainfall and schedu ling conflicts with hired harvesters delayed commercial harvesting during the winter season. Excavating the total silage corn root mass was difficult due to labor and time constraints. A preliminary experiment was therefore conducted to determine a samplin g method for estimating the total corn root mas rounded 25 cm shovel to dig up the main root mass in a circular area with a radiu s of 25 met root mass captured b The following procedure was used to determine the total root mass. The corn root ing depth was determined to be 40 cm by exc avating soil to a depth below which no
46 roots were observed in the soil excavation face (determined by washing the soil from the face and observing where rooting extended). A 76 cm x 76 cm x 40 cm sampling area was decided based on the corn root depth and t he 76 cm corn row spacing. The 76 cm width of the sampling area allowed for a ll the soil between the middle of adjacent row alleys to be collected. Within the 76 cm x 76 cm x 40 cm sampling area, first t he to sample the pl ll remaining soil and roots were then dug up and and root excavation within the 76 cm x 7 6 cm x 40 cm sampling area were repeated four times in Field J. All roots were dried at 40C and weighed. Root was determined to be that part of the plant below the soil surface but also included the brace roots below the soil surface. T he sum of the root mas s es sampled by and the remaining root mass dug up from the sample area were considered the tot al root mass found in the 76 cm x 76 cm x 40 cm area. The total root mass dry weight was compared to the root mass dry weight be 66% of the t otal root mass (standard deviation 18%) which was then used to ng Corn plant samples were harvested from a randomly selected 200 cm length of chosen randomly on the day of sampling. Plants were harvested by cutting stalks 25 cm
47 above the soil surface to mimic mechanical harvesting for silage. The remaining corn stalk from 25 cm above ground to the soil surface comprised the stubble. A rounded 25 described above. The harvested root mass was washed to remove soil from the roots and then the stubble, the part above the soil line, was cut from the main root mass. Corn samples were separated into leaves, ears, stalks, stubble, and roots. The total num ber of corn plants and sample fresh weight for the leaves, ears, and stalks was measured. The fresh weight of the harvested portion (leaves, ears, and stalks) was measured for comparison to farm records of silage corn yield. The rye ryegrass winter crop wa s sampled at the six preselected locations, A through F, and six random locations labeled S through X. The rye ryegrass mixture was sampled using a 0.25 m 2 quadrat ring that was randomly placed in the selected location to include two seeding rows of the rye/ryegrass mixture. Four samples using a 0.25 m 2 quadrat ring were taken at each location for a total of 48 samples. The rye/ryegrass mixture was clipped 10 cm abo ve the soil surface to mimic the height of the mechanical harvester. Rye ryegrass harvested sample fresh weights were recorded. Stubble samples, the remaining plant material above the soil surface, were collected at each of the forty eight 0.25 m 2 quadrat ring locations The rye/ryegrass root depth was determined to be 25 cm by excavating soil to a depth at which no more roots were observed, washing the soil from the soil face, and determining the deepest root extension. The rye/ryegrass mixture roots were sampled by digging up one quarter of the 0.25 m 2 quadrat ring location using a 25 cm rounded shovel. The quarter of the 0.25 m 2 quadrat ring was randomly selected. All of the soil and roots in the quarter of the
48 0.25 m 2 quadrat ring location were removed a nd washed through a screen to collect the roots. These measurements were multiplied by 4 to estimate the total root mass in the full 0.25 m 2 quadrat ring sample area. All plant samples were dried at 40C until a constant dry weight was measured A dry weig ht was recorded for all samples and the samples were ground to 2 mm using a Wiley mill to create a homogenous sample. The samples were analyzed for total N by combustion u sing an Elementar Vario Max CN A nalyzer (Elementar Americas Inc., Mt. Laurel, NJ). A sample of 250 to 255 mg ground plant tissue was measured into a crucible. Crucibles were loaded i nto the Elementar Vario Max CN A nalyzer for determination of total N by combustion. All samples were run in duplicate. A National Bureau of Standards tomato le af was run approximately every 25 to 30 samples to ensure the instrument was calibrated correctly. All sample measurements maintained a replication precision of 5%. Any samples not meeting this criterion were rerun. Leachate sampling and analysis Drainage lysimeters were used to capture leachate and estimate leaching for the N mass budget. Twelve drainage lysimeters were installed in Field J at the six lysimeter locations; two lysimeters were paired at each location shown in Figure 2 1. Lysimeters were paired at each location in order to provide a check for lysimeter leaching results at each location. Comparison of the paired leaching results was used to determine when a random spike in N load, as measured by a single lysimeter, did not actually re flect a high N leaching load for that particular location. Drainage lysimeters were constructed with a basin to collect the leachate and a reservoir. The basin was constructed from plastic 208 L capacity drum s cut in half lengthwise. The basin was 89 cm lo ng, 30 cm high, and had a diameter of 60 cm. A
49 drain at one end of the basin allowed leachate to drain into a 19 L reservoir. Pebbles covered the bottom of the basin to allow water to freely drain into the reservoir. A plastic screen was placed over the pe bbles prevented mixing of soil and pebbles. The basin was installed at a decline towards the drain. Water collected in the reservoir was sampled through Tygon tubing within a capped PVC pipe which connected the reservoir to the soil surface. A diagram of t he lysimeter design is in Figure 2 3. When the drainage lysimeters were installed, the soil profile above the intended location of the lysimeter was removed separately by horizon. The lysimeters were buried with the upper edge of the drainage basin at a de pth of 76 cm to prevent disturbance by farm tillage equipment during the study and also make sure the lysimeter basin was below the root zone. Placing the lysimeter basins below the root zone maintained natural leaching conditions by preventing plant roots from taking up leachate collected in the basin and only measuring leachate that had drained past the root zone. All of the removed soil was replaced by horizon into the basin to recreate the original soil profile and the soil was repacked to approximately the original bulk density. The drainage lysimeters captured leachate from an 89 cm by 60 cm area of soil profile. Drainage lysimeters were installed in Field J at the 6 lysimeter locations (Figure 2 1) on 4 and 10 March 2011. Because drainage lysimeters w ere installed less than 2 weeks prior to the silage corn spring planting, lysimeters were allowed to settle prior to the start of sampling. Lysimeters were pumped out 22 April 2011 to start the first sampling period for the spring. Water samples from the d rainage lysimeter reservoir were taken approximately every two weeks. Leachate was collected from the reservoir by using a ShopCraft Multi Use Pump (Part No. W1145) (Advanced Auto Parts, Inc.,
50 Roanoke, VA, USA) connected to the Tygon tubing to manually pum p leachate out of the reservoir. Leachate was allowed to flow into a 19 L bucket for a few seconds before 40 mL was captured in two 20 mL scintillation vials. One drop of 9 M sulfuric acid was added to each scintillation vial and vials were kept on wet ice to maintain a temperature less than 4C. An additional 20 mL scintillation vial of leachate was provided for 10% of the samples submitted for each analysis in order to fulfill lab quality control requirements. The rest of the leachate was captured in the bucket so that total leachate could be measured. The drainage lysimeter reservoirs were pumped dry at each sampling date. The time, lysimeter location, and volume of total leachate were recorded for each lysimeter. When lysimeter reservoirs were empty for any sampling date, it was recorded that no leaching occurred. Leachate was sampled according to the Environmental Protection Agency certifi cation guidelines (Autry, 2 003). Water samples were analyzed for nitrate N directly from the sample vial by automated colorimetric analysis (EPA Method 353.2) using an Alpkem Flow Solution IV (OI Analytical, College Station, TX, USA). For TKN analysis of water samples, 25 mL of water sample and 5 mL of digestion solution (Sulfuric acid mercuric sulfate potassium sulfate solution) were mixed in a digestion tube with a vortex mixer. After adding four to eight Teflon boiling chips to the digestion tube, tubes were placed in a block digester preheated to 160C for one hour. The temperature of the block was raised to 380C and tubes continued to heat for an additional 1.5 hours before tubes were removed and diluted with 25 mL of reagent water. Prepared solutions were analyzed for TKN by automated colorimetric analysis (EPA Method 351.2) using the Astoria Z Analyzer (Astoria Pac ific, Inc., Clackamas, OR, USA).
51 Water samples were analyzed at the University of Florida Analytical Research Laboratory (E72850) as National Environmental Laboratory Accreditation Conference (NELAC) certified water samples (EPA/600/R 04/003) Total N was the sum of nitrate N plus TKN. Nitrate N and TKN concentrations were multiplied by the volume of the total leachate collected to obtain the leached N load. Sampling date N loads were summed to quantify leaching for each crop season. Mineralization Experime nt Nitrogen m ineralization experiment s were carried out during the summer and winter growing season s on Field J at the University of Florida Dairy Unit. The experiment s were conducted to calculate the mineralized N input for the summer and winter seasons a s well as organic N is broken down into a plant available inorganic form. Because the primary source of N applied to crops at the Dairy Unit is manure effluent, correctly estimating mineralization of orga nic N in manure effluent is important to providing an optimal amount of plant available inorganic N to the crops Having a better estimation of mineralization in field conditions allows for farm management to improve decisions on the timing of manure efflu ent application. The PVC tube method (Wienhold, 2007) was used to measure N mineralization over three one month time periods during the summer and winter season. Because manure effluent was applied to Field J throughout the year, isolating soil over diffe rent time periods allowed the PVC tubes to capture varying amounts of previously applied manure effluent and soil N content. One month time periods also allowed for consistent intervals depicting the changes in speciation of N in the soil due to mineraliza tion.
52 PVC tubes approximately 30 cm long and 5 cm in diameter were used to isolate an area of soil. Holes (1 cm in diameter) in the sides of the PVC tubes helped maintain normal soil moisture conditions and prevented anaerobic conditions inside the tubes. Once the PVC tubes were installed in the ground, they were capped to prevent rainfall and manure effluent from entering the isolated soil. The PVC tubes were installed at the six preselected locations according to the schedule in Table 2 3. In the summer s eason, Time 0 was 24 July 2011, Time 1 was 24 Aug. 2011, Time 2 was 24 Sept. 2011, and Time 3 was 24 Oct. 2011. In the winter season, Time 0 was 8 Nov. 2011, Time 1 was 8 Dec. 2011, Time 2 was 8 Jan. 2012, and Time 3 was 9 Feb. 2012. The PVC tube caps were spray painted different colors to indicate which time period they were installed. In the winter season, the experimental protocol was improved and soil samples were taken at Time 0, Time 1, and Time 2 to account for the initial soil N content and N specia tion to use in conjunction with the measurements on the soil isolated by the PVC tubes over time. In the month preceding the installation of the first set of PVC pipes, 20 kg ha 1 TK N of manure effluent was applied to the rye ryegrass mixture in Field J. T he winter mineralization experiment captured 104 kg ha 1 TK N of manure effluent over the winter crop season ( Table 3 12 ). On each soil sampling date, three soil samples from the top 25 cm of the soil profile were taken at each of the preselected locations (Figure 2 1) by using the PVC tube as a soil probe. All soil samples were air dried at 38C and sieved through a 2 mm screen. The samples were analyzed for nitrate N, ammonium N, and TKN using the soil analysis methods outlined above by the Analytical Rese arch Laboratory at the University of Florida.
53 Data Analysis All data comparisons for each component of the N mass budget were analyzed separately using PROC NPAR1WAY, a non parametric Wilcoxon test due to non normal distributi ons of data (SAS Institute, 2 002 ). For the soil samples, the population and experimental unit were t he soil N content of Field J. The treatment was time. The mean and standard deviation of soil bulk density, mineral N, nitrate N, ammonium N, and TKN content were calculated for the 0 to 60, 0 to 30, and 30 to 60 cm soil profile on a kg ha 1 basis for the four sampling dates E ach season's initial soil N content was compared the final soil N content for soil mineral N, nitrate N, ammonium N, and TKN for the 0 to 60, 0 to 30, a nd 30 to 60 cm soil profiles The null hypothesis was the mean of the initial soil N content w as equal to the mean of the final soil N content. S ources of error would have been variation caused by uneven application of manure effluent or fertilizer, accide ntally sampling in the corn rows versus the alley, organic matter (corn stalks) that was not evenly distributed causin g differences in mineralization and immobilization rate s, soil texture, and soil structural differences in the fie ld impacting N transform ations, manur e infiltration, and losses. Soil b ulk density was also compared by sampling date, location, and depth (0 to 30 and 30 to 60 cm). The mean and standard deviation of soil bulk density was calculated for each soil sampling date. For the silage co rn samples, the population and experimental unit were the N content of the silage corn in Field J. The treatment was time. The N content of the silage corn plant parts (leaves, stalks, ears, stubble, and roots) the harvested crop uptake, and root and stubble uptake were compared between the spring and summer season. The null hypothesis was the mean of the N c ontent of the spring silage corn (each plant part separately, the harvested crop uptake, and the root and stubble uptake) was equal
54 to the mean of the N con tent of the summer silage corn (each plant part separately, the harvested crop uptake, and the root and stubble uptake). Sources of error were from variation caused by uneven availability of nutrients including mineral N for plant uptake, differe nces in water availability for the crop, impacts from pests and disease, soil texture, and soil structural differences in the field. For each sampling location, the N content of the harvested portion (leaves, stalks, and ears) was summed and converted to a kg ha 1 basis, then the mean and standard deviation were calculated. For the rye ryegrass samples, the population and experimental uni t were the N content of the winter rye ryegrass in Field J. The stubble and roots were summed for each sampling location to calculate the root and stubble uptake. The harvested crop uptake and root and stubble uptake were converted to a kg ha 1 basis. The mean and standard deviation for the harvested crop uptake and root and stubble uptake N content were calculated from the 12 replicates at 12 locat ions. Sources of error were the same as those of the silage corn samples. For leaching measurements, the N load was calculated for each leachate measurement and converted to a kg ha 1 basis. The N load for each lysimeter was summed for each season. The mean and standard deviation of the 12 lysimeter N loads were then calculated for each season. Comparisons between seasons were not made, because of seasonal differences in rainfall, manure effluent application, freshwater irrigation, crop grown, and length of the season. It was not reasonable to assume that the mean N loads for each season were equal.
55 Table 2 1. Soil particle distribution by soil series (NRCS, 2010) Chipley fine sand Tavares fine sand Soil Profile Sand (%) Silt (%) Clay (%) Sand (%) Silt (%) Clay (%) 0 30 cm 92 5 3 96 2 2 30 60 cm 92 4 4 94 5 1 Table 2 2. Inputs and outputs of N mass budget N Mass Budget Components Source Inputs Soil Initial Directly Measured Mineralized N Estimated from Mineralization Experiments Inorganic N Fertilizer Farm Records Atmospheric Deposition Estimated from NADP Rainfall Assumed Negligible Outputs Soil Final Directly Measured Harvested Crop Uptake Directly Measured Root and Stubble Uptake Directly Measured Leaching Directly Measured Runoff Assumed Negligible Balance Unaccounted for N Calculated from N Mass Balance Table 2 3. Installation and removal schedule for PVC N mineralization tubes for the summer silage corn crop and for the winter cover crop Time 0 w Time 1 x Time 2 y Time 3 z 15 tubes installed 10 tubes installed 5 tubes installed 5 tubes removed from T0 batch 5 tubes removed from T0 batch 5 tubes removed from T0 batch 5 tubes removed from T1 batch 5 tubes removed from T1 batch 5 tubes removed from T2 batch w Time 0 was 24 July 2011 (summer) and 8 Nov. 2011 (winter) x Time 1 was 24 Aug. 2011 (summer) and 8 Dec. 2011 (winter) y Time 2 was 24 Sept. 2011 (summer) and 8 Jan. 2012 (winter) z Time 3 was 24 Oct. 2011 (summer) and 9 Feb. 2012 (winter)
56 Figure 2 1. Preselected sampling locations in f ield J Figure 2 2. Selection method of six additional sampling locations For each sixth of the field, one additional sampling location was selecte d by the following method. Example: A) Field J was separated into sixths. The highlighted portion indicates the area where the first additional sampling location would be randomly selected. B) The first sixth was equally divided between the inner and outer portion. The outer portion was randomly selected by a coin toss. C) The outer potion was then equally divided into right and left sides. The right side of the outer portion was randomly selected by a second coin toss. The additional sampling location woul d then be randomly selected from this area by the researcher.
57 Figure 2 3. Drainage l ysimeter design
58 CHAPTER 3 RESULTS AND DISCUSSI ON Site Characterization The soil in the research field was mapped as two soil types (Table 2 1). Topsoil samples taken on 14 June 2012 from the upper 15 cm of the soil profile at the lysimeter locations (Figure 2 1) had an mean soil pH of 7.5, organic matter content of 1.90%, and electrical conductivity (EC) of 0.11 dS m 1 (Table 3 1). Mehlich 1 macro and micro nutrient analyses indicated high or very high concentrations of phosphorus (very high, 692 mg kg 1 ), potassium (high, 110 mg kg 1 ), and magnesium (very high, 314 mg kg 1 ) ( Mylavarapu et al., 2009) For calcium (1920 mg kg 1 ), manganese (19 mg kg 1 ), and zinc (10 mg kg 1 ), no additional fertilizer was recommended (Mehlich, 1953; Mylavarapu et al. 2009) (Table 3 1). According to the IFAS recommendations for irrigated corn, th ere may be a crop response to copper (Cu) application, but plant tissue testing was recommended before application of Cu (Wright et al., 2011 b ). No Cu fertilizer was added in addition to that already in the manure effluent. Nitrogen application was recomme nded based on the crop growth requirement. No lime was required due to the soil pH testing above the target pH of 6.5 (Mylavarapu et al., 2009). Soil bulk density was measured for the soil samples taken on 14 April 2011, 1 July 2011, 29 Sept. 2011, and 16 Mar. 2012 for use in the calculation of the N content in the upper 60 cm of the soil profile (Table 3 2). Soil bulk density was significantly lower (P < 0.0001) in the upper 0 to 30 cm (1.5 g cm 3 ) of the soil cores than in the lower 30 to 60 cm (1.6 g cm 3 ) (Table 3 3). Soil bulk density was reported to increas e with depth by Hillel (19 80). Soil bulk density did not change over the cropping year (Table 3 2). The n o till farming practices used at the Dairy Unit resulted in bulk densities within the
59 typical range of 1.3 to 1.8 g cm 3 for cultivated soils ( Mattos et al., 2003; Brady and Weil, 2008 ; Zotarelli et al., 2009 ; Constantin et al., 2010 ; Frazao et al., 2010) Spring Season N Inputs Silage corn was planted on 15 March 2011 at which point manure effluen t and inorganic N fertilizer were applied to supply N to the crop. The starting point for calculating the spring season N mass budget was marked by the initial soil mineral N measurement on 14 April 2011. Atmospheric deposition of N also contributed to the N pools in Field J. I nitial soil N content The initial soil N content for the spring season was measured on 14 April 2011 within the 0 to 60 cm soil profile. The initial soil mineral N content was 28 6 kg ha 1 N (Table 3 4) comprised of 161 kg ha 1 N in the 0 to 30 cm soil profile and 125 kg ha 1 N in the 30 to 60 cm soil profile (Table 3 5). In the 0 to 60 cm soil profile, the soil mineral N content was made up of 113 kg ha 1 nitrate N ( Table 3 6 ) and 17 3 kg ha 1 ammonium N ( Table 3 7 ). Soil TKN ( organic N plus ammonium N) was 33 1 0 kg ha 1 N ( Table 3 8 ) at the start of the spring season, but the organic N portion (31 40 kg ha 1 organic N) organic N was not readily available to provide N to the silage corn until further N mineralization. The soil TKN contained organic N from previous manure applications, plant residues, and N immobilized by microbes. The initial concentration of TKN in this soil was 373 mg kg 1 TKN. Similar concentrations of soil TKN (347 mg kg 1 TKN) were found on e ntisols in southern Spain by Cabrera et al. (2005 b ). Th at study, conducted on a sandy soil similar to the Dairy Unit, concluded that coarser textured soils have lower fertility when
60 compared to the soil TKN concentration of sandy clay loam e ntisols with 536 mg kg 1 TKN. Soil nitrate N in the root zone (0 to 30 cm) was significantly greater ( P < 0 .0001) than the soil nitrate N in the lower 30 to 60 cm of the soil p rofile, for all sampling dates ( Table 3 9 ). Jokela (1992) found soil nitrate N decreased with soil depth and that soil nitrate N was proportional to the amount of N input from f ertilizer and manure. This researcher also found the vertical distribution of s oil nitrate N suggested some movement of soil nitrate N into the bottom increment of the soil profile. It is difficult to determine if the soil nitrate N distributi on in the soil profile indicated prior leaching, because the pre fertil ization soil nitrate N content wa s not known. An increase of soil nitrate N in the lower soil profile over the season would in dicate that nitrate N moved downward in the soil profile which ultimately could result in leaching. Soil ammonium N content did not differ significantl y with depth for any sampling dates ( Table 3 10 ). Soil TKN content in the upper 0 to 30 cm was significantly greater ( P < 0 .0001) than the soil TKN content of the lower 30 to 60 cm for all sampling dates ( Table 3 11 ). Soil nitrate N and soil TKN content de creased with depth whereas ammonium N was not significantly different in the lower soil profile. N itrification could have prevented an accumulation of soil ammonium N in the upper 0 to 30 cm soil profile. Altom et al. (2002) studied soil N distributions o n a Minco fine sandy loam in Oklahoma and found surface accumulation of ammonium N in the 0 to 15 cm soil profile with 336 and 448 kg ha 1 N rate of ammonium nitrate fertilizer treatme nts. At lower depths, these researchers found no differences in ammonium N due to fertilizer
61 treatments and concluded that movement of ammonium N into the profile was limited and was not a leaching threat. That these researchers used inorganic N fertilizers as treatments instead of manure efflu ent applications, may explain why their results ( ammonium N buildup in the surface of the soil profile and the lack of ammonium in the lower profile ) differed from the results at the Dairy Unit. In Boigneville and Kerlavic, France, Constantin et al. (2010) observed a lower range, 19 to 68 kg ha 1 nitrate N, in the 0 to 90 cm soil profile than the nitrate N observed at the Dairy Unit in the 0 to 60 cm soil profile, 113 kg ha 1 nitrate N ( Table 3 6 ). Zotarelli et al. (2009) measured nitrate N concentrations i n the 0 to 30 cm soil profile on spring sweet corn plots with a cover crop previously grown ov er the winter season. These researchers found initial nitrate N concentrations from 2 to 9 mg kg 1 nitrate N, which was less than the nitrate N concentration (17 mg kg 1 ) observed at the Dairy Unit in the 0 to 30 cm soil profile. In Mato Grosso, Brazil, Frazao et al. (2010) measured concentrations of nitrate N and ammonium N in the 0 to 20 cm layer over several land uses (native, conventional till, pasture, and no till). In February 2006, they observed nitrate N concentrations between 0.04 and 1.73 mg kg 1 and ammonium N concentrations between 0.57 and 1.89 mg kg 1 Tho se researchers observed lower concentrations of ammonium N during the rainy season and determined the small amount of nitrate N to be a result of leaching. At the Dairy Unit, the concentrations of ammonium N and nitrate N were measured from soils sampled one month after planting. From the time of planting (15 March 2011) to when the soil samples were taken (14 April 2011), 74 kg ha 1 N of manure effluent and 35 kg ha 1 N of inorganic N fertilizer were applied to the spring
62 silage corn crop. Sampling the soil after manure effluent and inorganic fertilizer were applied explains the elevated concentration s of nitrate N and ammonium N compared to those observed by Zotarelli et al. (2009), Constantin et al. (2010), and Frazao et al. (2010) on similar soils. To accurately describe total crop available N fertilizer inputs during the spring season, the N fertil izer applications made prior to soil sampling were counted in the N mass budget inputs. Initial soil sampling avoided corn rows and would not have accurately sampled inorganic N fertilizer banded at planting. C rop uptake during the first month would have a lso taken up a portion of N fertilizer applied. Multiple years of manure effluent application have built up the organic N content of the soil in Field J to high levels (313 0 kg ha 1 ). The organic N pool in the soil provided a substantial pool of N that was subject to mineralization in turn providing mineral N for crop uptake. Although only mineral N in the soil provided N readily available for crop uptake, the organic N in the soi l provided a large reserve of N that mineralized throughout the crop year. Because manure effluent was surface applied, it built up t he N content of the upper 0 to 30 cm soil profile more than the lower 30 to 60 cm soil profile (Table 3 5) It was beneficial to the silage corn crop to have a majority of N in the upper 0 to 30 cm because corn root s only grew to approximately 40 cm in the soil profile. B ecause more mineral N (28 6 kg ha 1 N) was available in the soil than the crop requirement (235 kg ha 1 N) and additional N fertilizer s were applied, N leac hing was possible over the spring season. Manure effluent and mineralized N Farm management at the Dairy Unit applied fertilizers (inorganic N fertilizer and manure effluent) to the silage corn in order to provide N for plant uptake. Fertilizer application s to the spring silage corn were made according to the University of Florida
63 Institute of Food and Agricultural Sciences (UF IFAS) recommendations of 235 kg ha 1 N for irrigated corn or 470 kg ha 1 N applied annually for 2 seasons of irrigated corn (Mylava rapu et al., 2009). The general practice of the Dairy Unit is to apply the maximum amount of N allowed by UF IFAS Extension recommendations to produce the best silage corn yields. Manure effluent (353 kg ha 1 TKN) was applied through the center pivot irrigation system to the spring silage corn crop ( Table 3 12 ). The NMP of the Dairy Unit assumed 40% of the N in the manure effluent would be lost during and after application ( due to volatilization, denitrificati o n, leaching etc.) and adjusted the application rates to account for this projected loss (Martin, 2000) N content of the manure effluent was calculated from farm records of liters of effluent applied to Field J and analytical results from monthly samplin g. The mean manure effluent TKN concentration during the spring season was 159 mg L 1 TKN (organic N plus ammonium N) of which 92 mg L 1 (58%) was ammonium N. Daily manure effluent application for the spring season averaged 19 kg ha 1 TKN per application e vent and ranged from 4 to 29 kg ha 1 TKN (Figure 3 1). In the last month before the silage corn crop was harvested, 74 kg ha 1 TKN manure effluent was applied to the field. These applications were late in the growing season and could have increased N leach ing during the fallow period between the spring and summer silage corn seasons. A monthly schedule of manure effluent application to Field J is shown in Table 3 12 A large portion of the N applied through manure effluent was in the organic form and was no t readily available for crop uptake. Organic N was not subject to volatilization and remained in the soil until it was mineralized or lost to leaching. Ammonium N in the
64 manure effluent was available for plant uptake. Past measurements by farm management o f the nitrate N portion of manure effluent were small enough that nitrate N was not considered by farm management in N applicati ons. Manure effluent samples during the 2011 2012 crop year measured nitrate N in the range of 0.07 to 41 mg kg 1 nitrate N or 9 % of the total N concentration of manure effluent (198 mg kg 1 N). Because nitrate N was the primary form of N susceptible to leaching, leaching was not expected from manure effluent applications due to its small nitrate N concentration. In addition, a por tion of organic N from manure effluent application and the initial soil organic N pool (313 0 kg ha 1 organic N ) was mineralized by soil microbes and became available for plant uptake over the spring season. Although farm management had to apply manure effluent based on the TKN content, estimating mineralized N provided a better calculation of plant available N as a N input. For the spring season, mineralized N from manure effluent applicati ons was estimated to be only the ammonium N portion of the manure effluent. Ammonium N was 58% of manure effluent TKN during the spring. Estimated mineralized N was 204 kg ha 1 a mmonium N which was the portion of crop available N in the applied manure effluent. This estimate does not include any mineralized N from the organic N portion of manure effluent or account for any volatilization losses during application. This estimate wa s used because no mineralization experiment was carried out during the spring season to better determine mineralized N The pool of organic N in the soil was large enough that even large manure effluent applications (353 kg ha 1 TKN) during the spring seas on did not significantly increase the soil TKN content from the initial soil TKN content of 33 10 kg ha 1 N
65 Inorganic N fertilizer Inorganic N fertilizer (59 kg ha 1 N) was applied to the spring silage corn crop in addition to manure effluent ( Table 3 13 ). At planting, 35 kg ha 1 N as NH 4 NO 3 was applied as liquid fertilizer (28 0 0) band ed at planting. In April, 24 kg ha 1 N as (NH 4 ) 2 SO 4 of granular fertilizer (16 0 0) was broadcast by tractor. A monthly schedule of inorganic N fertilizer application is shown in Table 3 13 According to farm management, inorganic N fertilizer was typically applied at planting and then again when corn plan ts were 25 to 35 cm high. IFAS recommends 235 kg ha 1 N application to irrigated corn (Mylavarapu et al., due to gaseous losses of the total manure effluent applied, 212 kg h a 1 N was assumed to be available for the crops in Field J. The total N fertilizer (manure effluent assuming 40% losses plus inorganic N fertilizer) available to crops in Field J for the spring season was 271 kg ha 1 N which exceeds the IFAS recommendation (235 kg ha 1 N) for N application to irrigated corn plants in a single season The fertilizer application for the NMP receiving wastewater (174 kg ha 1 N) ( Martin, 2000 ) Mineralized N ( 204 kg ha 1 mineral N) plus inorganic fertilizer (5 9 kg ha 1 N) provided a total of 26 3 kg ha 1 mineral N for crop uptake. The available N for crop uptake was greater than the estimated N uptake for corn receiving wastewater and the IFAS rec ommended crop requirement of 235 kg ha 1 N Therefore, plant available N was more than sufficient for silage corn crop needs from mineralized N and inorganic fertilizer. Volatilization losses were expected due to the high concentration of ammonium N in the manure effluent and because it was surface applied and left unincorporated.
66 Atmospheric d eposition Atmospheric deposition was also an input of N to the spring cropping system. In 2010, the annual wet deposition was 0.6 kg ha 1 NH 4 + N and 1.2 kg ha 1 NO 3 -N (NADP, 2011) The total mineral N contributed by atmospheric deposition was 0.5 kg ha 1 N for the spring season (102 days) ( Table 3 14 ). N Outputs The spring silage corn was harvested on 25 June 2011. The soil mineral N content decreased over the spring season due to c rop uptake and leaching losses Unaccounted for N comprised the N inputs minus the measured N outputs of final soil mineral N content, harvested crop uptake, root and stubble uptake, and leac hing. Unaccounted for N was assumed to be from gas eous losse s. Final s oil N content The soil mineral N content (165 kg ha 1 N) at the end of the spring silage corn season was less ( P < 0.0001) than the initial soil mineral N content (28 6 kg ha 1 N) (Table 3 4). There was a difference ( P < 0.0001) between the final soil ammonium N content (6 6 kg ha 1 N) and the initial soil ammonium N content (17 3 kg ha 1 N) ( Table 3 7 ). The final soil nitrate N content ( 100 kg ha 1 N) in the 0 to 60 cm soil profile was not significantly different from the initial soil nitr ate N content (113 kg ha 1 N) ( Table 3 6 ). There was no significant difference in Soil TKN content in the initial spring soil TKN content (33 1 0 kg ha 1 TKN) and the final spring soil TKN content (343 0 kg ha 1 TKN)( Table 3 8 ). During the spring season, soil mineral N decreased by 12 1 kg ha 1 N due to crop uptake, leaching or gaseous losses. In the upper soil profile (0 to 30 cm), the initial and final soil mineral N contents were different ( P < 0.0001) (Table 3 5 ). In the 0 t o 30 cm soil profile, soil nitrate N
67 decreased from 73 to 5 6 kg ha 1 nitrate N and soil ammonium N decreased from 8 8 to 3 1 kg ha 1 ammonium N over the spring season ( Table 3 9 ) ( Table 3 10 ). Although soil mineral N decreased ( P < 0.0001) in the 30 to 60 cm soil profile from the soil initial to final measurements, the soil nitrate N increased (Table 3 5) (Table 3 9) The change in soil nitrate N in the 30 to 60 cm soil profile from the initial soil measurement ( 40 kg ha 1 ) to the final soil measurement (4 4 k g ha 1 ) indicated that nitrate N leached into the lower soil profile over the season ( Table 3 9 ). Zotarelli et al. (2009) found similar nitrate N trends over the spring sweet corn season with changes in nitrate N concentrations with depth indicating leachi ng. At the end of the season, soil ammonium N in the 0 to 30 cm soil profile (3 1 kg ha 1 ) was not significantly different from the soil ammonium N in the 30 to 60 cm soil profile (3 5 kg ha 1 ) ( Table 3 10 ). Although soil ammonium N decrea sed from 14 April 2 011 to 1 July 2011, soil ammonium N in the 0 to 60 cm soil profile remained evenly distributed ( Table 3 10 ). Frazao et al. (2010) observed no significant difference in ammonium N concentration by depth during the soil sampling in July 2005. Soil TKN in the 0 to 30 cm soil profile decreased from 241 0 to 2370 kg ha 1 TKN but increased in the 30 to 60 cm soil profile from 89 6 to 106 0 kg ha 1 TKN indicating movement of organic N deeper in the soil over the season Although manure effluent applications o ver the spring season were 353 kg ha 1 TKN they did not cause a significant change in soil TKN content due to the large pool of soil TKN already in the soil. The primary form of N in manure effluent was ammonium N, but the ammonium N content decreased from 173 to 66 kg ha 1 over the spring season. The depletion of soil ammonium N despite large manure effluent additions
68 suggested volatilization was a large factor in ammonium N losses. Crop uptake and nitrification would have also decreased soil ammonium N content. In the lower 30 to 60 cm soil profile, nitrate N and soil organic N increased over the season possibly indicating potential issues with leaching. Crop uptake For farm management, the most important output of the spring season nutrient budget was the silage corn crop. F arm management was interested in having as much as possible of the N inputs go into producing a crop. The dry weight of the sampled leaves, stalks, and ea rs was used to calculate a dry weight silage yield for Field J. For the spring seas on, the dry weight yield was 15.2 Mg ha 1 although the Dairy harvested yield was 8.8 Mg ha 1 ( Table 3 15 ). Spring silage corn yields at the Dairy Unit were less than those of the University of Florida spring silage corn variety trials conducted at t he Plant Science Research and Education Unit at Citra, FL (UF DAS, 2011) The mean yield for all varieties grown in 2011 was 20.1 Mg ha 1 (Table 3 16). Variety trials were also grown according to IFAS recommendations so yield differences were not due to a difference in fertilizer allowance. T he silage corn plants around the edges of the field were observed to be smaller than the plants clos er to the center of the pivot T his size difference was not reflected in the yield calcu l ated from the sampling locations because few sampling locations were on the edges of the field. Silage corn plants were sampled on 21 June 2011 and commercially harves ted on 25 June 2011. The time difference in sampling and commercial harvest meant the moisture content of the silage at commercial harvest was lower because the silage had more time to continue drying down. Fresh yield of silage corn changes based on the moisture content at harvest and the maturity of the corn plant.
69 N taken up by the leaves, stalks, and ears was harvested for silage, while the roots and stubble of the corn plants were left in the field. Although the roots and stubble were not removed from the field before the next c rop root and stubble N uptake were included in the N output pool at the end of the season because they represented a measured pool of N at the end of the season that was not lost Crop N uptake consisted of the total N ( organic N and nitrate N ) found in the plant tissue of the silage corn. Table 3 17 shows the N uptake (kg ha 1 ) of the silage corn plant parts divided into leaves, stalks, ears, stubble, and roots for the spring season. The leaves, stalks, and ears comprised the p ortion of the crop harvested commercially for silage. The spring harvested crop uptake was 20 2 kg ha 1 N ( Table 3 19 ). Harvested crop uptake N accounted for 76% of the total plant crop uptake (harvested crop uptake plus root and stubble uptake) (26 4 kg ha 1 N) in the spring season. The roots and stubble N uptake for the spring was 62 kg ha 1 N and was 24% of the total plant crop uptake ( Table 3 19 ). The percentage s of total N in dry silage corn plant parts for the spring crop a re shown in Table 3 18. For th e harvested portion of the plant, the leaves, stalks, and cobs were 2.18, 0.98, and 1.17%, respectively (Table 3 18). The total harvested silage mean was 1.48% N. The N percentage of the c rop wa s used to calculate crude protein content by multiplying the percent N by 6.25. The percent crude protein for the spring silage corn crop was 9.25% slightly higher than spring 201 1 corn silage variety trial ( 8.00% ) (Table 3 16) (UF DAS, 2011). 100 kg ha 1 ) was similar to the total N uptake of the corn grain y ield harvested by Eghball and Power (1999). Eghball
70 and Power measured an average of 66 to 99 kg ha 1 total N uptake for corn grain from 2 kg ha 1 N) was similar to results found by Van Horn et al. (1996) in Tifton, GA. Van Horn et al. (1996) harvested 129 and 197 kg ha 1 N from corn silage from yearly manure application of 421 and 493 kg ha 1 N, respectively. Althou gh the results of Eghball and Power (1999) and Van Horn et al. (1996) were similar to the results observed at the Dairy Unit, the N uptake was in the upper end of the ranges observed by past researchers. This may have been due to the large amount of manure effluent applied as N fe rtilizer. Fertilizer N use efficiency ( F NUE) was calculated for each season by dividing the N crop uptake by the N fertilizers applied. The total amount of N fertilizers applied was a combination of manure effluent and inorganic N fertilizer. The F NUE for the spring silage corn was calculated by dividing 20 2 kg ha 1 N of the harvested crop uptake by the manure effluent N application of 353 kg ha 1 plus the inorganic N fertilizer application of 5 9 kg ha 1 The F NUE of the spring silage corn was 49%. Because an unfertilized plot was not available, crop uptake solely from N fertilizers could not be determined. FNU E is overestimated, because a part of N crop uptake came from the soil mineral N content. A large portion of the manure effluent applications were organic N and were not readily available for crop uptake. Volatilization losses also subtracted from the amount of ammonium N available from manure effluent. Without an unfertilized crop to compare to the fertilized crop F NUE was difficult to evaluat e since a more accurate estimate of mineralized N was not available and soil mineral N likely provided N to the corn. M eng et al. (2012) reported apparent N recovery for maize grain yields, which was similar to the F NUE calculated in this study Apparent N recovery was calculated as the
71 N removed by the crop grown with fertilizer minus the N removed by an unfertilized crop divided by the N fertilizer application. The world average for apparent N recovery of maize grain yields is 33%. At the recommended N fe rtilizer application rate of 450 kg ha 1 N, the apparent N recove ry in the spring crop was 28%. M eng et al. (2012) determined an optimal N fertilization rate by subtracting the soil mineral N in the root zone from target N values for the crop. At the optim al N fertilization rate, the apparent N recovery rate was 72%. The spring silage corn F NUE (49%) was lower than the apparent N recovery rate of 72% achieved by the optimal N fertilization rate, but higher than the world average apparent N recovery (33%) an d the recommended N rate apparent N recovery (28%). Zotarelli et al. (2009) measured apparent N recovery of sweet corn (including stems, leaves, and ears) fertilized with 267 kg ha 1 N in 2004 and 2006 In 2004, apparent N recovery was 40% while in 2006 ap parent N recovery was 50%. For the 1 N assuming 40% gaseous losses) and F NUE (49%) were comparable to the results found by Zotarelli et al. (2009). Fertilization rates greater than the IFAS N recommendation for irrigated corn contributed to the F NUE of 49% being lower than other efficiencies for lower fertilization rates. In 2006, Zotarelli et al. (2009) observed apparent N recoveries of 80% for sweet corn fertilized with 133 kg ha 1 N and 70% for sweet corn fertilized with 200 kg ha 1 N. They found that N uptake in the ears and stover was equal for the 200 and 267 kg ha 1 N application treatments in 2004 and 2006. The harvested crop N uptake (20 2 kg ha 1 N) exceeded the Dairy estimated N uptake for corn receiving wastewater (174 kg ha 1 N) (Martin, 2000) The results of Zotarelli et al. (2009) indicated that a
72 higher F NUE could be achieved for silage corn at the Dairy Unit by applying less N fertilizer while yields would likely remain the same. The N fertilization rates of Zotarelli et al. (2009) were less than the N available from mineralized N (204 kg ha 1 ammo nium N ) plus inorganic N fertilizer (58 kg ha 1 N) applied during the spring season at the Dairy Unit. F ertilization of silage corn at the Dairy Unit was greater than the crop requirement and decreased the F NUE. T he F NUE was also lower than the results found by Zotarelli et al. because manure effluent contained a large amount of unavailable organic N which could not be uti lized by the crop without mineralization Without an unfertilized corn silage plot to compare crop uptake, it was difficult to determine the true N fertilizer use efficiency. Manure effluent applications added to the organic N pool in the soil which minera lized over time to provide adequate plant available N for crop uptake. It was unclear if manure effluent applications determined mineralization rates or if mineralization was driven solely by the soil organic N content because no mineralization experiment was completed during the spring season Mineralized N provided a substantial source of N for crop uptake during the spring season in addition to inorganic N fertilizer and soil mineral N. There was little poten tial for leaching losses because the large amo unts of manure effluent applied contained very little nitrate N, leaving inorganic N fertilizer as the only considerable source of nitrate N. Leaching Lysimeter N loads were calculated using the volume (liters) of l eachate collected, nitrate N concentration in the leachate, and the area of the lysimeter (0.534 m 2 ). Lysimeters which were empty were recorded as zero load for the sampling date.
73 Lysimeters which had less than the 40 mL of leachate needed for laboratory N analysis were recorded as no sample or zero load. The cause of the increase in nitrate N in the lower soil profile (30 to 60 cm) over the spring crop season was leaching measured by the drainage lysimeters. Only one leaching event on 21 May 2011 was observed during the spring season. Lysimeter reservoirs were empty for all other sampli ng dates over the spring. Nitrate N lea ching for the spring season (1.9 kg ha 1 N) was calculated as the mean N load per lysimeter for the one leaching event of the season on 21 May 2011 ( Table 3 20 ). Th e mean nitrate N concentratio n of the leach ate observed on 21 May 2011 was 35 mg L 1 nitrate N (Table 3 21) Over the spring silage corn season, 59 mm of fresh water was applied through the irrigation system. The greatest amount of fresh water applied on o ne day was 12 mm. Before the leaching event on 21 May 2011, a total of 11 mm of fresh water had been applied to the silage corn crop more than 2 weeks before. Therefore, fresh water irrigation was not a likely contributor to leaching during the spring crop season. Rainfall during the spring season (15 March 2011 to 25 June 2011) was low (170 mm) compared to the average rainfall of 283 mm (Figure 3 2) from 15 March to 25 June over the last 5 years (2011 2007) in Alachua County (FAWN, 2012 ). Figure 3 3 shows daily rainfall, fresh water irrigation, manure effluent, and leaching events. One high rainfall event on 14 May 2011 of 84 mm appears to have caused the spring season leaching event recorded on 21 May 2011 (Figure 3 3). Reference evapotranspiration during the spring season was 3.81 mm day 1 in Alachua County C orn crop coefficients were 0.7 for the first 40 days and 1.1 for the rest
74 of the season (SCS, 1967). Total evapo transpi ration during the spring season was 386 mm, whereas only 229 mm of water was adde d to the crop through freshwater irrigation and rainfall. This left a deficient in water availability to the silage corn crops. Constantin et al. (2010) reported 14 kg ha 1 year 1 N leached from a no till field with a cover crop. Cover crops and no till sy stems were found to be efficient techniques for reducing N leaching. The rye/ryegrass cover crop grown previous to the spring silage corn in combination with the no till agricultural practices of the Dairy Unit likely explains the small amount of leaching during the spring season. Morari et al. (2012) did not measure any nitrate N concentrations of lysimeter water greater than 25 mg L 1 nitrate N. T he mean leaching measured for maize crops was 10.9 kg ha 1 N (Morari et al., 2012). Leaching is highly variabl e and despite large additions of manure effluent during the spring crop season, little leaching was observed. So although N fertilizer applications exceeded IFAS recommendations, leaching losses during the spring season were very small, less than 1% of N outputs. H ypothesis #1 was due to overestimated gaseous losses. Farm management assumes 40% losses. This leads to an over fertilization of the crops and leaching of N to groundwater Although there was excessive N fertilizer application, leaching losses were small The decrease in soil N content over the spring season paired with little leaching losses and poor silage corn yields indicated the likelihood of high gaseous l osses. Unaccounted for N The inputs and measured outputs of the spring season were used to calculate unaccounted for N in the spring N mas s budget. For the spring season, 11 9 kg ha 1 N of the inputs were unaccounted for in the total mass budget ( Table 3 22 ).
75 Unaccounted for N was 22 % of the total N budget outputs (5 50 kg ha 1 N) in the spring season. A possible reason for unaccounted for N would be an overestimated initial soil mineral N measurement I nitial soil mineral N measurements were taken in the fi eld after the crop had been fertilized by m anure effluent and inorganic fertilizer applications (74 kg ha 1 TKN manure effluent plus 35 kg ha 1 N inorganic fertilizer ). The late soil sampling date may have resulted in a soil mineral N measurement on 14 April 2011 greater than the soil mineral N content would have been at planting (15 March 2011) Thus, a portion of the soil mineral N content measured on 14 April 2011 was also counted in the N fertilizer inputs. N total inputs for the spring season would therefore be overestimated due to that part of the N pool being counted tw ice. The overestimated N inputs would result in overestimated unaccounted for N. Unaccounted for N included volatilization and denitrification losses Farm management estimate d 40% N losses from manure effluent application due to gaseous losses. This would result in 141 kg ha 1 N of gaseous losses from the 353 kg ha 1 N of manure effluent applied as an input. Estimated unaccounted for N was 11 9 kg ha 1 N, which was 2 2 kg ha 1 N less than farm management assumptions. This suggested that N fertilizer application (353 kg ha 1 manure effluent minus 11 9 kg ha 1 N gaseous losses plus 5 9 kg ha 1 inorganic N fertilizer) was greater than the IFAS recommendation of 235 kg ha 1 N due to ov er estimated gaseous losses (Hypothesis #1). Van Horn et al. (1996) estimated volatilization losses to be 50 to 70% of N from surface applied animal manures. This would result in 17 7 to 247 kg ha 1 N losses from the 353 kg ha 1 N manure effluent applied. Al ternatively, Hall and Risser (1993) estimated vola tilization to be 25% of outputs. This estimate was similar to the Dairy
76 n unaccounted for N which was 2 2 % of the outputs (11 9 kg ha 1 N divided by 5 50 kg ha 1 N). Although manure effluent applications were large, organic N applied appeared to have been stored in the soil. This would suggest that gaseous losses were less than those proposed by Van Horn et al. (1996) and similar to those found by Hall and Risser (1993). Accounting for all of the N outputs in a system is very difficult. Nevertheless, the unaccounted for N was only 2 2 a 75% surplus of N between measured imports and exports on a dairy farm in the Netherlands. Bacon et al. (1990) found a N balance (inputs outputs) of 54% of the N inputs in 1985 and 51% of the N inputs in 1986, but Bacon did not account for environmental losses. By these standards, during the spring season at the Dairy Unit, a larger portion of N outputs wer e accounted for than in other studies. Summer Season N Inputs Silage corn was planted in Field J for the summer season on 2 July 2011, 7 days after the spring silage corn was harvested. The fin al measurements for soil N contents from the spring season were used as the initial soil N content of the summer season. No manure was applied in between the two corn crops. Manure effluent an d inorganic N fertilizers were applied to the summer season silage corn. Mineralized N w as estimated from the su mmer mineralization experiment and included as a N input in the mass balance. Crop available N inputs to Field J for the summer season were initial soil mineral N content, mineralized N inorganic N fertilizer, and atmospheric depos ition of N.
77 Initial s oil N content The initial soil mineral N (nitrate N plus ammonium N) in the total 0 to 60 cm soil profile was 165 kg ha 1 N (Table 3 4) made up of 100 kg ha 1 nitrate N ( Table 3 6 ) and 6 6 kg ha 1 ammonium N ( Table 3 7 ). The initial soil TKN (organic N plus ammonium N) content in the 0 to 60 cm soil profile for the summer season was 343 0 kg ha 1 TKN ( Table 3 8 ). Soil TKN content of the 0 to 30 cm soil profile was 2370 kg ha 1 TKN; in the lower 30 to 60 cm soil profile, it was 106 0 kg ha 1 TKN ( Table 3 11 ). Zotarelli et al. (2009) measured soil nitrate N in sweet corn plots, 15 days after harvest on 30 June 2006. Nitrate N concentrations measured between 3 and 5 mg kg 1 nitrate N. The timing of the soil sampling was similar to the initi al summer soil sampling on 1 July 2011. At the Dairy Unit, nitrate N concentrations averaged 11 mg kg 1 nitrate N in the 0 to 60 cm soil profile. The soil nitrate N concentration at the Dairy Unit is likely a higher concentration than measured by Zotarelli et al. (2009) due to the spring silage corn receiving higher N fertilization than the 133 kg ha 1 N fertilizer application to the sweet corn. Frazao et al. (2010) measured nitrate N and ammoni um N in July 2005. These researchers observed mean nitrate N co ncentrations in the 0 to 20 cm soil profile of 0.54 mg kg 1 nitrate N. The nitrate N concentrations in the 0 to 30 cm soil profile measured on 1 July 2011 were 12.6 mg kg 1 nitrate N. Constantin et al. (2010) observed the soil nitrate N content at the harv est of main crops. In Boigneville, France, the no till, cover crop experimental site measured 41 kg ha 1 nitrate N at the harvest of the main crop (spring barley). The soil nitrate N content measured at the Dairy Unit at the start of the summer silage corn season was 100 kg ha 1 nitrate N. Zotarelli et al. (2009), Frazao et al. (2010), and Constantin et al. (2010) observed soil N contents less than those
78 measured at the Dairy Unit but were comparable to those in agricultural systems with lower N fertilization of the previous crop Soil nitrate N contents higher than those found by other researchers were likel y due to the nitrification of manure effluent a pplications. During the spring season, high amounts of manure effluent were applied which would have raised soil N levels. In addition, during the last month of the growing season 74 kg ha 1 TKN was applie d t hrough manure effluent. Most of this N was likely not taken up by the spring crop, left in the soil, and subject to nitrification, raising the nitrate N concentration of the soil. Manure effluent and mineralized N Total N f ertilizer applications during the summer season were less than the N fertilizer applied in the spring season. Manure effluent applications in the summer season (31 kg ha 1 TKN) were only 9 % of those applied in the spring season (353 kg ha 1 TKN) ( Table 3 12 ) while summer inorganic fertilizer application s (56 kg ha 1 N) were similar to the quantity applied in the spring season (5 9 kg ha 1 N) ( Table 3 13 ) Manure effluent was applied on only three days in the summer growing season (Figure 3 1). One manure effluent application was 22 k g ha 1 TKN in August, while the other two applications in July were 4 and 5 kg ha 1 TKN, respectively (Figure 3 1). The mean N composition of manure effluent during the summer season was 82 mg L 1 organic N, 108 mg L 1 ammonium N, and 17.5 mg L 1 nitrate N In 2011, t he summer season had 378 mm of rainfall (Figure 3 4). The more frequent rainfall in the summer season compared to the spring season (170 mm) resulted in fewer opportunities for the farm management to apply manure effluent to the summer silage c orn because the water was not needed. Water availability did not limit crop production during the summer growing season. In the spring season, there were
79 19 days with rainfall, while in the summer season there were 29 days when it rained. Typically, farm m anagement avoided applying manure effluent when it was raining and when the field soil moisture was high due to recent heavy rainfall. A soil N mineralization experiment was conducted with the summ er silage corn crop to characterize mineral N (nitrate N pl us ammonium N) made available during the growing season. Soil mineral N could come from inorganic fertilizer, mineral N in manure effluent, atmospheric deposition, and from the mineralization of soil organic matter. Sources of soil organic matter include m anure effluent, roots, stubble, and other plant matter left in the field from previous seasons. The soil N mineralization experiment mineralization of organic N over the summer sil age corn crop. Mineralized N was a N input in the summer N mass budget and was available for crop uptake as well as environmental losses. Further, kno wing the amount of mineralized N will help the farm management make decisions about amounts of manure effl uent needed for crop growth requirement s Organic N from manure effluent applications during the summer season in addition to the pool of organic N remaining in th e soil from previous seasons was mineralized during the summer N mineralization experiment. Over the summer mineralization experiment, 31 kg ha 1 N of manure effluent was applied to Field J. The complete mineralization experiment sampling plan is described in Table 2 3. Staggering the in stallation of PVC pipes in monthly time periods allowed for the isolation of different initial soil mineral N contents throughout the summer. Varying initial soil mineral N contents resulted from applications of manure effluent. Over a one month period th e
80 observed change in soil mineral N was due to net mineralization (positive change) or net immobilization (negat ive change) One month periods also allowed for baseline soil mineral N contents to be established from each manure effluent application. For ex ample, a soil isolated in July and its subsequent mineral N changes over time could serve as a baseline for a soil isolated in August containing additional manure effluent. Soil mineral N changes different than those observed in the July soil baseline coul d be attributed to the additional manure effluent in the soil isolated in August. Therefore, t he In the month prec eding the start of the summer mineralization experiment (24 June to 24 July 2011), 9 kg ha 1 N of manure effluent was applied to the silage corn in Field J. The initial set of PVC pipes for the summer mineralization experiment was installed on 24 July 2011 The results of the soil mineral N content of the soil isolated by PVC pipes on 24 July 2011 are shown in Table 3 23 The monthly soil mineral N content w as used to calculate any increases in soil mineral N due to net miner alization of organic N in the so il over the time of the experiment. For the soil isolated by PVC pipes installed on 24 July, the change in mineral N content over the first month time period could not be calculated, because no initial soil samples were taken on 24 July. The change in mine ral N content at each lysimeter location was calculated from Aug ust to Sept ember and Sept ember to Oct ober by subtracting the mineral N content on 24 Aug ust from the content on 24 Sept ember and the content on 24 Sept ember from 24 Oct ober For each lysimeter location, the accumulations of mineral N in the soil isolated by the PVC pipes installed on 24 July
81 were summed. The sums for each lysimeter location were then averaged to get the N mineralized for the soil samples isolated by the first set of PVC pipes in stalled on 24 July. The mean mineralized N during the f irst measurement period was 18 kg ha 1 mineral N which was considered a N input in the summer N mass budget. The mineral N content of soil isolated on 24 July 2011 was used as a baseline to compare aga inst the mineral N content of soil samples isolated by PVC pipes installed on 24 Aug ust and sampled on 24 Sept ember and 24 Oct ober Between 24 July and 24 Aug ust 2011, 22 kg ha 1 N from manure effluent was added to Field J. Soil mineral N content of the soil isolated by PVC pipes installed on 24 Aug ust 2011 was measured on 24 Sept. and 24 Oct. The baseline changes in the soil mineral N content of soil isolated on 24 July 2011 were subtra cted from the change in soil mineral N isolated on 24 Aug. 2011. By subtracting the baseline, the difference showed only the net mineralization/immobilization after 24 Aug ust The mean mineralized N in the soil isolated by PVC pipes on 24 August was 24 kg ha 1 mineral N ( Table 3 24 ). Between 24 Aug ust and 24 Sept ember 2011, no manure effluent was added. Mineralization and immobilization still occurred due to the supply of organic N in the soil as of 24 Sept ember 2011. PVC pipes were installed on 24 Sept embe r and the isolated soil was sampled on 24 Oct. 2011. The mineral N content of the soil isolated on 24 Sept ember and sampled on 24 Oct ober was compared to the baseline mineral N content from the soil isolated on 24 Aug ust and sampled on 24 Oct ober. In each location, the mineral N content of the soil isolated on 24 September was lower than the baseline N content from the soil isolated on 24 Aug ust ( Table 3 24 ) resulting in net
82 immobilization or no mineral N added to the soil from mineralization from 24 Sept em ber to 24 Oct ober Over the summer mineralization experiment, the total mineral N available for plant uptake due to mineralizat ion was 4 3 kg ha 1 The mean daily mineralized N was 0.7 kg ha 1 mineral N. Figure 3 5 depicts the net mineralization and immobil ization for P VC pipes installed in July, August, and Sept ember Figure 3 6 shows the results of the soil mineral N content of the soil isolated by the PVC pipes installed on 24 July 2011. The general trend was a decrease or net immobilization of soil mineral N from the soil sampled one month after installation to the soil sampled two months after installation (24 Sept. 2011). Between soil sampled two months after installation and three months after installation (24 Sept ember to 24 Oct ober 2011), the general trend was a positive increase or net mineralization i n soil mineral N. Although it was unknown whether the mineral N content increased from 24 July to 24 Aug ust 2011, the general trend of the mineral N content followed the same results as the i ncubation trial of Cusick et al. (2006). They found a net mineralization in the first 21 days, a net immobilization from day 21 to 84, and then a net mineralization from day 84 to the end of the trial. re similar to the results s een in Figure 3 6 although it wa s likely that net mineralization began before 84 days in the summer mineralization experiment. The summer mineralization experiment did not provide a mineral N estimate for the entire season. Mineralized N for the summer sea son was calculated using the average daily mineralized N of 0.71 kg ha 1 mineral N. The summer season was 81 days long resulting in an estimated mineralized N of 57.4 kg ha 1 mineral N (0.71 kg ha 1 N multiplied by 81 days) Estimated mineralized N for the summer season was greater
83 than manure effluent applications (31 kg ha 1 TKN). Large amounts of manure effluent applied during the spring season continued to mineralize over the summer season providing mineralized N greater than the summer manure effluent applications. This was expected because manure effluent would not fully mineralize during one crop season. Mineralized N accounted for 21 % of total mineral N inputs for the summer crop season (279 kg ha 1 mineral N) During the summer crop season, we faile d to reject Hypothesis contributor to crop available N inputs followed by mineralized N (57 kg ha 1 mineral N) estimated from the summer mineralization experiment. Inorganic N fertilizer In the summer season, a total of 56 kg ha 1 of inorganic fertilizer was applied ( Table 3 13 ) At planting, 34 kg ha 1 N as NH 4 NO 3 of liquid fertilizer (28 0 0) was applied during the planting operation In August, 22 kg ha 1 N as NH 4 NO 3 of liquid fertilizer (28 0 0) was applied to the crop through the center pivot irrigation The total N fertilizer application for the summer silage corn season was less than the IFAS recommendation of 235 kg ha 1 N for irrigated corn (Mylavarapu et al., 2009). Assuming 40% losses of the manure effluent applied, only 19 kg ha 1 N of manure effluent was available for the corn crop in Field J during the summer season ( Martin, 2000) The total N fertilizer application assumed by farm management (manure effluent after 40% losses plus inorganic fertilizer) was 75 kg ha 1 N which was 160 kg ha 1 management plan estimated corn N uptake to be 174 kg ha 1 N. Applied N fertilizer for the summer season was less than the estimated crop uptake
84 Inorganic N fertilizer in addition to the estimated m in eralized N (57 kg ha 1 mineral N) provided a total crop available minera l N of 113 kg ha 1 mineral N. These sources of crop available N f e ll short of the 235 kg ha 1 N recommendation for irrigated corn. Although the soil mineral N content could have provided N to the plants, the N applications were not sufficient for crop requ irements. The lack of N fertilization suggested that crop yields would be limited. Leaching was also not expected to be substantial during the summer season due to the lack of N fertilization. Atmospheric deposition Atmospheric deposition of N contributed to the N inputs for the summer silage corn crop. The summer season wa s 81 days long, from 2 July to 21 Sept. 2011. Total NO 3 N and NH 4 + N from atmospheric deposition from the summer season was 0.4 kg ha 1 N ( Table 3 14 ) (NADP, 2011) N Outputs The summer silage corn was harvested on 21 Sept. 2011. Crop available N inputs were the initial soil mineral N content, estimated mineralized N, inorganic N fertilizer, and atmospheric deposition. Crop uptake was the most desirable N output. Other N output s were the final soil N content, leaching, and unaccounted for N. Unaccounted for N was assumed to be from gaseous losses. Final s oil N content The final soil mineral N content of the summer season was 5 3 kg ha 1 N. There was a significant difference ( P < 0.0001) between the initial soil mineral N content (165 kg ha 1 N) in the 0 to 60 cm soil profile and final soil mineral N content (5 3 kg ha 1 N) (Table 3 4). There was a change of 11 2 kg ha 1 N of mineral N from the beginning to the end of the summer grow ing season. Final soil nitrate N content (47 kg ha 1 nitrate N)
85 significantly differed ( P < 0.0001) from the initial soil nitrate N content ( 100 kg ha 1 nitrate N) in the 0 to 60 cm soil profile ( Table 3 6 ). Final soil nitrate N levels for the summer season were similar to results found by Jokela (1992) in Swanton, VT. In October 1988, he found 86 kg ha 1 of soil nitrate N in 0 to 150 cm of the soil profile after a corn harvest with N fertilizer (56 kg ha 1 N at planting) and manure application (240 kg ha 1 yr 1 N ) to Field J. Constantin et al. (2010) observed nitrate N levels of 25 kg ha 1 nitrate N in late autumn on no till plots with a cover crop grown on soils similar to the Dairy Unit. Soil TKN content at the end of the summer season was 346 0 kg ha 1 N ( Table 3 8 ). There was no significant difference between the initial and final soil TKN contents of the summer season ( Table 3 8 ) Because the soil TKN content was a larger pool of soil N compared to nitra te N and ammonia N, the small amount of manure effluent ap plied during the summer season did not significantly increase the soil TKN content. Soil nitrate N in the lower 30 to 60 cm decreased from 4 4 to 1 8 kg ha 1 nitrate N over the summer season. This sug gested possible nitrate N leaching, because corn roots only reached to approximately 40 cm making significant N uptake in the lower soil profile un likely Only one third of the initial soil mineral N content remained at the end of the summer season. Crop u ptake could have depleted mineral N in the soil, because N fert ilization was inadequate during the summer season. Crop uptake The dry weight yield for the summer silage corn crop was calculated from the corn samples t aken on 20 Sept. 2011. The dry weight y ield of the summer silage corn (11.8 Mg ha 1 ) was less than the spring silage corn dry weight yield of 15.2 Mg ha 1 ( Table 3 15 dry weight harvested yield ( 7.6 Mg ha 1 ) was less than the
86 calculated yield from the crop sampling due to smaller plants near the field edges which were not sampled in the current research The summer silage corn crop was commercially harvested on 21 Sept. 2011. Silage corn production yields are typically less for the summer season than the spring season due t o shorter growing seasons In the 2011 Corn Silage Field Day Corn Hybrid Variety Test (UF DAS, 2011) performed by the University of Florida in Citra, Florida, varieties of silage corn were planted in the spring and in the summer. The mean dry yield per hec tare was 20.2 Mg ha 1 for the spring planting whereas for the summer silage corn mean dry yield was 14.8 Mg ha 1 (Table 3 16). The variety test yield results were characteristic of silage corn production in Florida where summer yields are typically 25% le ss than spring yields (UF DAS, 2011). Because farm management expected greater silage yields in the spring season, more manure effluent was applied during the spring season than in the summer. The concentrations of total N in dry silage corn plant parts fo r the summer crop are shown in Table 3 18. For the harvested portion of the plant, the leaves, stalks, and ear s were 1.77, 0.47, and 1.34%, respectively (Table 3 18). The total harvested silage mean was 1.39% N. The concentration of crude protein for the summer silage corn crop was 8.69% slightly higher than summer 2011 corn silage variety trial (7.34%) (Table 3 16) (UF DAS, 2011). Both seasons of silage corn at the Dairy Unit had poorer yields than the variety trials, but a higher concentration of N and c rude protein. Adequate N fertilization would likely have increased the summer silage corn yields. Estimated mineralized N and inorganic N fertilizer applications during the summer season provided less than 50% of the N crop requirement of 235 kg ha 1 N. Cr op
87 available N inputs totaled 279 kg ha 1 mineral N for the summer season, but 78 kg ha 1 mineral N of that amount was soil mineral N in the lower soil profile to which plants had limited access Farm management did not provide adequate N fertilization for the summer silage corn. The crop N uptake of the summer silage corn season was calculated for leaves, stalks, ears, stubble, and roots and the results are shown in Table 3 17 There was a significant difference in the N uptake for all plant parts ( P < 0.0001 for the leaves, stalks, stubble, and roots; P = 0.0461 for the ears) between the spring and summer season crops. The summer season corn N uptake was less than plant N uptake in the spring silage corn. Greater yields during the spring crop season and lower fertilization in the summer explain the difference in crop N uptake. The harvested crop uptake (leaves, stalks, and ears) was 142 kg ha 1 N for the summer crop ( Table 3 19 ). The harvested crop N uptake was removed from Field J and used for c orn sila ge to feed to the dairy cows. Harvested crop uptake was 88% of the total plant crop N uptake for the summer season. During the spring season, the harvested crop uptake had been 76% of the total crop plant N uptake. Although the summer silage corn crop had a lower yield than the spring silage corn crop, it was more efficient in partitioning N to the harvested crop uptake plant parts. The root and stubble N uptake in the summer silage corn season was 20 kg ha 1 N ( Table 3 19 ). For the summer season, root and stubble N uptake accounted for 12 % of the total plant crop uptake. Root N content was greater than stubble N content for both the spring and summer silage corn crops. Root N uptake was 45 and 16 kg ha 1 N in the
88 spring and summer crops, respectively ( Tabl e 3 17 ). Stubble N uptake was 17 and 4 kg ha 1 N in the spring and summer crops, respectively ( Table 3 17 ). The F NUE for the summer silage corn was calculated by dividing 142 kg ha 1 N in the harvested crop uptake by the N fertilizer applied, which was the sum of manure effluent N application (31 kg ha 1 TKN) plus the inorganic N fertilizer application (56 kg ha 1 N). The calculated F NUE of the summer silage corn was 163%. The large difference in silage corn F NUE between the spring and summer seasons was pr imarily caused by the difference in N fertilizer application. The summer silage corn crop received much less manure effluent than the spring silage corn crop (31 vs. 353 kg ha 1 TKN). The summer silage corn harvested crop uptake was greater than the N fert ilizers applied which indicated that the summer silage corn drew N from the soil N content. The summer silage corn F NUE indicated that the silage corn crop yield was limited by inadequate N fertilization The results for apparent N recovery found by M eng et al. (2012) are not comparable to the F NUE calculated for the summer silage corn. Because the N crop uptake of the summer silage corn was greater than the total N applied through manure effluent and inorganic N fertilizer, a portion of N crop uptake cam e from the soil mineral N content. Without an unfertilized silage corn crop to compare to the summer silage corn N crop upta ke, the F NUE of the summer crop wa s difficult to evaluate based solely on N fertilizer applications. Although the high F NUE (163%) o f the summer crop seems optimal, it wa s clear that the soil mineral N content was depleted by crop uptake. In 2006, Zotarelli et al. (2009) reported the highest apparent N recovery (80%) on sweet corn fertilized with 133 kg ha 1 N. This N treatment did not result in the greatest
89 yield. The 133 kg ha 1 N treatment yielded 9.76 Mg ha 1 dry above ground biomass. With a lower apparent N recovery of 70%, the 200 kg ha 1 N treatment yielded 9.81 Mg ha 1 dry above ground biomass. It wa s likely that the summer yiel d would have been greater with greater N fertilization similar to the results observed by Zotarelli et al. (2009). Leaching During the summer seasons, some drainage lysimeters did not yield an observation at each sampling date because of problems with the lysimeter equipment. Lysimeter reservoirs at location C were dug up and repaired on 22 November 2011 due to problems incurred over the summer season. Reservoirs at locations A and B were dug up and repaired on 17 January 2012. All repaired drainage lysimet ers showed the same problem of crimping in the soft, Tygon tubing used to pump out the reservoir from the weight of the soil on top of the bucket reservoir pushing the PVC piping down into the tubing (Figure 2 3). Crimped tubing prevented leachate from bei ng pumped out of the reservoir and therefore no observation was recorded. Leaching was calculated from the 8 lysimeters which were not blocked during the summer season. Lysimeter locations A2, B2, C1, and C2 (Figure 2 1) were blocked for the 4 sampling dat es during the summer season and were not included in the mean load for the season. Lysimeter N loads were calculated using the volume (liters) of l eachate collected, the nitrate N concentration in the leachate, and the area of the lysimeter (0.534 m 2 ). The calculated load of each lysimeter was summed over all summer sampling dates. The individual lysimeter sums were then averaged to obtain the summer leaching total load. The leaching events for the summer season resulted in the greatest ( P = 0.0028) leachin g load for an y season (1 9 kg ha 1 nitrate N ) ( Table 3
90 20 ). Leaching observations were recorded for each sampling date. The mean nitrate N concentration of observed leachate events for the summer season was 66 mg L 1 (Table 3 21). The short fallow period between the spring and summer silage corn seasons had one leaching event of 4.8 kg ha 1 nitrate N. During the fallow period, Field J received 37 mm of rainfall on 30 June 2011 (Figure 3 3). After the summer silage crop was planted, Field J received 53 mm o f rainfall on 9 July 2011. There was no freshwater irrigation application during the fallow period. In June, relatively late in the spring growing season, 74 kg ha 1 TK application late in the growing season increases the likelihood of leaching losses in the rainfall events, or a combination of the two were the cause of the leaching event observed on 17 June 2011. Therefore, w e fail ed to reject h ypothesis #3. Because there were few manure effluent applications during the summer season, thus lower quantities of water, it wa s likely that rainfall events were the major cause of leaching events and not manure effluent applications. The summer season rainfall was 208 mm greater than the rainfall during the spring season. The summer season also had a greater number of days of rainfall (29 vs. 19 days in the spring) in addition to the summer season being 20 days shorter than the spring season. The greater amount of rainfall during the summer season most likely caused the increase in the number of leaching events and overall load during the summer season compared to the spring season as shown in Figure 3 3.
91 Rainfal l during the summer season (378 mm) was less than average rainfall over the last 5 years (508 mm) (Figure 3 4) The dry summer weather likely limited leaching compared to past summer seasons with greater rainfall. During a summer season with greater rainfa ll, it is likely that leachin g would be greater than the 1 9 kg ha 1 nitrate N observed during the 2011 summer season. During the summer season, a total of 31 mm of fresh water was applied to Field J through the center pivot irrigation system (Figure 3 3). The mean fresh water applied on a single day was 4 mm. This amount was less than the mean rainfall event on a single day during the summer season (10 mm). Fresh water was primarily applied later in the summer season with almost all applications in late Aug ust and Sept ember Because fresh water was applied in smaller quantities c ompared to rainfall events, it wa s not likely that fresh water application was a cause of leaching. The reference evapotranspiration rate for the summer season was 3.95 mm day 1 ( FAW N, 2012 ) The corn crop coefficients for the first 40 days of the season w ere 0.7 and 1.1 for the remainder of the season (SCS, 1967) The total crop evapotranspiration for the summer season was 286 mm. Total rainfall (378 mm) and freshwater irrigation (31 mm) provided 409 mm of water to the corn crop. Considering manure effluent applications added additional water to the crop, the water applied to Field J was greater than evapotranspiration losses. Therefore, l eaching during the summer season was expected. In Boigneville, France, Cons tantin et al. (2010) reported a mean 14 kg ha 1 year 1 N leached on a no till field with a cover crop. Cover crops and no till systems were found to be efficient techniques for reducing N leaching. Although leaching over the
92 su mmer season at the Dairy Unit was greater than the 14 kg ha 1 year 1 N reported by Constantin et al. (2010) it was less than the mean N leached (32 kg ha 1 year 1 ) on all experiment sites. Leaching events during the summer silage corn season can best be e xplained by high rainfall events. Because the quantity of the few manure effluent applications was similar to manure effluent application rates in th e spring season which only had 1.9 kg ha 1 nitrate N leached, manure effluent was not a likely cause of N l eaching during the summer. Freshwater irrigation was also in smaller quantities than rainfall events. Leaching during the summer silage corn season was more likely caused by the saturation of soil pores due to rainfall or a combination of rainfall, freshwa ter irrigation, and manure effluent. recommendation due to overestimated gaseous losses. Farm management assumes 40% losses. This leads to an over fertilization of the cr ops and leaching of N to xceed IFAS recommendations, it wa s possible that excessive N fertilizer applications over the spring season increased soil nitrate N in the lower 30 t o 60 cm soil profile and resulted in leaching during the summer season when rainfall was much greater (Figure 3 2) (Figure 3 4). Unaccounted for N The inputs and measured outputs of the summer season were used to calculate unaccounted for N in the summer N mas s budget. Unaccounted for N was calculated by subtracting the measured outputs of final soil mineral N content, crop N uptake, root and stubble N uptake, and leaching from the inputs of initial soil mineral N, estimated
93 mineralized N i norganic fertilizer, and atmospheric deposit ion. For the summer season, 4 4 kg ha 1 N of the N inputs were unaccounted for in the N mass balance ( Table 3 25 ). Unaccounted for N was composed of gaseous losses Unaccounted for N was lower in the summer season (4 4 kg ha 1 N) than in the spring season (11 9 kg ha 1 N), accounting for 16 % of the N outputs in the summer crop season. Farm management at the Dairy Unit assumed that 12 kg ha 1 N of manure effluent applications (40% of the 31 kg ha 1 N manure effluent applied) would be lost to gaseous losses. Un accounted for N was greater than gaseous losses assumed by farm management at the Da iry Unit. Unaccounted for N was greater than the estimate of 50 to 70% of volatilization from animal manures by Van Horn et al. (1996). The percentage of unaccounted for N relative to N outputs assumed to be gaseous losses, was less than mass budget on a Pennsylvania dairy farm. Ku ipers et al. (1999) and Bacon et al. (1990) found a greater percentage of their N mass budgets were unaccounted for than the 16% of N outputs as unaccounted for N in the summer season Winter Season N Inputs The final soil mineral N content for the summer season was the starting point for the winter season. The fallow period between the summer harvest on 21 Sept. 2011 and the rye/ryegrass mixture planted on 28 Oct. 2011 was included in the winter season. There were no inputs into Field J by farm management during the fallow period between summer corn and winter rye ryegrass During the winter season, soil mi neral N content, mineralized N inorganic N fertilizer, and atmospheric deposition of N provided inputs of N for the rye/ryegrass winter cover crop.
94 I nit ial soil N content Initial soil mineral N content for the winter season was 5 3 kg ha 1 consisting of 47 kg ha 1 nitrate N and 6 kg ha 1 ammonium N (Table 3 4, 3 6 and 3 7 ). Soil TKN (organic N plus ammonium N) was 346 0 kg ha 1 TK N at the start of the wint er season, but only the ammonium N in the soil TKN content was readily available to provide N to the rye ryegrass mixture ( Table 3 8 ). Soil nitrate N and soil TKN significantly decreased ( P < 0.0001) with sampling depth (Table 3 9 3 1 1 ). Soil TKN content of the upper 0 to 30 cm was 2350 kg ha 1 TK N; in the lower 30 to 60 cm, it was 111 0 kg ha 1 TK N ( Table 3 11 ). Jokela (1992) measured nitrate N in the 0 to 150 cm soil profile in Nov ember 1987 and Oct ober 1988 on a sandy loam soil in Ver mont. The experimental plot similar to the 1 year 1 TK N from manure and 56 kg ha 1 inorganic N fertilizer at planting of summer corn crops. The soil nitrate N measured in 1987 was 87 kg ha 1 nitr ate N and 86 kg ha 1 nitrate N in 1988. Although 47 kg ha 1 nitrate N measured in Sept ember 2011, the measurements at the Dairy Unit were from the 0 to 60 cm soil profil e instead of 0 to 150 cm When taking into account (1992) report that very little nitrate N was measured below 90 cm in the soil profile t he nitrate N measured at the Dairy Unit was similar to the 86 to 87 kg ha 1 nitrate N measured by Jokela (19 92) when the difference in soil sampling depth was taken into account. Manure effluent and mineralized N During the fallow period from 21 September to 28 Oct ober 2011, no manure effluent or inorganic fertilizer was applied to Field J. Once the winter rye ryeg rass crop
95 was planted on 28 October 2011, a total of 104 kg ha 1 TKN from manure effluent was applied through the center pivot irrigation system during the winter crop season ( Table 3 12 ). The majority of the manure effluent applications occurred in No vember and December, and consisted of 68 and 34 kg ha 1 TKN respectively ( Table 3 12 ). The N composition of manure effluent during the winter season was 115 mg L 1 organic N 93 mg L 1 ammonium N, and 9 mg L 1 nitrate N. Daily manure effluent application is shown in Figure 3 1. Mean manure effluent applied at each application event during the winter season was 17 kg ha 1 TK N. A winter soil N mineralization experime nt was conducted with the rye ryegrass cover crop to determine amounts of mineral N (nitrate N plus ammonium N) made available during the winter growing season. The results of the winter mineralization experiment are shown in Table 3 26 The design of the summer mineralization experiment was improved upon for the winter season by taking soil sampl es when PVC pipes were installed. Initial soil samples allowed for the change in mineral N concentration to be recorded over more periods. The results of the winter mineralization experiment were used to calculate the increase in mineral N in the soil due to mineralization of manure effluent and organic N in the soil over the time of the experiment. The calculations used were the same as those used for the summer mineralization experiment. The change in soil mineral N content and the mean field mineralizati on are shown in Table 3 27 The mean mineralized N of the soil isolated by the PVC pipes i nstalled on 8 Nov. 2011 was 7 kg ha 1 mineral N. The mineral N contents of soil isolated on 8 Nov 2011 and sampled on 8 Dec 2011, 8 Jan 2012 and 9 Feb. 2012 were used as baselines to
96 compare against the mineral N content of soil samples taken on 8 Dec 2011 and soil isola ted by PVC pipes installed on 8 Dec. 2011 and s ampled on 8 Jan. 2012 and 9 Feb. 2012. Between 8 Nov. and 8 Dec. 2011 82 kg ha 1 TK N manure effluent was added. Soil samples were taken at each lysimeter location on 8 Dec. 2011. Soil mineral N content of the soil isolated by PVC pipes installed on 8 Dec. 2011 was measured on 8 Jan. 2012 and 9 Feb. 2012. The baseline changes in the soil mineral N content of soil isolated on 8 Nov. 2011 were subtracted from the change in soil mineral N isolated on 8 Dec 2011. By subtracting the baseline, the difference showed only the net mineralization or immobilization after 8 Dec ember. The mean mineralized N fo r the soil isolated by PVC pipes inst alled on 8 Dec. 2011 was 94 kg ha 1 mineral N. Between 8 Dec. 2011 and 8 Jan. 2012, no manure effluent was added to Field J. PVC pipes were installed and a soil sample was taken on 8 Jan 2012 and the isolated soil was sampled on 9 Feb 2012 The mineral N content of the soil samples taken on 8 Jan. 2012 and the soil isolated on 8 Jan uary and sampled on 9 Feb ruary was compared to the baseline mineral N content from the soil isolated on 8 Dec 2011 and sampled on 8 Jan. 2012 and 9 Feb. 2012 Even though no additional manure effluent was added, un mineralized manure effluent from previous time periods and organic N in the soil w ere subject to mineralization. The mean mineral N added from mineral ization for the PVC pipes i nstalled at 8 Jan. 2012 was 44 kg ha 1 The total mineralized N over the winter mi neralization experiment was 14 5 kg ha 1 Figure 3 7 depicts the net mineralization or immobilization of PVC pipes installed in Nov ember Dec ember and Jan uary
97 Estimated mineralized N during the winter season was greater than the manure effluent application of 104 kg ha 1 TKN. Continued mineralization of both past manure effluent applications and the pool of organic N in the soil provided more minera l N than was applied. This was expected because large manure effluent applications were made during the spring season and manure effluent was mineralized slowly over time. The winter mineralization experiment lasted from Nov ember 2011 to Feb ruary 2012 Alt hough the winter season began when the summer silage corn was harvested on 21 Sept. 2011, Field J received no manure effluent additions during Sept ember and Oct ober The summer mineralization experiment found no mineralized N in October. Therefore, the win ter mineralization experiment was a good estimate of mineralized N during the winter season. This addition of mineralize d N provided N for possible rye ryegrass plant uptake over the winter season and was a N input in the winter N mass balance. Mineralized N was 64 % of N inputs for the winter cover crop s eason. As a result, Hypothesis # of the N in the tot Figure 3 8 shows the results of the s oil mineral N content of the PVC pipes installed at T0. The general trend was a n increase in soil mineral N due to net mineralization from installation to 30 days after installation, a decrease in soil mineral N or net immobilization from the pipes pulled 30 days after inst allation to the pipes pulled 90 days after installation T he general trend of the min eral N content was similar to the results of an incubation trial by Cusick et al. (2006). They found a net mineralization in the first 21 days, a net immobilization from day 21 to 84, and then a net mineralization from day 84 to the end of the trial. The results in Figure 3 8 do not show a net
98 mineralization in the soil mineral N content 90 days after ins tallation. The cooler weather of the winter months could have resulted in slowed mineralization. The overall trend in soil mineral N is consistent with the results of Azeez and Van Averbeke (2010). In the oratory incubation study, the initi al mineral N concentration of 27 mg kg 1 increased to 18 6 after 30 days, decreased to 3 3 mg kg 1 after 70 days, and decreased further to 20 mg kg 1 after 90 days. Inorganic N fertilizer Inorganic N fe rtilizer was applied to the rye ryegrass to provide an additional 24 kg ha 1 N as (NH 4 ) 2 SO 4 for plant growth. The monthly winter schedule of inorganic N fertilizer applications is shown in Table 3 13 Granular fertilizer (19 0 0) was broadcast by t ractor only one time to the rye ryegrass mixture in December. A nnual silage corn N fertilizer application in 2011 was below IFAS recommended N rates of 470 kg ha 1 N. Because the full amount of allowed N fertilizer was not applied during the spring plus summer corn season s supplemental N fertilizer was a llowed to be applied to the rye ryegrass winter cover crop. Assuming a 40% loss of N from the manure applications, 86 kg ha 1 N fertilizer (manure effluent plus inorganic fertilizer) was available to the rye/ryegrass cover crop. The N fertilizer application was less th an the estimated N uptake by rye/ryegrass receiving wastewater (135 kg ha 1 N) from the Dairy U Mineralized N estimated from the winter minerali zation experiment provided 14 5 kg ha 1 mineral N to the winter cover crop. The combination of mineralized N and inorganic N fertilizer provided mineral N (1 70 kg ha 1 mineral N ) greater than the estimated crop uptake of 135 kg ha 1 N (Martin, 2000). Since manure effluent wa s the
99 primary form o f N fertilizer applied at the Dairy Uni t, mineralization of organic N wa s an important source of N for crop uptake needs. Atmospheric deposition Atmospheric deposition of N was an input of the winter rye ryegrass season. The winter season was 17 2 days including the fallow period. Total nitrate N and ammonium N from atmospheric depositio n from the winter season was 0.8 kg ha 1 N made up of 0.2 8 kg ha 1 NH 4 + N and 0 .56 kg ha 1 NO 3 N ( Table 3 14 ). N Outputs The winter rye ryeg rass mixture was harvested on 13 Mar 2012. The final soil mineral N content decreased over the winter season. Crop uptake and leaching losses were measured outputs that decreased the final soil mineral N content. Unaccounted for N comprised the N inputs minus the measured outputs of final soil mineral N content, crop uptake, and leaching. Final soil N content In the winter season, the final soil mineral N content was 2 8 kg ha 1 (Table 3 4). Soil nitrate N content of the total 0 to 60 cm soil profile was significantly diffe rent ( P < 0.0001) between the initial soil nitrate N content and final soil nitrate N content (47 vs. 20 kg ha 1 nitrate N) ( Table 3 6 ). The final soil nitrate N level was significantly different ( P < 0.0001) between the upper 30 cm (1 6 kg ha 1 nitrate N) and the lower 30 to 60 cm ( 4 kg ha 1 nitrate N) of the soil core ( Table 3 9 ). There was a significant difference ( P < 0.0001) between the initial soil ammonium N and final soil ammonium N measurements for the total 0 to 60 cm soil profile ( Table 3 7 ). Duri ng the winter season, soil ammonium N increased from 6 to 8 kg ha 1 The final soil ammonium N content did not differ significantly by depth ( Table 3 10 ).
100 The final soil TKN content was 346 0 kg ha 1 TK N at the end of the winter season. There was not a significant change from the initial soil TKN content (346 0 kg ha 1 TK N) ( Table 3 8 ). Soil TKN differed significantly ( P < 0.0001) by depth. In the 0 to 30 cm soil profile, soil TKN content was 256 0 kg ha 1 TK N at the end of the winter season. In the 30 to 60 cm soil profile, soil TKN content was 900 kg ha 1 TK N ( Table 3 11 ). Constantin et al. (2010) measured nitrate N levels of 33 kg ha 1 nitrate N in mid winter on no till plots with a cover crop (planted late Aug ust to Sept ember ) grown on soils similar to those at the Dairy Unit. They observed an increase in soil nitrate N from sampling in late autumn to mid winter. The soil nitrate N content at the Dairy Unit significantly decreased from the initial winter measurements (47 kg ha 1 ) to the final winter mea surements ( 20 kg ha 1 ) in 2012. Constantin et al. (2010) determined that the cover crop diminished soil mineral N compared to the non cover crop treatments in both late autumn and mid winter, but more efficiently in late autumn. At the Dairy Unit, winter i nitial soil measurements were not reduced by a cover crop, because the initial soil measurements for the winter season were sampled in Sept ember before the winter rye ryegrass was planted. Woodard et al. (2003) reported rye to be highly effective at contr olling nitrate N leaching. Rye was reported to extract large amounts of N from the soil as well as soil mo isture Mineralized N contributed 14 5 kg ha 1 mineral N to the soil profile over the winter crop season. Despite the addition of mineralized N, there was a net loss of 25 kg ha 1 soil mineral N over the winter season. Possible soil mineral N fates were crop uptake, leaching, or gaseous losses.
101 Crop u ptake The dry weight yield for the winter rye ryegrass mixture was calculated from the samples of t he harvested portion of the rye ryegrass mixture taken on 17 and 20 Feb. 2012. T he stubble and roots of the rye ryegrass mixtur e were not included in the dry weight yiel d for the winter crop. The dry weight yield for the winter rye ryegrass was 4.1 Mg ha 1 ( T able 3 15 weight harvested yield (2.4 Mg ha 1 ) was less than th e calculated yield from the rye ryegrass samples. Differences in yiel d between the rye ryegrass dry weight we ight harvested yield were due to the crop drying down between the sampling date (17 and 20 February) and final commercial harvested on 12 March 2012. Harvested crop uptake contained the total N (organic N plus nitrate N ) found in the plant tissue of t he harvested portion of the rye rye grass mixture. The total N removed by rye ryegrass harvested crop uptake from the winter season was 79 kg ha 1 N ( Table 3 19 ). The total plant crop uptake was 128 kg ha 1 N made up of 79 kg ha 1 N of harvested crop uptake and 4 9 kg ha 1 N of root and stubb le uptake ( Table 3 19 ). The harvested portion of the rye ryegrass was 62% of the tota l plant crop uptake for the rye ryegrass mix ture. The root N uptake was 27 kg ha 1 N; the stubble N uptake was 2 2 kg ha 1 N ( Table 3 28 ). The root and st ubble uptake for the winter rye ryegrass was an output of the winter season for the N mass budget. Van Horn et al. (1996) grew corn silage and Abruzzi ry e in Tifton, GA. These researchers found rye crop N uptake of 53 to 249 kg ha 1 with manure effluent applications of 24 0 to 9 90 kg ha 1 yr 1 N. In the treatment most similar to the manure application of the Dairy Unit, Van Horn et al. (1996) applied 493 kg ha 1 yr 1 N from manure effluent and returned 172 kg ha 1 N of A bruzzi rye crop uptake. The rye
102 ryegrass harvested crop uptake (79 kg ha 1 N) at the Dairy Unit was lower than Van The mineral N (nitrate N plus ammonium N) available for crop uptake ( 19 6 kg ha 1 ) was from depletion of soil mineral N (25 kg ha 1 the diff erence between initial and final soil mineral N), estimated mineralized N ( 145.8 kg ha 1 ), inorganic fertilizer (24 kg ha 1 N ), and atmospheric deposition (0.8 kg ha 1 N) Rye ryegrass total crop uptake was 128.3 kg ha 1 N, leaving 67.3 kg ha 1 N subject to leaching or gaseous losses. The NUE of the winter rye ryegrass was calculated by dividing 79 kg ha 1 N crop uptake by the manure effluent N application of 104 kg ha 1 TK N plus the inorganic N fertilizer application of 24 kg ha 1 N. The c alcul ated NUE of the winter rye ryegrass was 62%. Raun and Johnson (1999) reported a worldwide NUE estimate for rye as 33%. Because mineralized N provided the majority of crop available mineral N, the NUE was difficult to evalu ate without an unfertilized rye ry egrass plot to compare to. NUE would have been lower if the yield from an unfertilized plot had been subtracted from the numerator of the calculation. Leaching All 12 lysimeters yielded observations during the winter crop season. The mean of these loads wa s then calculated to obtain the winter leaching total load. The mean leaching load for the winter season inc luding the fallow period was 1.9 kg ha 1 nitrate N ( Table 3 20 ). Two leaching events were observed during the fallow period on 2 Oct. 2011 (0.05 kg ha 1 nitrate N) and 13 Oct. 2011 (0.83 kg ha 1 nitrate N) (Figure 3 3). Although leaching events were more frequent in the winter season as seen in Figure 3 3, the nitrate N load leached was similar to the spring season when there was only one leaching eve nt. The mean nitrate N concentration of leaching events during the winter
103 season was 62 mg L 1 (Table 3 21). The nitrate N concentrations were variable with a range of 0.01 to 164.21 mg L 1 Rainfall over the winter season was 413 mm ( Figure 3 9 ) with one high rainfall event on 20 January 2012 (137 mm) (Figure 3 3), but overall quantity of leachate recorded was low compared to the summer season. Rainfall during the winter season was 39 mm above the 5 year historical mean for the same time period 21 Sept embe r to 12 Mar ch ( Figure 3 9 ). A total of 118 mm of fresh water was applied to Field J through the center pivot irrigation system (Figure 3 3). The mean fresh water applied on a single day was 10 mm. This amount was the same as the mean rainfall event on a single day during the summer season (10 mm). Fresh water was primarily applied later in the summer season with almost all applications in Jan uary and Feb ruary Fresh water was applied in larger quantities compared to rainfall events (5 mm mean excluding the 136 mm rainfall event on 20 Jan. 2012). It is likely that fresh water application contributed to leaching, but the leachate N concentration was low and so total N leaching was less than 1% of N outputs for the winter N mass balance. During the fallow season, evaporation from the soil was approximately 14 mm total over the 38 day fallow period. During the winter rye ryegrass crop season the daily reference evapotranspiration rate was 1.5 mm day 1 in Alachua County ( FAWN, 2012 ). Takin g into account pasture crop coefficients of 0.8 for the season t otal evapotranspiration during the winter season was 164 mm (SCS, 1967). Freshwater irrigation provided 118 mm of water in addition to 325 mm of rainfall. Approximately 280 mm of water was ap plied above the crop needs and was subject to leaching.
104 removing N and water from the soil. The 128 kg ha 1 N total crop uptake left 60 kg ha 1 mineral N (of the available 18 9 kg ha 1 mineral N) subject to leaching and gaseous losses. The leaching losses (1.9 kg ha 1 total N) only accounted for a sma ll portion of the left over 60 kg ha 1 mineral N. It must be concluded that the majority of left over mineral N was lost to gaseous losses. During the fallow period fr om 21 Sept. to 28 Oct. 2011, 0.9 kg ha 1 N was leached. There was 103 mm of rainfall during this time period. No freshwater irrigation or manure effluent applications were made during the fallow period as per the nutrie nt actively growing crops. The last manure effluent application during the summer season was on 3 Aug. 2011, 47 days before the summer silage corn was harvested. The hypo effluent was not applied late in the growing season. osses of N are mostly associated with rainfall after the crop is harvested and not with excessive irrigation during the gro rejected Leaching losses during the fallow period were only 4% of the total leaching losses for the 2011 2012 crop system. 96% of N leaching occurred during the growing season. It does appear that rainfall was the main cause of leaching during the fallow period since no freshwater irrigation or manure effluent was applied. Unaccounted for N The inputs and measured out puts of the winter season were used to calculate unaccounted for N in the winter N mass budget. Unaccounted for N was calculated by
105 subtracti ng the measured outputs of final soil mineral N, harvested crop uptake, root and stubble uptake, and leaching from the inputs of initial soil m ineral N, mineralized N inorganic fertilizer, and atmospheric deposition. Unaccounted for N was 66 kg ha 1 N in the winter rye/ryegrass season ( Table 3 30 ). Unaccounted for N was composed of gaseous losses from volatilization a nd denitrification Typically, volatilization slows during the winter crop season due to cooler temperatures (Van Horn et al., 1996). Farm mana gement at the Dairy Unit used the same estimate for gaseous losses (40%) from manure effluent application year ro und. The Dairy Unit estimated 4 2 kg ha 1 N of manure effluent would be lost to gaseous losses during the winter season. Unaccounted for N (6 6 kg ha 1 ) was slightly more than Rye ryegrass available min eral N for crop u ptake was composed of min eral N from initial soil mineral N (5 3 kg ha 1 ), mineralized N ( 14 5 kg ha 1 ), inorganic fertilizer (24 kg ha 1 ) and atmospheric deposition (0.8 kg ha 1 ) Of the known available mineral N, 128 kg ha 1 mineral N was taken up by crops and 2 8 kg ha 1 mineral N remained in the soil, ve ry little of it was leached ( 2 kg ha 1 nitrate N). Using these data, approxima tely 6 6 kg ha 1 mineral N was lost to the environment by gaseous losses and was unaccounted for N Assuming all unaccounted for N was gaseou s losses they w ere 29 % of the N total outputs ( 22 4 kg ha 1 ) for the winter N mass bud get. This estimation was slightly higher than N mass budget on a Pennsylvania dairy farm. Unaccounted for gaseous losses were within the range Van Horn et al. (1996) estimate d 50 to 70% for volatilization from
106 animal manures. Since greater than 40% of manure effluent applications were gaseous losses then we reject the greater than the IFAS recommendation due to overestimated gaseous losses. Farm management assumes 40% losses. This leads to an over fertilization of the crops and leaching of N to groundwater Gaseous losses were greater than 40% during the winter season. Farm management underestimated gaseous losses in the winter season, but instead of over fertilization, soil N was depleted and leaching was minimal. Mineralized N over the wi nter season provided substantial mineral N for crop uptake. 2011 2012 Cropping System Mass Balance N Inputs The N mass balance for the 2011 2012 cropping system is sh own in Table 3 31 Crop available N inputs over the 2011 2012 cropping system were 28 6 kg ha 1 mineral N from the soil, 407 kg ha 1 N from m ineralized N 138 kg ha 1 N fro m inorganic fertilizer, plus 1.7 kg ha 1 N from atmospheric deposition ( Table 3 31 ). The initial soil mineral N content was higher than expected because soil samples were tak en a month after the spring silage corn was planted. Manure effluent (74 kg ha 1 TKN) and inorganic fertilizer (35 kg ha 1 N) were applied to the crop before soil sampling, increasing the initial soil mineral N content. Van Horn et al. (1996) found soil mi neral N content was 57 kg ha 1 N for the medium manure application rate of 600 kg ha 1 N per year. The findings of Van 8 kg ha 1 content was more than 7 times the mineral N found by Van Horn et al. (57 kg ha 1 N). Using an estimation of 40% losses from application of manure effluent, the farm management determined the total N fertilizer applied was 430 kg ha 1 N. Farm
107 management con sidered 40% to be a conservative estimate from the 45 50% volatilization loss estimation specified in their comprehensive nutrient management plan (Martin, 2000 ) N fertilizer application fell within the UF IFAS recommended guidelines for yearly application of N fertilizer to irrigated silage corn of 470 kg ha 1 N. Spring N fertilizer application (270 kg ha 1 N) exceeded the IFAS recommendation for an irrigated corn crop (235 kg ha 1 N). More equal distribution of N fertilizers over the two silage corn crops might i ncrease the summer crop yields. Inorganic N ferti lizer was 17 % of N inputs in the N mass balance ( Table 3 31 ). Reducing inorganic N fertilizer could have a cost benefit to farm management since it is the only N input which the Dairy Unit must purchase from off the farm. Whereas the cost of application of manure effluent for forage corn production is only determined by the overhead expense of operating the irrigation system, the cost of inorganic N fertilizer is determined by the variable costs of the amount of fertilizer purchased, equipment costs to apply the fertilizer, and the number of applications. During the soil N m ineralization experiments, 407 kg ha 1 mineralized N was a measured N input during the 2011 2012 crop system. A lack of fertilizer or manure effluent additions for long time periods could result in little to no mineralized N available for plant uptake. This was seen during Oct ober in the summer mineralization experiment. Atmospheric deposition had very little impact on the N inputs of the cropping system accounting for less than 1% of total N inputs ( Table 3 31 ). The mean monthly mineralized N from the summer and winter mineralization experiments was 3 8 kg ha 1 mineral N. Using this estimate, the mineralized N for the spr in g season would have been 125 kg ha 1 mineral N (approximately 3.3 months in the
108 spring season). This was less than the estimated mineralized N (204 kg ha 1 ammonium N) from the spring season used in the spring N mass budget. Manure effluent applications du ring the spring season were much larger than either the summer or the winter applications. It would be expected that mineralized N for the spring season would be greater than what was seen in the summer and winter mineralization experiments for this reason Although mineralized N was highly variable and dependent on many changing climactic factors, the estimated mineralized N used for the spring season seems reasonable. N Outputs During the 2011 2012 cropping system, t he re was a build up of soil TK N conten t due to manure applications, but soil mineral N content decreased over the crop season. The final soil mineral N content was 2 8 kg ha 1 N ( Table 3 31 ) which meant 258 kg ha 1 N was removed by crop uptake, leaching, and gaseous losses. The Dairy Unit appli ed 626 kg ha 1 total N fertilizers over the entire corn corn rye and ryegrass cropping system ( Table 3 31 ). The total harvested crop uptake was 423 kg ha 1 N. Dry matter yields were less than those measured by the 2011 Corn Silage Field Day Corn Hybrid Variety Test (UF DAS, 2011). The silage corn dry harvested yield at the Dairy Unit was 15.1 Mg ha 1 versus the mean yield of the variety trial of 20.2 Mg ha 1 for the spring and 11.7 Mg ha 1 versus the mean yield of the variety trial of 14.8 Mg ha 1 for the summer. (423 kg ha 1 N harvested crop uptake/ 626 kg ha 1 N fertilizers). The tot al plant crop uptake (harvested crop uptake plus root and stubble uptake) was 554 kg ha 1 N, representing 51 % of the total outputs. In a corn corn rye and clover cropping system in Tifton GA, Newton et al. (2003) applied 560 kg ha 1 N of dairy manure and commercial
109 fertilizer and harvested 384 kg ha 1 N total crop uptake. They reported a 69% recovery of N fertilizer applied very similar to the results found at the Dairy Unit. The NUE (68%) for the 2011 2012 cropping system was close to apparent N recovery (72%) at the optimal N fertilizer rate found by Meng et al. (2012). Harvested crop uptake (423 kg ha 1 N) during the 2011 2012 cropping system was greater than the N uptake estimated by the 1 N for corn silage corn silage rye) (Martin, 2000) Leaching over the 2011 2012 cropping system was a small part of the total N mass balance, only 3 % of N outputs ( Table 3 31 ). The mean rainfall in the spring and summer seasons was below average and resulted in fewer leaching events. Average to high rainfall would likely have caused much greater N leaching especially during the spring season when large amounts of manure effluent were applied. Although rainfall during the winter 2011 2012 season was above the historical 5 year average ( Figure 3 9 ), N leaching was 2 kg ha 1 N. During the fallow period from 21 Sept. to 28 Oct. 2011, there was little leaching most likely due to the low mineral N content of the soil (5 3 kg ha 1 N) and the low manure effluent application during the summer season (31 kg ha 1 N). If manure effluent applications were increased over the summer season, leaching during the fallow period could increase. Figure 3 3 shows daily rainfall, freshwater irriga tion, manure effluent applications, and leaching events. Yearly leaching was determined to be 2 3 kg ha 1 nitrate N ( Table 3 31 ). In France, Constantin et al. (2010) found a similar mean N leachate of 27 kg ha 1 yr 1 N. They reported leaching to be highly v ariable, ranging from 11 to 71 kg ha 1 yr 1 N depending on climate, site, and crop type. Rotz et al. (2005) estimated leaching to be 45
110 kg ha 1 yr 1 nitrate N o n a non grazed, grass and maize silage simulated dairy farm. The estimation of leaching load by Rotz et al. (2005) is greater than the 2 3 kg ha 1 nitrate N leached during the year on Field J at the Dairy Unit. Measuring leaching over a year with higher rainfall could result in estimates similar to Rotz et al. (2005). In Germany, annual leaching rate s for arable sandy soils were 63 kg ha 1 nitrate N (Richter and Roelcke, 2000). This exceeds the annual leaching of total N found at the Dairy Unit in 2011 of 2 3 kg ha 1 nitrate N. Richter and Roelcke (2000) calculated surplus N to be the difference betwee n total inputs of inorganic fertilizer, manures, and atmospheric deposition and N removed by the harvested part of crops. The mean N surplus for arable crops in Germany in recent decades wa s 110 to 130 kg ha 1 yr 1 N. 1 yr 1 N manure effluent, 138 kg ha 1 yr 1 N inorganic fertilizer, 1.7 kg ha 1 yr 1 N atmospheric deposition) 423 kg ha 1 yr 1 N harvested cr op uptake), a N surplus of 20 5 kg ha 1 yr 1 N was calculated ( Table 3 31 ). Differences in the agricultural systems of Germany versus the Dairy Unit are evident from the estimates Richter and Roelcke (2000) make of gaseous losses. Gaseous losses are estimated to be less than 20 kg ha 1 yr 1 N for arable systems in Germany (Richter and Roelcke, 2000). The climate and soil type of northeast Florida causes gaseous losses to be a much greater output of N in the N mass balance than it is in Germany. Unaccounted for N over the entire crop year wa s 229 kg ha 1 N or 27 % of the total outputs for the N mass budget ( Table 3 31 ). The percent age of harvested crop uptake (51 %) of the 37% of mean annual N outputs from harvested crops on a Pennsylvania dairy farm.
111 They found the percentage of N leaching to be much higher (38%) th N leached (3 %) although rainfall was lower than historical averages in the spring and summer seasons in Alachu a County. Unaccounted for N ( 27 %) at the Dairy Unit was approximately the same as the estimated volatilization losses (25%) of Hall and Risser (1993). The lar ge amount of unaccounted for N during the spring season (53 % of total unaccounted for N ) appeared to be a direct result of the high manure effluent application during that season. Reducing manure effluent application would likely have resulte d in less unaccounted for N without impacting silage corn yield. During the winter mineralization experiment, greater manure effluent applica tion (104 kg ha 1 N) resulted in more ( 14 6 kg ha 1 ) mineralized N as opposed to the summer experiment which had 31 kg ha 1 N of manure effl uent applied and resulted in 57 kg ha 1 mineralized N. Spring crop loading of manure effluent above the crop requirement did not increase soil mineral N or significantly increase soil TKN. Therefore, N from applied manure effluent was su bject to gaseous losses since little N was found in leachi ng.
112 Table 3 1. Physical and chemical properties of topsoil at lysimeter locations on 14 June 2012 Lysimeter Location A B C D E F Mean pH w 7.4 7.3 7.6 7.7 7.4 7.7 7.5 % OM x 2.44 2.1 1.76 1.22 1.83 2.04 1.90 EC y dS m 1 0.13 0.11 0.12 0.10 0.09 0.10 0.11 Mehlich 1 z P, mg kg 1 719 796 860 558 615 605 692 Mehlich 1 z K, mg kg 1 129 88 93 124 103 123 110 Mehlich 1 z Mg, mg kg 1 316 287 322 281 305 373 314 Mehlich 1 z Ca, mg kg 1 1944 2016 2057 1553 1755 2197 1920 Mehlich 1 z Cu, mg kg 1 0.07 0.16 0.11 0.12 0.11 0.02 0.10 Mehlich 1 z Mn, mg kg 1 19.75 20.28 17.03 16.66 18.88 22.96 19.26 Mehlich 1 z Zn, mg kg 1 8.75 10.46 11.28 10.04 8.74 8.27 9.59 w pH (1:2 soil: deionized H 2 O ratio) x OM Walkley Black Method (Mylavarapu, 2002) y EC (1:2 soil:water) soluble salts z Mehlich 1 (Mehlich, 1953) Table 3 2. Bulk density of soil samples (0 60 cm) on four sampling dates Lysimeter Locations Field Mean STDEV T ime g cm 3 4/14/2011 1.4 0.14 7/1/2011 1.5 0.16 9/29/2011 1.5 0.17 3/16/2012 1.5 0.12 Table 3 3. Bulk density of soil samples by depth Field Mean STDEV Depth g cm 3 0 30 cm 1.5 0.17 30 60 cm 1.6 0.13 P < 0.0001 Table 3 4. Soil mineral N (nitrate N plus ammonium N) content in the 0 to 60 cm soil profile by season Initial 0 60 (cm) STDEV Final 0 60 (cm) STDEV P N (kg ha 1 ) N (kg ha 1 ) Spring 28 6 4 2 165 5 8 < 0.0001 Summer 165 5 8 5 3 2 2 < 0.0001 Winter 5 3 2 2 2 8 1 4 < 0.0001
113 Table 3 5 Soil mineral N (nitrate N plus ammonium N) content divided into 30 cm increments by season Initial STDEV Final STDEV P N (kg ha 1 ) N (kg ha 1 ) --------------------------------0 30 (cm) -------------------------------Spring 161 23 87 40 < 0.0001 Summer 87 40 33 13 < 0.0001 Winter 33 13 21 12 0.0002 -----------------------------30 60 (cm) -----------------------------Spring 125 22 78 28 < 0.0001 Summer 78 28 20 12 < 0.0001 Winter 20 12 7 2 < 0.0001 Table 3 6 Soil nitrate N content in the 0 to 60 cm soil profile by season Initial 0 60 (cm) STDEV Final 0 60 (cm) STDEV P kg ha 1 kg ha 1 Spring 113 18 100 4 7 NS Summer 100 4 7 47 21 < 0.0001 Winter 47 21 20 1 3 < 0.0001 Table 3 7 Soil ammonium N content in the 0 to 60 cm soil profile by season Initial 0 60 (cm) STDEV Final 0 60 (cm) STDEV P kg ha 1 kg ha 1 Spring 17 3 38 6 6 23 < 0.0001 Summer 6 6 23 6 2 < 0.0001 Winter 6 2 8 3 < 0.0001 Table 3 8 Soil TKN content in the 0 to 60 cm soil profile by season Initial 0 60 (cm) STDEV Final 0 60 (cm) STDEV P kg ha 1 kg ha 1 Spring 3310 66 0 343 0 700 NS Summer 343 0 700 346 0 5 40 NS Winter 346 0 540 346 0 5 50 NS
114 Table 3 9 Soil nitrate N content in 0 to 30 cm and 30 to 60 cm soil profile by sampling date 0 30 (cm) STDEV 30 60 (cm) STDEV P kg ha 1 kg ha 1 4/14/2011 73 10 40 14 < 0.0001 7/1/2011 5 6 32 44 27 < 0.0001 9/29/2011 30 13 18 11 < 0.0001 3/16/2012 16 12 4 2 < 0.0001 Table 3 10 Soil ammonium N content in 0 to 30 cm and 30 to 60 cm soil profile by sampling date 0 30 (cm) STDEV 30 60 (cm) STDEV P kg ha 1 kg ha 1 4/14/2011 8 8 20 85 2 1 NS 7/1/2011 3 1 13 3 5 13 NS 9/29/2011 3 1 2 1 NS 3/16/2012 5 2 3 1 NS Table 3 11 Soil TKN content in 0 to 30 cm and 30 to 60 cm soil profile by sampling date 0 30 (cm) STDEV 30 60 (cm) STDEV P kg ha 1 kg ha 1 4/14/2011 241 0 530 900 18 0 < 0.0001 7/1/2011 2370 6 80 106 0 3 20 < 0.0001 9/29/2011 2350 48 0 111 0 2 50 < 0.0001 3/16/2012 256 0 380 900 2 50 < 0.0001 Table 3 12 Manure effluent N (TKN) application by month Spring TKN Summer TKN Winter TKN kg ha 1 kg ha 1 kg ha 1 March 46 July 9 October 0 April 91 August 22 November 68 May 142 September 0 December 34 June 74 January 0 February 2 Spring Total 353 Summer Total 31 Winter Total 104
115 Table 3 13 Inorganic N fertilizer (Total N) application by month Spring N Summer N Winter N kg ha 1 kg ha 1 kg ha 1 March 35 July 34 October 0 April 24 August 22 November 0 May 0 September 0 December 24 June 0 January 0 February 0 Spring Total 5 9 Summer Total 56 Winter Total 24 Table 3 14 Wet ion deposition by season (NADP, 2011 ) NH4+ NO3 Total N --------kg ha 1 ---------Spring 0.2 0.3 0.5 Summer 0.1 0.3 0.4 Winter 0.2 0.4 0.7 Table 3 15 Crop dry weight yield by season Mean Yield STDEV Mg ha 1 Spring Sampling 15.2 3.1 Dairy Unit 8.8 Summer Sampling 11.8 1.0 Dairy Unit 7.6 Winter Sampling 4.1 0.8 Dairy Unit 2.4 Note : Sampling yields were calculated from harvested portion of plants sampled. Dairy Unit yield was measured by farm management at commercial harvest. Silage corn was grown during the spring and summer seasons. A rye/ryegrass cover crop was grown during the winter season. Table 3 16 University of Florida corn silage variety trial dry matter yields and crude protein concentration in 20 1 1 (UF DAS, 2011) Mean Yield STDEV Crude Protein STDEV Mg ha 1 % Spring 20.2 1.1 8.00 0.22 Summer 14.8 0.9 7.34 0.36 Note : Corn silage variety trials were conducted at the Plant Science Research and Education Unit in Citra, FL. Corn was fertilized with 235 kg ha 1 N and harvested at 35% dry matter. For the spring season, corn was planted on 16 Mar. 2011 and harvest dates were from 24 to 30 June 2011. For the summer season, corn was planted on 13 July 2011 and harvest dates were from 7 to 19 Oct. 2011.
116 Table 3 17 Mean N content of dry silage corn plant parts by season Spring STDEV Summer STDEV P N (kg ha 1 ) N (kg ha 1 ) Leaves 68.6 18.6 42.1 3.7 < 0.0001 Stalks 33.1 6.8 13.4 3.9 < 0.0001 Ears 99.9 19.6 86.6 9.5 0.0461 Stubble 17.3 4.6 3.9 2.7 < 0.0001 Roots 44.9 11.4 15.6 3.3 < 0.0001 Table 3 18. Mean N concentration of silage corn plant parts by season Spring STDEV Summer STDEV P %N %N Leaves 2.18 0.23 1.77 0.23 0.0002 Stalks 0.98 0.16 0.47 0. 15 < 0.0001 Ears 1.17 0.07 1.34 0.06 < 0.0001 Stubble 2.01 0.50 0.49 0.31 < 0.0001 Roots 1.40 0.20 0.85 0.14 < 0.0001 Table 3 19 Harvested crop uptake and root and stubble uptake (N kg ha 1 ) on Field J at the Dairy Unit by season Harvested Crop Uptake STDEV Root and Stubble Uptake STDEV N (kg ha 1 ) N (kg ha 1 ) Spring 20 2 38 62 14 Summer 142 14 20 5 Winter 79 16 4 9 12 Table 3 20 Sum of leachate loads (N kg ha 1 ) over 2011 2012 season Leachate Load STDEV N (kg ha 1 ) Spring 1.9 3.5 Summer 18.8 20.2 Winter 1.9 2.2 Table 3 21 Nitrate N concentration (mg L 1 ) of leaching events by season Nitrate N Concentration mg L 1 STDEV Range Spring 34.63 26.74 1.66 57.66 Summer 65.84 37.19 10.30 148.79 Winter 62.44 42.19 0.01 164.21
117 Table 3 22 N mass balance of spring silage corn crop at the Dairy Unit Inputs Outputs N (kg ha 1 ) N (kg ha 1 ) Initial Soil Mineral N 28 6 Final Soil Mineral N 165 Mineralized N 204 Harvested Crop Uptake 20 2 Inorganic N Fertilizer 5 9 Root and Stubble U ptake 62 Atmospheric Deposition 1 Leaching nitrate N Load 2 Unaccounted For N 11 9 Table 3 23 Soil mineral N (nitrate N plus ammonium N) content of each location over three time periods for the summer mineralization experiment on Field J at the Dairy Unit Summer Soil Mineral N Content Field Mean STDEV N (kg ha 1 ) Installed July Pulled Aug. 6 4 25 Pulled Sept. 5 3 3 8 Pulled Oct. 71 39 Installed Aug. Pulled Sept. 60 2 7 Pulled Oct. 84 28 Installed Sept. Pulled Oct. 56 18 N ote: Installation and removal of PVC pipes was on the 24 th day of each month. 9 kg ha 1 N of manure effluent was added in the month before 24 July 2011. 22 kg ha 1 N of manure effluent was added between 24 July and 24 Aug. 2011. No manure effluent was added after 24 Aug. 2011 for the remainder of the summer mineralization experiment. Soil Mi neral N contents represent the kg ha 1 N in the upper 60 cm of the soil profile with a soil bulk density of 1.5 g cm 3
118 Table 3 24 Change in soil mineral N content over three time periods for the summer mineralization experiment on Field J at the Dairy Unit Change in Soil Mineral N Content Field Mean Installed July N (kg ha 1 ) Pulled Aug. 6 4 Pulled Sept. 5 3 11 Baseline for Installed Aug. Pulled Oct. 71 +18 Baseline for Installed Aug. Mineralized N 18 Installed Aug. Pulled Sept. 60 8 +1 9 Pulled Oct. 84 2 4 +5 Baseline for Installed Sept. Mineralized N 24 Installed Sept. Pulled Oct. 56 2 8 33 Mineralized N 0 Net Mineralization 4 3 Note : Installation and removal of PVC pipes was on the 24 th day of each month. Baselines are the change from the previous installed set. Positive incremental changes denote net mineralization for the time period. Negative incremental changes denote net immobilization for the time period Net Mineralization is the sum of positive incremental changes.
119 Table 3 25 N mass balance of summer silage corn crop at the Dairy Unit Inputs Outputs N (kg ha 1 ) N (kg ha 1 ) Initial Soil Mineral N 165 Final Soil Mineral N 5 3 M ineralized N 57 Harvested Crop Uptake 142 Inorganic N Fertilizer 56 Roots and Stubble Uptake 20 Atmospheric Deposition 0 Leaching nitrate N Load 1 9 Unaccounted For N 4 4 Note : Atmospheric deposition was calculated to be 0.4 kg ha 1 N from estimates of wet deposition from the National Atmospheric Deposition Program (NADP, 2011). Table 3 26 Soil mineral N content of each location over three time periods for the winter mineralization experiment on Field J at the Dairy Unit Soil Mineral N Content Field Mean STDEV N (kg ha 1 ) Installed Nov Pulled Nov 70 2 2 Pulled Dec 77 36 Pulled Jan 5 9 28 Pulled Feb. 41 1 2 Installed Dec Pulled Dec 43 3 4 Pulled Jan 11 9 55 Pulled Feb. 6 7 46 Installed Jan Pulled Jan 4 8 1 2 Pulled Feb 57 2 2 Note : Installation and removal of PVC pipes was on the 8 th day of each month with the exception of 9 Feb. 2011. 20 kg ha 1 N of manure effluent was added in the month before 8 Nov. 2011. 82 kg ha 1 N of manure effluent was added between 8 Nov. and 8 Dec. 2011. No manure effluent was added between 8 Dec. and 8 Jan. 2011. 2 kg ha 1 N of manure effluent was added between 8 Jan. and 9 Feb. 2011. Soil Mineral N contents represent the kg ha 1 N in the upper 60 cm of the soil profile with a soil bulk density of 1.5 g cm 3
120 Table 3 27 Change in the soil mineral N content over three time periods for the winter mineralization e xperiment on Field J at the Dairy Unit Change in Soil Mineral N Content Field Mean Installed Nov. N (kg ha 1 ) Pulled Nov. 70 Pulled Dec. 77 +7 Pulled Jan. 5 9 1 9 Baseline for Installed Dec. Pulled Feb. 41 17 Baseline for Installed Dec. Mineralized N 7 Installed Dec. Pulled Dec. 43 3 4 41 Pulled Jan. 11 9 7 6 +94 Baseline for Installed Jan. Pulled Feb. 6 7 52 3 5 Baseline for Installed Jan. Mineralized N 94 Installed Jan. Pulled Jan. 4 8 71 165 Pulled Feb. 57 10 +44 Mineralized N 44 Net Mineralization 14 6 Note : Installation and removal of PVC pipes was on the 8 th day of each month with the exception of 9 Feb. 2011. Baselines are the mineral N change from the previous set during the same time period. Positive incremental changes denote net mineralization for the time period. Negative incremental changes denote net immobilization for the time period. Net Mineralization is the sum of positive incremental cha nges.
121 Table 3 28 N content of rye/ryegrass plant parts Total N STDEV N (kg ha 1 ) Harvested 79 16 Stubble 2 2 6 Roots 27 9 Table 3 29 Mean percent N concentration of rye ryegrass plant parts Harvested STDEV Stubble STDEV Roots STDEV % N % N % N Winter 1.93 0.22 1.69 0.25 1.34 0.17 Table 3 30 N mass balance of winter rye/ryegrass crop at the Dairy Unit Inputs Outputs N (kg ha 1 ) N (kg ha 1 ) Initial Soil Mineral N 5 3 Final Soil Mineral N 2 8 Mineralized N 14 6 Harvested Crop Uptake 79 Inorganic N Fertilizer 24 Roots and Stubble Uptake 49 Atmospheric Deposition 1 Leaching nitrate N Load 2 Unaccounted For N 66 Table 3 31 Overall N mass balance of 2011 2012 corn corn rye/ryegrass cropping system at the Dairy Unit Inputs Outputs N (kg ha 1 ) N (kg ha 1 ) Initial Soil Mineral N 28 6 Final Soil Mineral N 2 8 Mineralized N 40 7 Harvested Crop Uptake 423 Inorganic N Fertilizer 13 9 Root and Stubble Uptake 13 1 Atmospheric Deposition 2 Leaching nitrate N Load 2 3 Unaccounted For N 22 9
122 Figure 3 1. Manure effluent daily N application 0 5 10 15 20 25 30 35 40 2/26/11 4/17/11 6/6/11 7/26/11 9/14/11 11/3/11 12/23/11 2/11/12 Daily Manure Effluent Application (kg ha 1 ) Date Manure effluent N (kg ha 1)
123 Figure 3 2. Historical rainfall by year (2007 2011) from March 15 to J une 25 in Alachua county ( FAWN, 2012 )
124 Figure 3 3. N leaching versus rainfall, fresh water irrigation, and manure effluent application (kg ha 1 ) 1/7/11 2/26/11 4/17/11 6/6/11 7/26/11 9/14/11 11/3/11 12/23/11 2/11/12 4/1/12 5/21/12 0 5 10 15 20 25 30 35 40 0 25 50 75 100 125 150 Leaching and Manure Effluent Application (kg ha 1 ) Rainfall and Freshwater Irrigation (mm) Date Rainfall (mm) Freshwater irrigation (mm) Leaching N (kg ha 1) Manure effluent N (kg ha 1)
125 Figure 3 4. Historical rainfall by year (2007 2011) from June 2 to September 21 in Alachua county ( FAWN, 2012 ) Figure 3 5 Net mineralization/immobilization of summer mineralization experiment
126 Figure 3 6 Mineral N content of summer mineralization experiment PVC pipes installed on 24 July 2011 Note : Error bars denote standard error of the mean. Figure 3 7 Net mineralization/immobilization of winter mineralization experiment 0 20 40 60 80 0 30 60 90 Soil Mineral N (kg ha 1 ) Days after installation PVC pipes installed 24 July 2011
127 Figure 3 8 Mineral N concentrations of winter mineralization experiment PVC pipes installed at 8 Nov. 2011 Note : Erro r bars denote standard error of the mean. Figure 3 9 Historical rainfall by year (2007 2012) from September 21 to March 12 in Alachua county (FAWN, 2012 ) 0 20 40 60 80 0 30 60 90 Soil mineral N (kg ha 1 ) Days after installation PVC pipes installed 8 Nov. 2011
128 CHAPTER 4 CONCLUSIONS In this study, quantifying the N mass balance of Field J revealed that leaching was not a substantial source of N losses to the environment during this period of below average rainfall N lost to leaching was 2 3 kg ha 1 nitrate N for the sampling year. Leaching events appeared to be caused most often by high rainfall events and not manure effluent application or freshwater irrigation. Although farm management took into account weather forecasts when applying manure effluent, rainfa ll was unpredictable. Therefore occasional leaching events were unavoidable, especially during the summer months. Leaching losse s during the fallow periods were 25 % of leaching over the crop year and not a majority of total leaching Farm management of man ure effluent applications and freshwater irrigation during the crop season is critical to minimize leaching and should continue to be managed carefully Initially, one concern of correctly estimating N fertilizer applications was the possibility of overest imation of gaseous losses. Unaccounted for N in this study was assumed to be comprised entirely of gaseous losses (volatilization and denitrification) because all other sources of N were accounted for Unaccounted for N was ve ry high in the spring season ( 11 9 kg ha 1 N ). High manure effluent application in the spring season accompanied by warm temperatures likely caused large amounts of N to be lost to volatilization making it difficult to correctly calculate N fertilizer application. This loss pathway may be expected due to the high pH of the soil (>7.0) and the significant amounts of ammonium N applied in the manure effluent. Since gaseous losses were 45 % of manure effluent applications (488 kg ha 1 TKN) s 40% N loss as sumption was a conservative
129 estimate of gaseous losses Therefo re, N fertilizer application was within the IFAS recommendation of 470 kg ha 1 yr 1 under current management practices. N fertilizer applications (29 4 kg ha 1 ) for the spring silage corn crop ex ceeded I FAS recommendations for the season (235 kg ha 1 N) and ga seous losses (119 kg ha 1 N) during the spring season were less than 40% of manure effluent inputs (353 kg ha 1 TKN). D uring the summer season, gaseous losses (44 kg ha 1 N) were greater than 100% of manure effluent inputs (31 kg ha 1 TKN). Although N fertilizer recommendations and gaseous loss assumptions were made for the crop year as a whole, seasonal application and losses were highly variable and may deserve individua l consideration by farm management. Taking into account the high F NUE (163%) of the summer season silage corn, it is likely that the summer silage corn yield would have been higher with increased N application. As increasing N fertilizer through manure ef fluent application in the summer would likely result in high volatilization losses, applying ammonium free inorganic N fertilizer during the summer months would be a better alternative to improve yields. As found in this study, the large amount of manure effluent applied during the spring season mineralized and provided mineral N throughout the crop year. Crop uptake of N is greatest approximately 40 to 60 days after corn is planted. Therefore, inorganic fertilizer applications at planting may not be neede d or should be applied at a more appropriate time for crop uptake. Also, manure effluent applications (74 kg ha 1 TKN) in the first month of the spring season and continued manure effluent application may have supplied adequate N to crops without a second application of inorganic fertilizer during the spring season. Shifting supplemental inorganic fertilizer applied
130 during the spring season to the summer season would not increase costs to the dairy but would likely increase summer silage corn yields while h aving little effect on spring corn yields. The summer and winter mineralization experiments provided an estimate of the amount of mineral N made available through mineralization of manure effluent and organic N in the soil. T he mineralized N percentage of crop N inputs varied greatly between the summer and winter experiments. Measuring possible volatilization of mineral N from mineralization experiment soils would more accurately describe all mineralized N. Repeating the experiments and including multiple l evels of manure effluent treatments would improve understanding of how soil N level changes based on N application. Measuring soil moisture content and taking into account evapotranspiration rates would lead to a greater understanding of when the combinat ion of manure effluent, freshwater irrigation, and rainfall result in leaching events. Measuring leaching in years with average to above average rainfall would provide a better understanding of whether leaching increases and may be a greater concern than i t was in this study. Although leaching losses were a small portion of the overall budget, the nitrate N concentration of leachate was above the MC L so leaching is a concern for water quality. Further study of gaseous losses would provide a better character ization of unaccounted for N. Volatilization losses are maximized in conditions with surface application of manures, high pH, soils with low cation exchange capacity, and warm, slightly moist environments (Pierzynski et al., 2000). All of these conditions apply to the forage production system at the Dairy Unit. In Field J, somewhat poorly drained soils
131 and surface applied manure effluent applications were conditions conducive to anaerobic activity and therefore possible denitrification. Determination of vol atilization versus denitrification losses would give farm management better information to combat N losses with possible use of urease and nitrification transformation inhibitors. Finally, i t is recommended that farm management change their practices for N fertilizer applic ation by more evenly distributing applications throughout the year as opposed to applying heavily concentrated fertilization in the spring season. A lso, a dopting seasonal estimations of gaseous losses instead of using a 40% annual a verage would more accurately estimate N losses based on climate conditions that vary throughout the year. In addition, m ore investigation into volatilization and denitrification would provide a better understanding of potential environmental losses
132 LIST O F REFERENCES Altom, W., J.L. Rogers, W.R. Raun, and W.E. Thompson. 2002. Changes in total inorganic profile nitrogen in long term rye wheat ryegrass forage production system. J. Plant Nutrition. 25:2285 2294. Autry, L. P. 2003. National Environmental Laboratory Accreditation Conference: Constitution, Bylaws, and Standards EPA/600/R 04/003. (TIP # 04 009, Published Report.) L. Autry, PO. Azeez, J.O. and W. Van Averbeke. 2010. Nitrogen mineralization potential of three animal manures applied on a sandy clay loam soil. Bioresour. Technol. 101:5645 5651. B acon S., L. L anyon and R. S chlauder 1990. Plant nutrient flow in the managed pathways of an intensive dairy farm. Agron. J. 82:755 761. Brady N. C. and R. R. Weil. 2008. The Nature and Properties of Soils. 14 th ed. Pearson Prentice Hall, Pearson Education Inc. Upper Saddl e River, New Jersey. Bremner, J.M. and C.S. Mulvaney, 1982. Nitrogen. In: A.L. Page et al., editors, Methods of Soil A nalysis. Agron. Monogr 9. Part 2. 2 nd ed. ASA and SSSA Madison, WI. p. 595 624. Cabrera, V.E., N.E. Breuer, P.E. Hildebrand, and D. Letson. 2005a. The dynamic north Florida dairy farm model: A user friendly computerized tool for increasing profits while minimizing N leaching under varying conditions. Comput. Electron. Agric. 49:286 308. Cabrera, F., P. Martin Olmedo, R. Lopez, and J.M. Murillo. 2005 b Nitrogen mineralization in soils amended with compo sted olive mill sludge. Nutr. Cycl. Agroecosys. 71:249 258. Constantin, J., B. Mary, F. Laurent, G. Aubrion A. Fontaine, P. Kerveillant and N. Beaudoin. 2010. Effects of catch crops, no till and reduced nitrogen fertilization on nitrogen leaching and balance in three long term experiments. Agric. Ecosyst. Environ. 135:268 278. Cusick, P.R., J.M. Powell, K.A. Kelling, R.F. Hensler and G.R. Munoz. 2006. Dairy manure N mineralization estimates from incubations and litterbags. Biol. Fertility Soils 43:145 152. De Vries, A. and R. Giesy. 2009. Florida Dairy Farm Situation in 2009. EDIS Factsheet AN 215.
133 Dou, Z., R .A. Kohn, J.D. Ferguson, R.C. Boston and J.D. Newbold. 1996. Managing nitrogen on dairy farms: An integrated approach .1. model description. J. Dairy Sci. 79:2071 2080. Duan, R., C.B. Fedler and C.D. Sheppard. 2010. Nitrogen leaching losses from a wastewater land application system. Water Environ. Res. 82:227 235. Eghball, B. and J. Power. 1999. Composted and noncomposted manure application to conventional and no tillage systems: Corn yield and nitrogen uptake Agron. J. 91:819 825. [FAWN] Florida Automated Weather Network. 2012. University of Florida IFAS Extension. http://fawn.ifas.ufl.edu Follett J.A. and R.F. Delgado. 2002. Carbon and nutrient cycles. J. Soil Water Cons. 57:455 464. Follett R.F. and D.J. Walker 1989. Groundwater quality concerns about nitrogen in Nitrogen Management and Groundwater Preservation. Elsevier Science Publishers B.V., Amsterdam, Netherlands. [FDEP] Florida Department of Environmental Pro tection. 2010. Florida Administrative Code, Chapter 62 640, 670. www.dep.state.fl.us/legal/rules/rulelist.htm Frazao, L.A., M. Piccolo, B.J. Feigl, C.C. Cerri, C.E.P. Cerri 2010. Inorganic nitrogen, microbial biomass and microbial activity of a sandy Brazilian Cerrado soil u nder different land uses. Agric. Ecosys. Environ 135: 161 167. Gazula, A., E. Simonne, M. Dukes, G. Hochmuth, B. Hochmuth and B. Studstill. 2006. Opti mization of drainage lysimeter design for field determination of nutrient loads. HortScience 41:508 508. Gilbertson, C.B., D.L. Vandyne, C.J. Clanton and R.K. White. 1979. Estimating quantity and constituents in livestock and poultry manure residue as re flected by management systems. Trans. ASAE 22:602 &. Haas, G., C. Deittert and U. Koepke. 2007. Farm gate nutrient balance assessment of organic dairy farms at different intensity levels in G ermany. Renew. Agr. Food Syst. 22:223 232. Haefele, S., M. Wope reis and H. Wiechmann. 2002. Long term fertility experim ents for irrigated rice in the West African S ahel: Agronomic results. Field Crops Res. 78:119 131. Hall, D. and D. Risser 1993. Effects of agricultural nutrient management on nitrogen fate and trans port in Lancaster county, P ennsylvania. Water Resour Bull 29:55 76.
134 Hanlon E.A., J.S. Gonzalez, and J.M. Bartos. 1996. IFAS Extension Soil Testing Laboratory (ESTL) and Analytical Research Laboratory (ARL) chemical procedures and training manual. EDIS Circular 812. Hillel D. 1980 Fundamentals of soil physics. Academic Press. New York. Hutson, J., R. Pitt, R. Koelsch, J. Houser and R. Wagenet. 1998. Improving dairy farm sustainability II: Environmental losses and nutrient flows. J. Prod. Agric. 11:233 239. Jokela W. 1992. Nitrogen fertilizer and dairy manure effects on co rn yield and soil nitrate. SSSA J. 56:148 154. Klausner, S., D. Fox, C. Rasmussen, R. Pitt, T. Tylutki, P. Wright, L. Chase and W. Stone. 1998. Improving dairy farm sustainability I: An approach to animal and crop nutrient management planning. J. Prod. Agric. 11:225 233. Kohn, R., Z. Dou, J. Ferguson and R. Boston. 1997. A sensitivity analysis of nitrogen losses from dairy farms. J. Environ. Manage. 50:417 428. Korevaar H. 1992. Th e nitrogen balance on intensive Dutch dairy farms : a review Livest. Prod. Sci. Elsevier Science Publishers B.V. Amsterdam, Netherlands. 31:17 27. Kuipers, A., F. Mandersloot and R. Zom. 1999. An approach to nutrient management on dairy farms. J. Anim. Sc i. 77:84 89. M agette W., R. W eismiller J. A ngle and R. B rinsfield 1989. A nitrate groundwater standard for the 1990 farm bill. J. Soil Water Conserv. 44:491 494. M artin, J 2000. Nutrient Management Plan for UF, IFAS Dairy Research Unit Hague, FL. Florida Department of Environmental Protection, Northeast District Office, Jacksonville, FL. Mattos, D., A.K. Alva, S. Paramasivam, and D.A. Graetz. 2003. Nitrogen volatilization and mineralization in a sandy entisol of Flor ida under citrus. Commun. Soil S ci. Plan 34: 1803 1824. Mehlich, A. 1953. Determination of P, Ca, Mg, K, Na, and NH 4 Pub. No. 1 53. N.C. Soil Testing Div., Raleigh, NC. Meng, Q.F., X.P. Chen., F.S. Zhang, M.H. Cao, Z.L. Cui, J.S. Bai, S.C. Yue, S.Y. Chen, and T. Muller. 2012. In season root zone nitrogen management strategies for improving nitrogen use efficiency in high yielding maize production in China. Elsevier Science Publishers B.V. Amsterdam, Netherlands. 22:294 303.
135 Mikha, M.M., C.W. Rice and J.G. Benjamin. 2006. Estimating soil mineralizable nitrogen under diff erent management practices. SSSA J. 70:1522 1531. Morari, F., E. Lugato, R. Polese, A. Berti, and L. Giardini. 2012. Nitrate concentrations in groundwater under contrasting agricultural management practices in the l ow plains of Italy. Agric. Ecosyst. Environ 147: 47 56. Mylavarapu, R. 2002. UF/IFAS Extension Soil Testing Laboratory (ESTL) Analytical procedures and training manual. EDIS Circular 1248. Mylavarapu, R 2003. Role of an Extension Soil Testing Program in the Development of Best Management Practice s: A Florida Case Study. J. Extension. 45(4) http://www.joe.org/joe/2003august/a7.shtml Mylavarapu, R., D. Wright, G. Kidder, and C.G. Chambliss. 2009. UF/IFAS standardized fertilizer recommendations for agronomic crops. EDIS fact sheet. SL 129. Newton, G.L., J.K. Bernard, R.K. Hubbard, J.R. Allison, R.R. Lowrance, G.J. Gascho, R.N. Gates, G. Vellidis. 2003. Managing manure nutrients through multi crop forage production. J. Dairy Sci. 86:2243 2252. [NADP] National Atmospheric Deposition Program (NRSP 3). 2011. NADP Program Office, Illinois State Water Survey, Champa ign, IL. http://nadp.sws.uiuc.edu [NPDES] Na tional Pollutant Discharge Elimination System. 2011. Animal Feeding Operations. United States Environmental Protection Agency, Washington, D.C. http://cfpub.epa.gov/npdes/home.cfm?program_id=7 [ NRCS ] National Resources Conservation Service. 2010 Web Soil Survey. Washington, D.C. http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx Overcash, M.R., F.J. Humenik and J.R. Miner. 1983. Livestock waste management. Volume I. CRC Press Inc., Boca Raton, FL. [PSAG] Penn sylvania State Agronomy G uide. 1982. The Pennsylvania State Univ. College Agric. Ext Serv., University Park, PA. Pierzynski, G.M., J.T. Sims, and G.F. V ance. 2000. Soils and Environmental Quality, 2 nd ed CRC Press, CRC Press Inc., Boca Raton, FL Powell, J.M., D.B. Jackson Smith, D.F. McCrory, H. Saam and M. Mariola. 2007. Nutrient management behavior on Wisconsin dai ry farms. Agron. J. 99:211 219. Raun, W.R. and G.V. Johnson. 1999. Improving nitrogen use efficiency for cereal production. Agron. J. 91:357 363.
136 Richter, J. and M. Roelcke. 2000. The N cycle as determined by intensive agriculture examples from ce ntral Europe and China. Nutr. Cycl. Agroecos ys 57:33 46. Rotz, C., L. Satter, D. Mertens and R. Muck. 1999. Feeding strategy, nitrogen cycling, and profitability of dairy farms. J. Dairy Sci. 82:2841 2855. Rotz, C., F. Taube, M. Russelle, J. Oenema, M. Sanderson and M. Wachendorf. 2005. Whole farm perspectives of nutrient flows in grassland agriculture. Crop Sci. 45:2139 2159. SAS Institute 2002. The SAS system for windows. Release 9.2. SAS Inst. Cary, NC. [SCS] Soil Conservation Service. 1967. Irrigation Water Requirements Technical Release No. 21. USDA Soil Conservation Serv ice Engineering Division. Shepard, R. 2005. Nutrient management planning: Is it the answer to better management? J. Soil Water Conserv. 60:171 176. Shober, A. and R. Mylavarapu. 2009. Soil sampling and testing for the home landscape or vegetable garden. EDIS Document SL 281. Silva, R.G., K.C. Cameron, H.J. Di and E.E. Jorgensen. 2005. A lysimeter study to investigate the effect of dairy effluent and urea on cattle urine N losses, plant uptake and soil retention. Water Air Soil Poll ut. 164:57 78. [UF DAS] University of Florida, Department of Animal Sciences. 2011. Corn silage field day corn hybrid variety test. UF/IFAS Plant Sci. Rsrch. Edu. Unit. Citra, FL. http://animal.ifas.ufl.edu/extension/CSFD/2011csfd/2011present.shtml [ US EPA ] United States Environmental Protection Agency. 2005. Protecting Water Quality from Agricultural Runoff. EPA841 F 0 5 001. http://www.epa.gov/owow/NPS/Ag_Runoff_Fact_Sheet.pdf van Es, H.M., J.M. Sogbedji and R.R. Schindelbeck. 2006. Effect of manure application timing, crop, and soil type on nitrate leaching. J. Environ. Qual. 35:670 679. VanHorn, H., G. Newton and W. Kunkle. 1996. Ruminant nutrition from an environmental perspective: Factors affecting whole farm nutrient bala nce. J. Anim. Sci. 74:3082 3102. Van H orn H., A. W ilkie W. P owers and R. N ordstedt 1994. Components of dairy manure management systems. J. Dairy Sci. 77:2008 2030.
137 Wang, S., D. Fox, D. Cherney, L. Chase and L. Tedeschi. 2000. Wh ole herd optimization w ith the C ornell net carbohydrate and protein system. III. application of an optimization model to evaluate alternatives to reduce nitrogen and phosphorus mass balance. J. Dairy Sci. 83:2160 2169. Wienhold, B.J. 2007. Comparison of laboratory methods and an in situ method for estimating nitrogen mineralization in an irrigated silt loam soil. Commun. Soil Sci. Plant Anal. 38:1721 1732. Woodard, K.R., E.C. French, L.A. Sweat, D.A.Graetz, L.E. Sollenberger, B. Macoon, K.M. Portier. S.J. Rymph, B.L. Wade, G.M. Prine, and H.H. Van Horn. 2003. Nitrogen removal and nitrate leaching for two perennial, sod based forage systems receiving dairy effluent. J. Environ. Qual. 32:996 1007. Wright D., J. Marois, J. Rich, and R. Sprenkel. 2011 a Field Corn Production Guide. EDIS Document SS AGR 85. Wright, D., E.B. Whitty, and C.G. Chambliss. 2011 b Fertilization of Agronomic Crops. EDIS Document SS AGR 152. Zhang, J., R.T. James, G. Ritter, and B. Sharfstein. 2007. Lake Okeechobee Protection Program State of the Lake and Watershed. In: South Florida Environmental Report. South Florida Water Management District, West Palm Beach, FL. Zotarelli, L., L. Avila, J.M.S. Scholberg, and B.J.R. Alves. 2009. Benefits of vetch and rye cover crops to sweet corn u nder no tillage. Agron. J. 101: 252 260.
138 BIOGRAPHICAL SKETCH Rebecca Jean Hellmuth was born in 1986 in Baltimore, MD to Martha P. Hellmuth and Dr. John H. Hellmuth. Rebecca and h er family moved to Middleburg, Flor ida in 1988. Rebecca graduated cum l aude from the University of Florida with a bachelor of science in b usiness a dministration with a concentration in f inance in August 2008. In 2010, Rebecca returned to the University of Florida to pursue a master of s cie nce in the Soil and Wa ter Science Department with a minor in agronomy Rebecca graduated with