Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The fate of fertilizer nitrogen in raspberry production as affected by nitrogen inputs and irrigation… Armstrong, Nathalie Monica 2015

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

Item Metadata

Download

Media
24-ubc_2015_september_Armstrong_Nathalie.pdf [ 5.54MB ]
Metadata
JSON: 24-1.0166453.json
JSON-LD: 24-1.0166453-ld.json
RDF/XML (Pretty): 24-1.0166453-rdf.xml
RDF/JSON: 24-1.0166453-rdf.json
Turtle: 24-1.0166453-turtle.txt
N-Triples: 24-1.0166453-rdf-ntriples.txt
Original Record: 24-1.0166453-source.json
Full Text
24-1.0166453-fulltext.txt
Citation
24-1.0166453.ris

Full Text

THE FATE OF FERTILIZER NITROGEN IN RASPBERRY PRODUCTION AS AFFECTED BY NITROGEN INPUTS AND IRRIGATION REGIME: AN EXPERIMENT IN THE FRASER VALLEY, BRITISH COLUMBIA  by  Nathalie Monica Armstrong   B.Sc., University of the Fraser Valley, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  THE COLLEGE OF GRADUATE STUDIES  (Environmental Sciences)   THE UNIVERSITY OF BRITISH COLUMBIA  (Okanagan)  July 2015   © Nathalie Monica Armstrong, 2015 ii Abstract  The lower Fraser Valley is an important agricultural region in British Columbia, Canada whose central portion is underlain by the Abbotsford-Sumas Aquifer. The aquifer has been identified as vulnerable and high concentrations of nitrate that have been linked to application of nitrogen (N) fertilizers to crops, particularly red raspberry. The region has the most concentrated raspberry production in Canada and therefore raspberry management has the potential to influence N losses to the aquifer. The goal of this study was to determine how variations in irrigation and N fertilization affect the fate of N from fertilizer (NFF) from one year’s application in order to identify practices that maximize crop acquisition of N and minimize the risks of nitrate leaching.  Specific practices that were compared include rate of N inputs (50 vs. 100 kg N ha-1), mode of N application (fertigation vs. broadcast), and irrigation method (fixed schedule vs. plant need based).  To determine the influences of different management practices, the fate of NFF was traced by applying 15N labeled fertilizer in April and May 2011, as a split application, and then analyzing movement of the 15N into plant tissues, leachate, and soil at four critical times during the subsequent year.  This study found that irrigation regime had a strong influence on N losses by leaching.  Reduced water inputs reduced nitrate -N leaching and did not negatively affect productivity. Fertigation, which employed a combination of reduced N and water inputs, improved plant acquisition of NFF, resulting in reduced leaching of nitrate-N.  The 50 kg N ha-1 treatments produced berry yields comparable to the 100 kg N ha-1 treatments while the higher N inputs resulted in greater NFF being removed from the system by berry harvest and pruning. Peak plant uptake of NFF was found to be in July 2011 with the maximum held in the floricane component. By August, NFF content in both the floricane and primocane components had decreased and the primocane component had relocated NFF from the leaves to the cane.  After the drop between July and August, NFF in the system remained relatively stable into the following growing season.    iii Preface  The data presented in this work arose from observations and analyses conducted primarily by the author with some exceptions as noted here.  As noted in Chapter 1, the broader field experiment was designed and installed by Shawn Kuchta, Dr. Denise Neilsen, Dr. Tom Forge and Dr. Bernie Zebarth of Agriculture and Agri-Food Canada (AAFC), as presented in Kuchta (2012).  During the duration of this experimental program, Shawn Kuchta and other AAFC technicians helped maintain this experiment, providing assistance with pruning, sampling, berry harvest, water sampling and sample processing.    The majority of sample collection and sample preparation was undertaken by the author. The analysis of total nitrogen, nitrate and ammonium in water samples was undertaken by Shawn Kuchta of AAFC.  The analysis of samples for 15N was undertaken by the University of California Stable Isotope Facility in Davis, California (UC SIF), as indicated in Chapter 2.  All statistical analyses were undertaken by the author. iv Table of Contents  Abstract .................................................................................................................................... ii Preface ..................................................................................................................................... iii List of Tables .......................................................................................................................... ix List of Figures ....................................................................................................................... xiii Acknowledgements ............................................................................................................ xviii Chapter 1 Introduction .......................................................................................................... 1 1.1 Raspberry production in the Fraser Valley ............................................................ 2 1.2 Raspberry crop management ................................................................................... 3 1.2.1 Raspberry cropping cycle ...................................................................................... 3 1.2.2 Current management ............................................................................................. 4 1.2.2.1 Current nutrient management .......................................................................... 4 1.2.2.2 Current irrigation Management ....................................................................... 4 1.2.2.3 Fertigation as a management alternative ......................................................... 4 1.3 Management practices and their influence ............................................................. 5 1.3.1 N cycling and the fate of NFF ............................................................................... 5 1.3.2 Implications of crop management ......................................................................... 5 1.3.2.1 Implications of nutrient management .............................................................. 5 1.3.2.2 Implications of irrigation management ........................................................... 6 1.3.3 Methods to determine fate of NFF ........................................................................ 7 1.3.4 Research problem .................................................................................................. 7 1.4 Research objectives .................................................................................................... 8 1.5 Thesis Overview ......................................................................................................... 9 Chapter 2 Site History and Methods ................................................................................... 10 2.1 Study site ................................................................................................................... 10 2.1.1 Site location and description ............................................................................... 10 2.1.2 Site soil properties ............................................................................................... 10 2.2 Experimental design ................................................................................................ 12  v 2.2.1 Experiment establishment ................................................................................... 12 2.2.2 Instrumentation .................................................................................................... 14 2.2.3 Irrigation management ........................................................................................ 16 2.2.4 Nutrient management .......................................................................................... 16 2.2.5 Experimental treatments ...................................................................................... 17 2.3 Experimental methods ............................................................................................. 18 2.3.1 2011 nutrient and irrigation management ........................................................... 18 2.3.2 2012 nutrient and irrigation management ........................................................... 19 2.4 Sampling methods .................................................................................................... 19 2.4.1 Leachate sample collection ................................................................................. 20 2.4.2 Plant tissue collection, excluding roots ............................................................... 22 2.4.3 Soil sampling ....................................................................................................... 23 2.5 Analysis ..................................................................................................................... 25 2.5.1 Chemical analysis ................................................................................................ 25 2.5.2 Calculations ......................................................................................................... 26 2.5.3 Statistical analysis and considerations ................................................................ 28 Chapter 3 Results .................................................................................................................. 31 3.1 Nitrogen budget of major system components ......................................................... 31 3.1.1 Nitrogen budget in growing season: July 4th, 2011 ............................................. 31 3.1.1.1 Total N July 4th, 2011 .................................................................................... 31 3.1.1.2 Nitrogen from fertilizer July 4th, 2011 ........................................................... 32 3.1.1.3 Percent NFF of applied July 4th, 2011 ........................................................... 33 3.1.1.4 Percent NFF of total nitrogen July 4th, 2011 ................................................. 34 3.1.2 Nitrogen budget post-harvest: August 22nd, 2011 ............................................... 35 3.1.2.1 Total N August 22nd, 2011 ............................................................................. 35 3.1.2.2 Nitrogen from fertilizer August 22nd, 2014 ................................................... 35 3.1.2.3 Percent NFF of applied August 22nd, 2014 ................................................... 36 3.1.2.4 Percent NFF of total nitrogen August 22nd, 2011 .......................................... 39 3.1.3 Nitrogen budget in mid-winter: January 24th, 2012 ............................................ 39 3.1.3.1 Total nitrogen January 24, 2012 .................................................................... 39 3.1.3.2 Nitrogen from fertilizer January 24, 2012 ..................................................... 40  vi 3.1.3.3 Percent NFF of applied January 24, 2012 ..................................................... 41 3.1.3.4 Percent from NFF of total nitrogen January 24, 2012 ................................... 42 3.1.4 Nitrogen budget in the following growing season: June 25th, 2012 .................... 43 3.1.4.1 Total nitrogen June 25, 2012 ......................................................................... 43 3.1.4.2 Nitrogen from fertilizer June 25, 2012 .......................................................... 44 3.1.4.3 Percent NFF of applied June 25, 2012 .......................................................... 45 3.1.4.4 Percent NFF of total nitrogen June 25, 2012 ................................................. 45 3.2 Temporal dynamics ................................................................................................. 46 3.2.1 Entire system nitrogen over time ........................................................................ 46 3.2.1.1 General leachate dynamics ............................................................................ 51 3.2.1.2 System changes between preharvest and postharvest ................................... 54 3.2.1.3 System changes into the dormant season ...................................................... 54 3.2.1.4 System changes into the next growing season .............................................. 55 3.3 Internal Cycling of Plant Nitrogen ......................................................................... 57 3.3.1 Timing of uptake ................................................................................................. 57 3.3.2 Partitioning of N .................................................................................................. 64 3.4 Environmental Conditions ...................................................................................... 67 Chapter 4 Discussion ............................................................................................................ 69 4.1 N application rate .................................................................................................... 69 4.2 Mode of application ................................................................................................. 70 4.3 Irrigation regime ...................................................................................................... 72 4.4 Timing of uptake and internal redistribution of NFF in the plant ..................... 73 4.5 Unaccounted for fates of NFF ................................................................................. 75 4.5.1 Losses from Soil .................................................................................................. 76 4.5.2 Losses from the plant .......................................................................................... 77 Chapter 5 Conclusion ........................................................................................................... 79 Bibliography .......................................................................................................................... 81 Appendices ............................................................................................................................. 88 Appendix A: Equations .................................................................................................... 88 Appendix B: Treatment Comparisons ............................................................................ 93  vii B.1 Entire System: Total Nitrogen .................................................................................. 93 B.2 Entire System: Fertilizer Nitrogen ........................................................................... 95 B.3 Entire System: Percent Fertilizer Nitrogen of Applied ............................................ 97 B.4 Entire System: Percent Fertilizer Nitrogen of Total Nitrogen ................................. 99 B.5 Soil: Total Nitrogen ................................................................................................ 101 B.6 Soil: Fertilizer Nitrogen ......................................................................................... 103 B.7 Soil: Percent Fertilizer Nitrogen of Applied .......................................................... 105 B.8 Soil: Percent Fertilizer Nitrogen of Total Nitrogen ............................................... 107 B.9 Plant (Corrected for Removal): Total Nitrogen ..................................................... 109 B.10 Plant (Corrected for Removal): Fertilizer Nitrogen ............................................. 111 B.11 Plant (Corrected for Removal): Percent Fertilizer Nitrogen of Applied .............. 113 B.12 Plant (Corrected for Removal): Percent Fertilizer Nitrogen of Total Nitrogen ... 115 B.13 Removed: Total Nitrogen ..................................................................................... 117 B.14 Removed: Fertilizer Nitrogen ............................................................................... 119 B.15 Removed: Percent Fertilizer Nitrogen of Applied ............................................... 121 B.16 Removed: Percent Fertilizer Nitrogen of Total Nitrogen ..................................... 123 B.17 Leachate: Cumulative Total Nitrogen .................................................................. 125 B.18 Leachate: Cumulative Fertilizer Nitrogen ............................................................ 127 B.19 Leachate: Cumulative Percent Fertilizer Nitrogen of Applied ............................. 130 B.20 Leachate: Cumulative Percent Fertilizer Nitrogen of Total Nitrogen .................. 132 B.21 Leachate: Total Nitrogen ...................................................................................... 134 B.22 Leachate: Fertilizer Nitrogen ............................................................................... 146 B.23 Leachate: Fertilizer Nitrogen of Applied ............................................................. 158 B.24 Leachate: Fertilizer Nitrogen of Total Nitrogen .................................................. 169 B.25 Leachate: Cumulative Total Nitrogen .................................................................. 181 B.26 Leachate: Cumulative Fertilizer Nitrogen ............................................................ 193 B.27 Plant Component: Total Nitrogen ........................................................................ 204 B.28 Plant Component: Fertilizer Nitrogen .................................................................. 207 B.29 Plant Component: Percent Fertilizer Nitrogen of Applied ................................... 209 B.30 Plant Component: Percent Fertilizer Nitrogen of Total Nitrogen ........................ 211 B.31 Floricane: Total Nitrogen ..................................................................................... 213  viii B.32 Floricane: Fertilizer Nitrogen ............................................................................... 215 B.33 Floricane: Percent Fertilizer Nitrogen of Applied ................................................ 217 B.34 Floricane: Percent Fertilizer of Total Nitrogen .................................................... 219 B.35 Floricane Leaf: Total Nitrogen ............................................................................. 221 B.36 Floricane Leaf: Fertilizer Nitrogen ...................................................................... 222 B.37 Floricane Leaf: Percent Fertilizer Nitrogen of Applied ....................................... 224 B.38 Floricane Leaf: Percent Fertilizer of Total Nitrogen ............................................ 225 B.39 Laterals: Total Nitrogen ....................................................................................... 227 B.40 Laterals: Fertilizer Nitrogen ................................................................................. 228 B.41 Laterals: Percent Fertilizer Nitrogen of Applied .................................................. 230 B.42 Laterals: Percent Fertilizer of Total Nitrogen ...................................................... 232 B.43 Fruit: Biomass ...................................................................................................... 233 B.44 Fruit: Total Nitrogen ............................................................................................ 235 B.45 Fruit: Fertilizer Nitrogen ...................................................................................... 236 B.46 Fruit: Percent Fertilizer Nitrogen of Applied ....................................................... 238 B.47 Fruit: Percent Fertilizer of Total Nitrogen ........................................................... 239 B.48 Primocane: Total Nitrogen ................................................................................... 241 B.49 Primocane: Fertilizer Nitrogen ............................................................................. 243 B.50 Primocane: Percent Fertilizer Nitrogen of Applied .............................................. 245 B.51 Primocane: Percent Fertilizer of Total Nitrogen .................................................. 247 B.52 Primocane Leaf: Total Nitrogen ........................................................................... 249 B.53 Primocane Leaf: Fertilizer Nitrogen .................................................................... 251 B.54 Primocane Leaf: Percent Fertilizer Nitrogen of Applied ..................................... 252 B.55 Primocane Leaf: Percent Fertilizer of Total Nitrogen .......................................... 254 B.56 Root: Total Nitrogen ............................................................................................ 255 B.57 Root: Fertilizer Nitrogen ...................................................................................... 257 B.58 Root: Percent Fertilizer Nitrogen of Applied ....................................................... 259 B.59 Root: Percent Fertilizer of Total Nitrogen ........................................................... 261     ix List of Tables   Table 2.1    Soil properties and soil characterization in the research field (Kuchta 2012) ..... 11 Table 2.2    Nitrogen, irrigation and alley management treatments used for the 15N study .... 18 Table 2.3    Destructive and non-destructive sampling schedule intended to trace fertilizer-N in P32. ................................................................................................................ 20 Table 3.1    Effects of N input by irrigation type treatment combinations on total N in soil, plant components, leachate and the entire system on July 4, 2011. .................. 31 Table 3.2    Effects of N input by irrigation type treatment combinations on nitrogen from fertilizer (NFF) in soil, plant components, leachate and the entire system on July 4, 2011. Letters (a, b) denote significant pairwise comparisons. ...................... 32 Table 3.3    Effects of N input by irrigation type treatment combinations on percent fertilizer of applied (%FOA) in soil, plant components, leachate and the entire system on July 4, 2011. Letters (a, b) denote significant pairwise comparisons. ............... 33 Table 3.4    Effects of N input by irrigation type treatment combinations on percent fertilizer N of total N (%FOT) in soil, plant components, leachate and the entire system on July 4, 2011. Letters (a, b) denote significant pairwise comparisons. .......... 34 Table 3.5    Effects of N input by irrigation type treatment combinations on total N in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons. .......................... 35 Table 3.6    Effects of N input by irrigation type treatment combinations on NFF in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons. .................................... 36 Table 3.7    Effects of N input by irrigation type treatment combinations on %FOA in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons. .......................... 37 Table 3.8    Unaccounted for losses between July 4th and Aug 22nd, 2011 of NFF as a proportion of that applied (%FOA).  These values were not statistically analyzed. ............................................................................................................ 39  x Table 3.9    Effects of N input by irrigation type treatment combinations on %FOT in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons. .......................... 40 Table 3.10    Effects of N input by irrigation type treatment combinations on total N in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons. .......................... 40 Table 3.11    Effects of N input by irrigation type treatment combinations on NFF in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons. .......................... 41 Table 3.12    Effects of N input by irrigation type treatment combinations on %FOA in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons. .......................... 42 Table 3.13    Effects of N input by irrigation type treatment combinations on %FOT in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons. .......................... 43 Table 3.14    Effects of N input by irrigation type treatment combinations on Total N in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. .................................... 44 Table 3.15    Effects of N input by irrigation type treatment combinations on NFF in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. .................................... 45 Table 3.16    Effects of N input by irrigation type treatment combinations on %FOA in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. .................................... 46 Table 3.17    Effects of N input by irrigation type treatment combinations on %FOT in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. .................................... 47 Table 3.18    An estimate of cumulative NO3- -NFF (kg N ha-1) in leachate between January and June 2012.  Estimates were derived using the product of cumulative values of NO3- -TN in leachate between January and June 2012 and %FOT in leachate  xi from April 2012.  April was used as it was the only date leachate was analyzed during that time period. ..................................................................................... 52 Table 3.19    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on July 4, 2011. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. ..................................................................... 59 Table 3.20    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on August 22, 2011. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. ..................................................................... 60 Table 3.21    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on January 24, 2012. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest  xii were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. ..................................................................... 61 Table 3.22    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on June 25, 2012. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. ..................................................................... 62 Table 3.23    Yield (kg ha-1) of Saanich raspberry as influenced by N and irrigation management in 2010 and 2011.  No treatment differences were found. ........... 64 Table 3.24    Monthly precipitation and mean monthly air temperatures at the Abbotsford Airport (~1 km from the experimental site) 2011 – 2013 (Environment Canada, 2011, 2012) ........................................................................................................ 68     xiii List of Figures  Figure 1.1    The geographical location of the Abbotsford Sumas Aquifer and the site location (Kuchta, 2012; originally modified from Graham et al. 2006) ........................... 2 Figure 2.1    Soil profile with horizons identified, Clearbrook substation (Kuchta, 2012). ... 11 Figure 2.2  Detailed planar view of the experimental layout .................................................. 13 Figure 2.3  Detailed planar view of individual full plot and sub-plot used for 15N study. ..... 14 Figure 2.4  PCAPS sampler installation sites under raspberry row management and associated alley management (Kuchta, 2012). .................................................. 15 Figure 2.5  Soil sampling locations on the raspberry row, within the 15N application area. .. 24 Figure 3.1  The percent fertilizer of applied (%FOA) lost from the entire system by an unknown mechanism at each sampling date. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fertilizer was applied as a split application on April 17 and May 24, 2012.  These data were not statistically analyzed. ......................................................................................... 38 Figure 3.2  The proportion of nitrogen from fertilizer (NFF) to applied N expressed as a percent (%FOA), grouped by treatment. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for treatment, date and treatment x date interaction are 0.2032, <0.0001 and 0.6542 respectively. 48 Figure 3.3  The effects of nitrogen input by irrigation type treatment combinations on NFF in the entire system at each sampling date. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fertilizer was applied as a split application on April 17 and May 24, 2012.  P-values for effects of treatment, date and treatment by date are <0.0001, <0.0001 and 0.0308 respectively.  Error bars attached to S-100N points are (+/- 6.33 kg N ha-1) standard error calculated from the pooled variance. .................................................................................. 49 Figure 3.4  The NFF balance on July 4, 2011. Negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components. At this date no plant material had yet been removed from the system. ......................................................................................................... 49  xiv Figure 3.5  The NFF balance on August 22, 2011. Negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components.  At this date harvest fruit had been removed from the system. The removed portion is cumulative and includes N recovered in fruit harvested between July 25 and Aug 19, 2011.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop. ...................... 50 Figure 3.6  The NFF balance on January 24, 2012. Negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components. The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012. Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop. Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in. The decomposing laterals and leaves were not separated from the cane. ........................................................................................................................... 50 Figure 3.7  Nitrogen from fertilizer balance on June 25, 2012. The negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components.  The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. .............................................................................. 51 Figure 3.8  Effects of N input by irrigation type treatment combinations on total cumulative NO3-N in leachate. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 5.48 kg N ha-1) standard error calculated from the pooled variance.  xv Cumulative values are incomplete after Dec 27, 2011 because after that date only a select few sample dates were analyzed. .................................................. 53 Figure 3.9  Effects of N input by irrigation type treatment combinations on NFF cumulative NO3-N in leachate. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, 0.0016 and 1.000 respectively.  Error bars represent (+/- 1.90 kg N ha-1) standard error calculated from the pooled variance. Cumulative values are incomplete after Dec 27, 2011 because after that date only a select few sample dates were analyzed. ....................................................................... 53 Figure 3.10  Effects of N input by irrigation type treatment combinations on total NO3-N in leachate at each sampling event. Between April and December 2011 all sampling dates are depicted.  After Dec 27, 2012 only three more events in January, April, October and November were statistically analyzed. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 0.81 kg N ha-1) standard error calculated from the pooled variance. ........................................................ 56 Figure 3.11  Effects of N input by irrigation type treatment combinations on NO3-NFF in leachate at each sampling event. Between April and December 2011 all sampling dates are depicted.  After Dec 27, 2012 only three more events in January, April, October and November were statistically analyzed. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are 0.2729, <0.0001 and 0.5538 respectively.  Error bars represent (+/- 0.33 kg N ha-1) standard error calculated from the pooled variance. ................................................................. 57 Figure 3.12  Effects of N input by irrigation type treatment combinations on NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) at each sampling event. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are 0.0002, <0.0001 and 0.0780 respectively.  Error bars represent (+/- 1.78 kg N ha-1) standard error calculated  xvi from the pooled variance. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. .................................................................................... 58 Figure 3.13  Effects of N input by irrigation type treatment combinations on NFF concentration in floricane leaves at each sampling event.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 0.5 !g N mg-1 sample) standard error calculated from the pooled variance. The decomposing leaves were left to drop to the ground. ............................................................................................. 63 Figure 3.14  Effects of N input by irrigation type treatment combinations on NFF concentration in primocane leaves at each sampling event. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 0.72 !g N mg-1 sample) standard error calculated from the pooled variance. The decomposing leaves were left to drop to the ground. ......................................................................... 63 Figure 3.15  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on July 4th, 2011 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. .................................................................................................................. 65 Figure 3.16  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on August 22, 2011 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes  xvii were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. ......... 66 Figure 3.17  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on January 24, 2012 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. ......... 66 Figure 3.18  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on June 25, 2012 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. ......... 67   xviii Acknowledgements  I want to sincerely thank my supervisors, Dr. Craig Nichol and Dr. Denise Neilsen as well as the other members of my supervisory committee, Dr. Tom Forge and Dr. David Scott, for their support, guidance and encouragement throughout my graduate studies. Thank you to Craig for your all your generous time, thoughtful input and humorous methods of motivation. Thank you to Denise for your patience and availability in answering my many questions over the last four years. Thank you to Tom for your open door policy and immeasurable contributions to this project. Thank you to Ben Frey, Naomi Hashimoto, Shaobing Yu and Shawn Kuchta who spent countless hours helping me with field and lab work.  I am very grateful to Agriculture and Agri-Food Canada and the SAGES program as well as UBC Okanagan for the support in this endeavour. I would also like to thank my parents for their unquestioning support over these last four years and for always believing in me. Finally, thank you to my partner Mike for standing by me throughout this experience, for understanding when I was preoccupied with schoolwork and for being a good listener.    1 Chapter 1 Introduction The production of anthropogenically-fixed nitrogen as fertilizers, and their widespread use in agricultural production, has significantly altered the nitrogen (N) cycle in both terrestrial and aquatic systems (Galloway et al. 2004; Burt et al. 2008).  One of the prominent concerns that has been identified is the increase of nitrate (NO3-) loading to groundwater, resulting in the contamination of aquifers with elevated concentrations of NO3- (Spalding and Exner 1993; Wassenaar 1995).  As the primary user of fertilizers, agriculture can be the most important contributor to the leaching of NO3- from the soil zone.  The problem continues to be a growing concern although many regions have developed best management practices in order to minimize N losses.  A global study that quantified the long-term fate of nitrogen from fertilizer (NFF) in agricultural soils found that over a 30 year period only 61 to 65 % of N applied to crops was taken up by plants and 8 to 12 % was lost to the hydrosphere (Sebilo et al. 2013). As Earth’s population continues to grow so does the demand for increased agricultural productivity and clean water.  The global challenge of meeting increased food requirements has necessitated the need for an improved understanding of how to minimize and mitigate land use impacts on groundwater.  There are many regional studies worldwide that seek to achieve this necessity by meeting crop N demands while optimizing yield and profit and minimizing environmental damage to water resources (Tilman et al. 2002).   The lower Fraser Valley is an important agricultural region in British Columbia (BC), Canada whose central portion is underlain by the Abbotsford-Sumas Aquifer, which supplies drinking water to more than 100,000 people (Figure 1.1).  The aquifer has been identified as vulnerable and in the past, agricultural irrigation and nutrient management have been linked to the nitrate contamination of this aquifer (Chesnaux et al. 2007; Wassenaar 1995; Wassenaar et al. 2006; Zebarth et al. 1998).  Although the region has some of the most fertile soils in the world, crops grown there require nutrient inputs to produce yields that are economically viable for growers.  Approximately one third of the farms in the lower Fraser Valley are categorized as high or very high risk for residual nitrate-nitrogen (NO3--N) in soil (Kowalenko et al. 2007).  Much research on crop management and environmental mitigation has been conducted in the region due to the intensive agricultural production and its risk to groundwater quality.   2  Figure 1.1    The geographical location of the Abbotsford Sumas Aquifer and the site location (Kuchta, 2012; originally modified from Graham et al. 2006)  1.1 Raspberry production in the Fraser Valley The Fraser Valley has the most concentrated red raspberry (Rubus idaeus) production in Canada, covering an estimated area of 1600 ha (Dean et al. 2000; Statistics Canada 2011).  The majority of this raspberry production is located over top of the Abbotsford-Sumas aquifer and thus its management has the potential to influence the quality of groundwater (Statistics Canada 2011).  The region also hosts an intensive poultry industry, and historically a high volume of poultry manure was applied to raspberry fields. Past intensive use of poultry manure or fertilizer in Fraser Valley raspberry production likely increased the amount of soluble nitrogen ions available to the plants as well as losses by leaching, surface runoff, denitrification and ammonia volatilization (Zebarth et al. 1998).  The use of poultry manure in Fraser Valley raspberry production was originally identified as the primary source of NO3- in the aquifer (Wassenaar 1995; Zebarth et al. 1998). Recent implementation of Good Agricultural Practices (e.g. CanadaGAP, EuroGAP), which are voluntary guidelines to ensure market access, have made it difficult for growers to use raw manure on established crops.  This has encouraged greater reliance on mineral fertilizers and resulted in a shift from  3 the use of manure to mineral fertilizers which have since been linked to the continued loading and transport of nitrate through the vadose zone of the Abbotsford-Sumas aquifer (Chesnaux et al. 2007; Wassenaar et al. 2006).   1.2 Raspberry crop management 1.2.1 Raspberry cropping cycle Raspberry is a perennial, berry-producing crop with a biennial growth habit.  In summer-bearing raspberries, the first year biennial shoots are known as primocanes or vegetative growth.  Primocanes are longer and thicker than floricanes and do not form long lateral branches.  In SW British Columbia, primocanes begin to emerge in May. In the second year, primocanes become the reproductive floricanes which are woody, have lateral branches and bear fruit.  Floricanes begin to bud in March and begin growing fruit-bearing laterals in April.  The floricanes begin to flower in May and berries are usually ripe by mid-July.  Once berries are harvested, the floricanes are usually pruned out because they die and decay and are no longer useful to the plant.  A raspberry crop has both primocane and floricane growth each year, with the exception of the first year after planting. Each year, some primocanes are killed with herbicides or hoed out in the early season because if too many are allowed to grow they can compete with the fruit bearing floricanes for nutrients and therefore their growth density is managed to maximize production. Raspberry crops in the Fraser Valley are managed by grouping floricanes into bunches and training them to a post or as a solid hedge row trained to a high tensile wire system.  The current Berry Production Guide provided by the British Columbia Ministry of Agriculture provides management guidelines regarding nutrient, water, pests and soil to ensure profitable yields of quality fruit (British Columbia Ministry of Agriculture 2012).  Despite extensively researched nutrient and irrigation management guidelines, there is a wide range of management practices for raspberry production in the Fraser Valley.         4 1.2.2 Current management 1.2.2.1 Current nutrient management Generally, raspberry growers apply N fertilizers at a rate of 50-100 kg N ha-1 as a split application.  Fertilizers are applied in April just before the new primocanes emerge and then again in May, approximately six weeks later.  The split application is intended to ensure that the NFF is taken up by both the earlier budding floricanes and the primocanes, which emerge later. Splitting the application is also intended to ensure that more NFF will be retained in the soil and to prevent the leaching of N before the plant is able to use it.  Although raspberries use NO3- more readily than NH4+ most growers apply urea (CO(NH2)2) because it is cheaper than nitrate fertilizers. Most raspberry growers in the lower Fraser Valley use the BC Ministry of Agriculture nutrient management guidelines (British Columbia Ministry of Agriculture 2012; Kuchta 2012).   1.2.2.2  Current irrigation Management Generally, raspberry growers in the Fraser Valley use high efficiency drip micro-irrigation. Typically, this consists of drip tape shallowly buried in the row when hills are formed prior to planting or suspended from post to post on wire over the center of a raspberry row.  Raspberry crops in the Fraser Valley are generally watered between June and September and durations and frequencies vary by grower (Chesnaux and Allen, 2008).  Growers rely on in-field observation, hand-texturing and past experience to guide the duration and frequency of irrigation (Kuchta, 2012).  One consistency in irrigation is that it is that it is usually a fixed regime with durations ranging between 4 to 8 hours daily or every second day (Kuchta, 2012).  Irrigation is generally increased during peak demand (July-August) and is tapered off at the end of the growing season.  Some growers may alter their irrigation regime in response to large precipitation events by turning irrigation off.  The prescribed annual water requirements using drip irrigation on coarse soils, such as those in Abbotsford, are approximately 300 mm of water (Van der Gulik, 1999). 1.2.2.3  Fertigation as a management alternative Fertigation is another method some growers use to apply N fertilizers, although it is not commonplace in the Fraser Valley.  Fertigation is the application of fertilizer through an  5 irrigation system.  Research has shown the method to optimize yield and raspberry quality in other parts of the world (Gurovich, 2008).  In BC, little research has been conducted on the viability of fertigation in raspberry production, however much has been done on apple production (Neilsen et al. 2009; Neilsen and Neilsen 2002). The promise of fertigation is that by better matching N supply to plant requirements, it may be more efficient, i.e. a greater proportion of applied N may be taken up by the crop, and therefore may allow for lower rates of fertilizer applications and, possibly reduced risk of nitrate leaching (Neilsen et al. 2009).  1.3 Management practices and their influence 1.3.1 N cycling and the fate of NFF To better understand the implications of nutrient and irrigation management practices it is first important to understand N cycling.  Nitrogen is generally found as an insoluble organic form in soil. Through mineralization and nitrification, ammonium and nitrate are released from the organic N pool. Alternative fates for ammonium applied as fertilizer are immobilization by soil microbes (depending on the C/N ratio of organic matter available for heterotrophic microbial activity), removal by plant uptake and volatilization.  The fate of nitrate fertilizers includes immobilization by soil microbes, removal by plant uptake, losses by leaching, and denitrification.  Nitrogen fertilizers are used extensively in agriculture because deficiencies of the element in soil are known to cause problems in plant growth, development and physiology (Drew and Siworo 1977; Novoa and Loomis 1981). Nutrient and irrigation management in crops influence these possible fates of N.  1.3.2 Implications of crop management 1.3.2.1 Implications of nutrient management Fertilizer applications can affect natural soil processes such as mineralization and nitrification as well as crop yield and quality (Neilsen and Neilsen 2002). The application of N fertilizers typically leads to greater nitrogen losses from the soil (Schepers and Raun 2008).  Higher rates of N application can lead to greater losses if the applied N is not well timed to plant demand or is in excess of plant requirements.  The main mechanisms through which nitrogen is lost are leaching and volatilization. N losses to leaching are prevalent in  6 many forms of agriculture because NFF quickly mineralizes and nitrifies after application, increasing soil nitrate levels. Mineral fertilizers introduce ammonium (NH4+) and NO3- ions into the soil that are taken up by plants and participate in the N cycle in the same way as naturally derived ammonium and nitrate.  Kowalenko (1989) found that mineralization of soil organic N occurs rapidly after fertilizer application in Fraser Valley soils.    Applications of ammonium or urea increase the amount of NH4+ available for nitrification, which can lead to higher levels of nitrate in amended soils and an increased risk of nitrate leaching to groundwater (Dean et al. 2000; Zebarth et al. 1998, 2002). Nitrate is easily lost from soil by leaching because it is negatively charged and does not adsorb to the soil particle surfaces.  Prolonged mineralization and nitrification during the growing season and into autumn has been linked to the buildup of nitrate in Fraser Valley soils (Kowalenko, 1989; Dean et al. 2000; Wassenaar 1995).  The combination of fall rains and excess soil nitrate may lead to leaching in Fraser Valley soils. Research has shown that precipitation moving through the soil profile during the rainy season (autumn, winter and early spring) is linked to N losses to leaching (Chesnaux and Allen, 2008; Kuchta, 2012).  In contrast, the summer months experience a water deficit (Zebarth et al. 1998).  This deficit makes irrigation a necessity for profitable raspberry production in the Fraser Valley (Dale 1989).    1.3.2.2 Implications of irrigation management When a large volume of water is moving down through the unsaturated soil as a result of irrigation and/or precipitation, nitrate in the water will move down through the root zone and into ground water.  In 2009 and 2010, Kuchta (2012) found that drainage and nitrate losses from the root zone have a positive relationship.  Therefore, higher rates of irrigation have the potential to increase the amount of nitrate entering groundwater.  Conversely, alternative methods such as irrigation based on plant need, which apply much less water, have the potential to reduce nitrate entering the groundwater.  Kuchta (2012) found that growing season losses were sensitive to fixed irrigation management as commonly practiced in Fraser Valley raspberry production and found nitrate losses during the growing season were 50% lower in treatments irrigated based on plant need than under fixed rate irrigation.        7  1.3.2.3 Other complicating factors Some other complicating factors are differences among cultivars, inherent site soil fertility, N cycling within the plant and the age of the plantings (Rempel et al. 2004).  For example, site fertility in the Fraser Valley is highly variable due to historic regular manure amendments.  Fields with a history of manure use have been identified as at risk of leaching nitrate (Dean et al. 2000; Jeffries et al. 2008; Zebarth et al. 1997).    1.3.3 Methods to determine fate of NFF Variations of raspberry crop management can affect the efficiency of one year’s N fertilizer application. Despite much research on the N cycling in the Fraser Valley and more limited research on the fate of NFF, efficient N application rates for raspberry crops have been difficult to identify (Rempel et al. 2004; Strik 2008).  This is due to the inability to obtain an accurate estimate of N contributions to the system from other sources such as soil N mineralization, N applied in irrigation water and atmospheric N deposition (Zebarth et al. 1997).  Research on the fate of NFF is necessary to determine the efficiency of a single year’s N application.  Much research on the fate of fertilizer N has been conducted on berry crops to determine the optimum N application rates.  A common method used to research the fate of NFF is tracer studies using 15N, as 15N is generally in low natural abundance in most soils and its known proportion in a region allows for detection of enrichment. 15N studies have been used successfully in the past to assess the fate of NFF in a variety of crops under a variety of management strategies (Neilsen et al. 2001; Rempel et al. 2004; Strik et al. 2006; Yasmin et al. 2006).    1.3.4 Research problem Efficient crop management regimes should match N and water application rates to meet plant demand while minimizing potential nitrogen losses to leaching and volatilization.  The timing of N uptake is very relevant from a management perspective because, for example, if there is a delay between N application and plant uptake, N is available to loss mechanisms and may be lost from the system. It is essential to match NFF inputs and timing in raspberry production to optimize yield and crop health while decreasing environmental risk (Kuchta  8 2012; Rempel et al. 2004; Zebarth et al. 1997). N inputs that are not timed properly or in excess may not be taken up by the crop is at risk of being leached (Hughes-Games and Zebarth 1999; Mitchell et al., 2003; Portela et al., 2006; Zebarth et al., 2007; Zebarth et al., 1997; Zebarth et al., 1998).    Current raspberry crop management guidelines aim to maintain soil organic matter (SOM) such that it provides long-term nitrogen supply, regulates soluble forms of nitrogen to ensure plant needs are met. and minimizes nitrogen losses (British Columbia Ministry of Agriculture 2012; Kowalenko et al. 2007; Hart et al. 2006).  However, there is a wide range of management practices for raspberry production in the Fraser Valley (British Columbia Ministry of Agriculture 2012).  As raspberry production is very common in the region, its management has the potential to influence the regional fate of NFF, more specifically NFF uptake by the plant, losses and residual soil N (Kowalenko et al. 2007; Hart et al. 2006).  The range of current nutrient and irrigation management regimes in Fraser Valley raspberry production needs to be assessed to determine how different regimes influence NFF uptake by the plant, residual soil N. and losses to leaching and volatilization.    1.4 Research objectives The goal of this study was to determine how variation in irrigation and nitrogen (N) management affects the fate of N from fertilizer application, with an emphasis on leaching losses, in order to help identify practices that maximize crop acquisition of N and minimize the risks of nitrate leaching.  Specific practices that were compared include mode of N application (fertigation vs broadcast), amount of N applied, and irrigation method (fixed schedule vs based on plant need).    The objectives of this study were to: 1. Determine the partitioning of NFF  from one year’s application among plant, soil and losses to drainage over one growing season and thus 2. Determine the efficiency of different management strategies by comparing the effects of mode of application (fertigation vs. broadcast), irrigation method (fixed schedule  9 vs. based on plant need) and N inputs (50 vs. 100 kg N ha-1) on plant uptake of NFF and losses to leaching.  The fate of NFF was traced by applying 15N labeled fertilizer and then analyzing the movement of 15N into plant tissues, leachate, and soil at four critical times during the subsequent year.  1.5 Thesis Overview The work is presented in self-contained chapters.  Chapter 2 describes the study site, its history and the research methodology.  Chapter 3 presents the research results, organized by sample date and parameter. Chapter 4 discusses the results, their implications, and how they relate to previous studies on similar topics.  Chapter 5 summarizes the main findings of this study and offers suggestions for further research. Two appendices provide supporting information.  Appendix A provides details of the equations used to calculate essential parameters of the nitrogen balance.  Appendix B provides the statistical outputs from PROC MIXED, which was used to statistically test the significance of effects of management practices on the various parameters.     10 Chapter 2 Site History and Methods The field experiment was established in 2008 at the Agriculture and Agri-Food Canada (AAFC) Clearbrook research substation in Abbotsford, BC (Lat. 49°0.702’N and Long. 122°20.097’W), as part of a broader AAFC initiative to mitigate nitrate contamination of vulnerable aquifers (AAFC-SAGES project 1459: Mitigating nitrate contamination of vulnerable aquifers by agricultural production). The experiment was designed to assess the effects that varying fertilizer and water inputs, types of nitrogen (N) inputs (manure vs fertilizer), and the use of between-row cover crops can have on nitrate leaching and crop production.  Analysis of data collected from 2008 through 2011 formed the M.Sc. thesis of Shawn Kuchta (Kuchta 2012).  This study used the existing field experiment as a platform for a more detailed subsequent investigation, using 15N enriched fertilizer, of the movement of NFF into components of the system.  2.1 Study site 2.1.1 Site location and description The study site is located over the Abbotsford-Sumas Aquifer at 510 Clearbrook Rd., Abbotsford, BC (Figure 1.1). The surrounding area consists of agricultural, municipal and industrial lands.  The majority of the agricultural land is used for raspberry and blueberry production as well as for poultry barns, forage and dairy cattle. The nearby industrial lands include a gravel mine and the Abbotsford Airport (YXX). The mean annual precipitation measured between 1981 and 2010 at the Abbotsford Airport Environment Canada weather station was 1537 mm (Environment Canada 2011). This precipitation falls mostly as rain during the months between October and April.  Over the same period the mean monthly minimum temperature in January was 0.4℃ whereas the August mean monthly maximum was 18.2℃.   2.1.2 Site soil properties The soils at the experimental site are well drained and productive for most crops.  They belong to the Marble Hill series and are classified as Orthic Humo-Ferric Podzols under the Canadian soil classification system (Luttmerding 1981). These soils developed on a 20-50 cm deep aeolian deposit which overlays coarse, gravelly glaciofluvial deposits.   11  Kuchta (2012) reported results of basic soil property tests (dry bulk density (BD), soil extract pH and electrical conductivity (EC), soil organic matter (OM) content and hand texturing prior to establishing the research plots (Figure 2.1)(Table 2.1).   Table 2.1    Soil properties and soil characterization in the research field (Kuchta 2012)  Horizon Depth (cm) Dry BDa (g cm-3)  pHb ECb (mS cm-1) OMc (%)  Texture Ap 0-26 1.18 6.0 0.13 7.4 Loam Bfj 26-61 1.27 6.2 0.042 3.2 Loam IIc 62-78+ 1.80 6.3 0.030 1.3 Sand to gravelly sand a Using brass cores ~137 cm3 bpH and EC determined on 1:2 by mass soil water extract c Loss on ignition   Figure 2.1    Soil profile with horizons identified, Clearbrook substation (Kuchta, 2012).  Kuchta (2012) determined the mean soil nitrate-N (NO3-N) content to 30 cm to be 24 mg N kg-1 prior to establishing the research plots.  As per the BC Berry Production Guide classification, this NO3-N content identified the site as a low-fertility site that would require  12 annual N inputs of up to 100 kg N ha-1 N to achieve commercially acceptable levels of production (British Columbia Ministry of Agriculture and Lands 2009).  2.2 Experimental design Randomized complete block design (RCBD) was chosen for this study as it is often used in agronomy and enables scientists to obtain more information from field experiments with fixed resources (Gbur et al. 2012).  Blocking is advisable in agricultural fields because there is often variation in soil characteristics, pathogens and other confounding factors (Gbur et al. 2012).  The choice of which experimental design to use is also influenced by which design maximizes the statistical power and precision for the treatment comparisons that address the original research questions (Gbur et al. 2012).    2.2.1 Experiment establishment The field was left unplanted and received no nutrient inputs between 2004 and 2008 (Kuchta 2012).  The entire experiment was set up in a randomized complete block (RCB) design consisting of eight treatments replicated in four 64 m blocks (Figure 2.2). The blocks were established by planting rows of the red raspberry variety ‘Saanich’ on April 24, 2008.  Each 64 m block included one raspberry row (1.2 m x 64 m) and two alleys (1.8 m x 64 m) (Figure 2.2). Each block was divided into eight, 8 m x 3 m long plots to which eight treatments were randomly allocated.  The four experimental blocks were separated by guard rows of mixed raspberry cultivars.  There were a total of thirty-two treatment plots.  Only four of the eight treatments were used for this study, for a total of sixteen treatment plots (Figure 2.2).  Sub-plots (4 m x 1.2 m) were established within the row of these sixteen treatment plots for this particular 15N tracer study (Figure 2.3).       The raspberry canes were planted on a 0.15 m strip centered in the 1.2 m width row. An approximately 60 cm wide strip on each side of the raspberry canes was maintained free of competing vegetation through the use of herbicides including: Casoron (Chemtura Canada, Dichlobenil [4%]); and Simazine (Drexel Chemical Co., 2-Chloro-4, 6 bis(ethylamino)-s-triazine [90%]).  These herbicides were applied every spring, alternating between the two herbicides each year.  The year of planting (2008) was used to establish a healthy, uniform  13 crop by following the management guidelines prescribed by the BC Berry Production Guide (British Columbia Ministry of Agriculture and Lands 2009).  A more detailed description of research plot establishment can be found in Kuchta (2012).        Figure 2.2  Detailed planar view of the experimental layout  64 m  14  Figure 2.3  Detailed planar view of individual full plot and sub-plot used for 15N study.  2.2.2 Instrumentation A lysimeter is required to measure nitrogen transport using water caught at the bottom of a particular volume of soil. The passive capillary wick sampler (PCAPS) was used in this study.  The PCAPS collect and measure the total flux of soil water moving through a specific volume of unsaturated soil by exerting passive suction on the soil at the top of the sampler (Knutson and Selker 1994). Passive suction is achieved by a saturated fibreglass rope wick which acts as a hanging water column (Boll et al. 1992; Knutson and Selker 1996). These samplers have been shown to accurately measure total flux of a contaminant as well as to be an excellent choice for continuous soil solution monitoring in unsaturated field conditions such as in this study (Boll et al. 1991; Boll et al. 1992).  The main benefit of the PCAPS is that it reduces the diversion of water around the lysimeter, a problem with many other models, by matching the matric potential at the top of the wick to the expected matric potential in the soil (Knutson and Selker 1994). The PCAPS work continuously without electricity and match soil tension better than other instruments (Brandi-Dohrn et al. 1996a).  Furthermore, they can be calibrated to specific soil types by wick matching and adjusting the pan size, increasing the sampler’s accuracy.    Prior to planting, a single PCAPS was installed in the alley and raspberry row of each of the 32 plots to intercept pore water and N moving past the root zone that would potentially enter 15N Sub-Plot Length 4 m  8 m Full Plot Length  Drip Line N 1.2 m  15 the aquifer. Each of the 32 plots contains two PCAPS with their tops located at 0.55 m below the soil surface (Figure 2.4). For each plot, one PCAPS is located at the center of the east facing alley and the other with one edge under the center of the raspberry row below the growing plants.  The matric potential at the top of the PCAPS was matched by Kuchta (2012) to the expected matric potential of the site soil as a function of flux by using the equation of Knutson and Selker (1994).  A wick length of 40 cm was determined to be optimal for the Clearbrook site.  The hanging wetted wick develops suction between 0 and 40 cm/H2O depending on flux and does not require another means of creating suction.  The PCAPS in this study drain water into a container accessible by permanently installed suction tubing (Boll et al. 1992; Knutson and Selker 1994, 1996; Kuchta 2012).  A more detailed description of the PCAPS design and installation can be found in Kuchta (2012).     Figure 2.4  PCAPS sampler installation sites under raspberry row management and associated alley management (Kuchta, 2012).  A Campbell Scientific CR10X datalogger was used to record climate and soil temperature and moisture data, as well as to control the irrigation system. Campbell Scientific CS616 Water Content Reflectometers (based on time domain reflectometry (TDR)) were installed 30 cm from the drip line and approximately 1 m away along the row from the row PCAPS Guard row raspberriesAlleyAccess to sampling linesSampling linesWater and Nitrate movementPCAPSExperimental row raspberries0.55 m0.6 m3.0 m1.2 m0.8 m0.4 m0.6 m 16 location to provide continuous monitoring of the soil volumetric water content between 0 to 30 cm (Kuchta 2012).  2.2.3 Irrigation management The irrigation described here was ongoing between 2008 and 2014.  Irrigation water was supplied by an on-site well and delivered through a drip micro-irrigation system by 2 L h-1 drip emitters. The drip line was located at the center of the 1.2 m wide raspberry row.  Irrigation was applied from spring to fall and controlled by a Campbell Scientific CR10X datalogger (Kuchta 2012).  Two different irrigation managements were utilized; fixed schedule and atmometer scheduled.    The fixed irrigation scheme was intended to mimic the existing practice.  Irrigation timing and duration was determined by ongoing consultation during the growing season with leading growers in the area. This management was scheduled to drip every other day during the summer starting with 4 hours a day early in the summer and increasing to 6 hours a day during peak water need.  In the fall, the irrigation was again reduced to 4 hours.  The second irrigation management applied water to meet plant need (actual evapotranspiration), estimated based on potential evapotranspiration measured on a daily basis using an atmometer combined with crop-specific calculations (Kuchta 2012).  For this type of management the irrigation was run each day to replace water calculated to have been used the previous day.    2.2.4 Nutrient management The general nutrient management for all treatments was the same between 2008 and 2014.  Two nutrient management methods were used: broadcast and fertigation.  Some plots received hand-broadcast granular urea, at an appropriate treatment rate, to the entire 8 m x 1.2 m raspberry row area.  Fertigation management plots received calcium nitrate (Ca(NO3)2) by injection through the irrigation line directly beneath the drip-line along the entire 8 m plot.  The CR-10X datalogger was used to control a Jaeco AgriFarm injection pump to inject fertilizer into the irrigation line.  A weekly amount of Ca(NO3)2 was dissolved into  17 approximately 50 L of water and then injected into the irrigation line at a rate of 0.6 L minute-1 for 10 minutes each day.    In addition to the N applications, all plots received split applications of P, K and micronutrients (as 0-20-20 + micros) on the split application dates.  These were hand-broadcast over the entire 8 m x 1.2 m raspberry row area in each full 8 m plot in order to meet the crop requirements as identified in the BC Berry Production Guide (British Columbia Ministry of Agriculture and Lands 2009).  2.2.5 Experimental treatments The four treatments used in this study were F-100N, F-50N, S-100N and S-50NF (Table 2.2).  Treatment F-100N mimics the management strategy most commonly practiced in the Fraser Valley.  This involved annual fertilizer inputs of 100 kg N ha-1 as urea broadcast over the row. This is at the upper end of the spectrum of fertilizer inputs used by commercial growers on similar soils (Kuchta 2012 as per Mark Sweeny personal communication; 2011 and 2012).  The treatment was irrigated using a fixed schedule that mimicked irrigation practices of nearby commercial growers.   Treatment F-50N involved annual fertilizer inputs of 50 kg N ha-1 as urea broadcast over the row, representing a more modest conventional application rate.  It was also irrigated using the same fixed schedule as F-100N.       Treatment S-100N involved annual fertilizer inputs of 100 kg N ha-1 as urea surface broadcast over the row. This treatment was irrigated according to crop demand using atmometer scheduled irrigation. The aim of S-100N was to investigate whether reduced water use would minimize leaching although having received a high fertilizer application rate. Irrigation was scheduled on daily evaporative demand as determined by atmometer-measured evapotranspiration and a plant growth model (Atmometer Scheduled)  Treatment S-50NF involved annual fertilizer inputs of 50 kg N ha-1 as dissolved calcium nitrate (Ca(NO3)2).  This was applied for 10 minutes a day over 6 weeks as fertilizer injection  18 through the irrigation line.  It was also irrigated according to crop demand using the same method as S-100N.  The aim of treatment S-50NF was to apply less fertilizer more efficiently, to maximize uptake and prevent losses and thus better match N supply to plant requirements.   Table 2.2    Nitrogen, irrigation and alley management treatments used for the 15N study Irrigation Regimea Nitrogenb  kg N ha-1 Fertilizer Type Treatment Fixed Interval 100 Urea - Broadcast 3241 Fixed Interval 50 Urea - Broadcast  85 Atmometer Scheduled 100 Urea - Broadcast  0 Atmometer Scheduled 50 Ca(NO3)2 – Fertigatedc 7 a Irrigation regime was determined by replicating irrigation practices of leading berry producers in the region, (Fixed) or scheduled on daily evaporative demand as determined by atmometer-measured evapotranspiration and a plant growth model (Atmometer Scheduled) b All nitrogen treatments were applied as a split application of urea (46-0-0), per field hectare, unless otherwise stated c Nitrogen applied as dissolved calcium nitrate (Ca(NO3)2).  Delivered 10 minutes per day over 6 weeks as fertilizer injection through the irrigation line.  2.3 Experimental methods 2.3.1 2011 nutrient and irrigation management In 2011, five atom% 15N enriched fertilizers were applied to treatments F-100N, F-50N, S-100N and S-50NF in order to trace NFF throughout the raspberry system.  The conventional fertilizer (urea) treatments (F-100N, F-50N, S-100N) received five atom % enriched 15N labeled urea fertilizer applied as a split application, on April 13, 2011 and on May 12, 2011 on a per field hectare basis. The urea was dissolved in a handheld portable sprayer and sprayed evenly on the ground within the 4 m x 1.2 m sub-plot area.  Buffer zones were left on either end of each full 8 m long plot to prevent cross contamination between sub-plots and the buffer zones received the treatment application rate on a per field hectare basis in the form of hand broadcast non-labeled granular urea.    The fertigation treatment, S-50NF, received N in the form of five atom% 15N enriched calcium nitrate (Ca(NO3)2) by injection through the irrigation line. Fertilizer injection was matched with the timing of the first split N application and the plots received Ca(NO3)2 for  19 six weeks between April 11 and May 22, 2011 on a per field hectare basis.  Due to the use of the irrigation line, Ca(NO3)2 was applied to the whole 8 m length of S-50NF plots.   In 2011, fixed irrigation involved irrigating 241 minutes every even Julian day between June 14 and July 7 and again between September 15 and September 19.  The irrigation was increased to 361 minutes every even Julian Day between July 9 and September 13. In 2011 the fixed irrigation applied 367 mm. In 2011 the atmometer scheduled irrigation applied an average of 136 minutes of irrigation every other day for a seasonal total of 157 mm ha-1.    2.3.2 2012 nutrient and irrigation management In 2012, all treatments received their usual treatment application rate in the form of non-labeled fertilizer.  The urea fertilizer was applied to the entire 8 m x 1.2 m raspberry row in the form of granular urea on April 17 and May 24, 2012.  Fertigation was timed to start at the first split application and ran between April 17 and May 29, 2012 at the same rate, 0.6 L minute-1, 10 minutes per day for six weeks.    In 2012, fixed irrigation involved 241 minutes every even Julian day between June 30 and July 12 and again between September 22 and October 8 and one day at 121 minutes on October 10.  The irrigation was increased to 361 minutes every even Julian Day between July 14 and September 20. In 2012, the fixed irrigation treatment applied 361 mm ha-1. In 2012, the atmometer scheduled irrigation applied an average of 124 minutes of irrigation daily for a seasonal total of 152 mm ha-1.    2.4 Sampling methods A wide range of sample types was taken over the course of two years between April 2011 and November 2013 (Table 2.3).  An experimental year was defined as April to the following March to be able to trace the effects of the growing season into the following winter and rainy period.  Samples included leachate, soil, and raspberry cane, lateral, leaf and fruit samples so as to capture all the possible paths of NFF. A detailed version of the prior sampling history between 2008 and 2011 can be found in Kuchta (2012).  All samples for this study were analyzed for total-N and 15N content by the University of California Stable  20 Isotope Facility in Davis, California (UC SIF). The analysis results were returned with the following data: Total N in solids (!g), Total N in water (mg), 15N (atom %), !15N vs. air.     The N parameters analyzed included:  Total Nitrogen (TN) expressed as kg N ha-1, Nitrogen From Fertilizer (NFF) expressed as kg N ha-1, NFF expressed as a percentage of the amount of NFF applied (%FOA), and NFF expressed as a percentage of TN in the tissue, soil or leachate (%FOT).  At the fourth sample time, June 25, 2012, the N balance is incomplete, as not all leachate samples between January and June 2012 were analyzed in time for inclusion in this analysis.  Table 2.3    Destructive and non-destructive sampling schedule intended to trace fertilizer-N in P32.  Repeated Measure Sample Dates Interval Sample Type Data Generated March 28, 2011 – Nov 2012 Biweekly Leachate Total volume and  NO3-N, NO3-NFF  April 29th  - Nov 11th, 2011 Weekly Primocane and floricane leaves Total N and NFF  April 9th  - Nov 2nd, 2012  Weekly Primocane and  floricane leaves Total N and NFF  Harvest Dates    July 25th -  Aug 19th, 2011 Weekly Ripe fruit Total N and NFF  July 13th - July 30th, 2012 Weekly Ripe fruit Total N and NFF  Destructive Sample Dates    July 4th, 2011 1 Time All plant partsa  and soilb (0-30 cm) Total N and NFF Aug. 22nd, 2011 1 Time All plant parts  and soil (0-30 cm) Total N and NFF  Jan. 24th, 2012 1 Time All plant parts  and soil (0-30 cm)  Total N and NFF  Jun. 25th, 2012 1 Time All plant parts  and soil (0-30 cm)  Total N and NFF  a All plant parts refers to the roots, primocanes, primocane leaves (if present), floricanes, floricane leaves (if present), laterals (if present) and fruiting structures (if present) of 6 plants. b Soil was collected using a 3 cm diameter auger from 6 random spots in the row. Two soil cores were taken from directly below the drip line, two cores 10 cm out from the drip line and two cores 30 cm out from the drip line.  The drip line runs down the center of the raspberry row.  The 6 soil cores were combined to form a composite sample.  2.4.1 Leachate sample collection The PCAPS were sampled bi-weekly between April 2011 and April 2013 for a total of 60 samplings (Table 2.3).  The first sampling times in both the 2011/2012 and 2012/2013 PCAPS sampling years were two weeks after the initial spring application of fertilizer to the experimental plots.  There was a total of 30 sampling times in 2011/2012 and 30 in  21 2012/2013.  The growing season of each year is represented by the initial 13 sampling times and covers a five-month period (mid-April to mid-September).  The period between October and the beginning of April is referred to as the rainy season.  Bi-weekly sampling was carried out using a sampling cart fitted with a Welch dry vacuum pump, vacuum manifold and suction lines and volume calibrated collection cylinders (Kuchta 2012).  Sampling the experimental plots took 1 to 3 days depending on the volume of water present in the lysimeters.  Blocks were always completed in one day and never left to be sampled over multiple days.  The individual plot sampling lines were hooked to the vacuum manifold at each sampling time and all the water present in the lysimeters was pumped out.  The total volume of leachate pumped was recorded and a 200 mL subsample of leachate was collected in high-density polyethylene NalgeneTM sample bottles. The excess drainage was collected in buckets and discarded at the end of each plot onto the guard row, approximately 5 m away from the row PCAPS in order to prevent reintroducing already accounted for 15N into the system and the PCAPS.  The samples were placed in a cooler after sampling and frozen immediately after each days sampling.    The samples remained frozen until they were prepared for analysis at the AAFC Summerland Research station and the UC SIF.  The samples were first analyzed for nitrate (NO3) concentrations at the AAFC Summerland Research station, as UC SIF required accurate nitrate concentrations for 15N-NO3 analysis.  Samples were analyzed using colorimetric analysis by a segmented flow analyzer (SFA, Model 305D, Astoria Pacific International, Clackamas, OR) according to the procedure described and used by Kuchta (2012).  The samples were thawed immediately prior to their analysis.  The sample preparation for UC SIF consisted of defrosting and subsampling 20 ml of the 200 ml sample.  These 20 ml subsamples were filtered with a 0.2 !m pore size filter and 20 ml syringe as per UC SIF instructions.  The filter was changed each time 2 samples had gone through it.  The filter and syringe were rinsed with deionized water after each sample was processed to prevent contamination.  The filtered samples were refrozen in 20 ml high density polyethylene NalgeneTM sample bottles and transported on ice overnight to UC SIF  22 for analysis. Ten percent of the samples were duplicated to check the UC SIF analysis precision.  2.4.2 Plant tissue collection, excluding roots Plant tissue was sampled from the experimental sub-plots on four individual dates over the course of 2011 and 2012 (Table 2.3).  The July 4, 2011 sampling represented system N during the pre-harvest growing season whereas August 22, 2011 represented system N post-harvest.  January 24th, 2012 was intended to give a window into where N was located during the dormant winter season and the June 25, 2012 sampling was intended to provide insight on system N one year after a fertilizer application.   On each sampling date, 6 random floricanes and primocanes were sampled from each sub-plot.  If present at that sampling, the laterals, leaves, and fruiting structures were separated from the cane. At the third sampling, January 24th, 2012, the spent floricanes were sampled.  The leaves and laterals were decomposing and were therefore not separated from the cane.  All tissue samples were stored in paper bags in a dry greenhouse until they were dried in a large industrial drying oven at 60℃ for 72 hours.  After 48 hours and 72 hours of drying the samples were weighed.  If the 72 hour weight was different from the 48 hour weight the samples were dried using additional 24 hours steps until the weights stabilized.  Final weights were recorded for later use.  Once the samples were dry, they were ground to a fine powder using an Arthur H. Thomas Co. grinder equipped with No.30 mesh.  After each sample was ground the grinder was vacuumed clean to prevent sample contamination.  The ground samples were then stored in plastic bags, within dark paper bags, in a dry greenhouse.    In preparation for analysis at the UC SIF, up to 10 mg of each sample was subsampled into tin capsules, pinched closed and set in numbered 96 well holding trays.  The UC SIF sample weight was dependent on the estimated N content and samples were optimized to contain approximately 20 !g of N.  The sample weight calculator on the UC SIF website was used to determine sample target weight.  A duplicate of every 10th sample was included to check the UC SIF precision of analysis. The samples were then shipped to the UC SIF to be sampled for total nitrogen (TN) and 15N content.           23  In addition to the four comprehensive sample dates, leaves were sampled weekly in both 2011 and 2012 throughout the entire growing season (Table 2.3).  Berries were harvested weekly between July 25 and August 19, 2011 and between July 13th and July 30th, 2012.  For every sub-plot, 10 harvest berries from each harvest week were combined to form a composite sample.  These tissue samples were handled according to the aforementioned sampling protocols for tissue samples.  In addition to dry weights, primocane and floricane sub-plot counts were recorded in April and October (after pruning) in both 2011 and 2012.   2.4.3 Soil sampling Soil was collected down to 30 cm using a 3 cm diameter auger from 6 random spots in the row at all four sampling dates in 2011 and 2012 (Figure 2.5, Table 2.3). Three soil cores were taken on both the west and east side of the raspberry row for a total of six cores per sub-plot.  Two cores were taken from 10 cm, 20 cm and 30 cm out from the drip line running down the center of the raspberry row. The sampling locations were systematically rotated for each new sampling date to ensure representative samples. The 6 soil cores were combined and mixed to form a composite sample and then later subsampled.  Soil samples were stored in a refrigerator immediately after sampling was completed.  The remaining soil was replaced by horizon and the stamped to reduce the potential for preferential flow at the sampling locations.  The soil was sieved using a 2mm sieve and the roots were separated out using tweezers within a few days after sampling.  The roots were then washed, air dried and stored in paper bags.  The roots were then handled according to the procedure described in the methods of plant sampling section.       24  Figure 2.5  Soil sampling locations on the raspberry row, within the 15N application area.      After sieving, the soil samples were air dried on plates in a dry greenhouse, which is also where they were stored until ready to be packed for shipping.  They were dried in a large industrial drying oven at 60℃ for 72 hours.  The samples were weighed after 48 hours and 72 hours of drying.  If the 72 hour weight was different from the 48 hour weight, the samples were dried for additional 24 hour steps until the weights measured 24 hours apart stabilized.  Final weights were recorded for later use.  Once dry, the soil was packed into plastic NalgeneTM cups for storage.  In preparation for analysis at the UC SIF, approximately 10 mg of the samples was subsampled into tin capsules, pinched closed and set in numbered 96 well holding trays.  The UC SIF sample weight was dependent upon the estimated N content and samples were optimized to contain approximately 20 !g to 150 !g of N.  The sample weight calculator on the UC SIF website was used to determine sample target weight which was 10 mg of soil.  A replicate was included for every 10th sample to check the analysis precision. The samples were then shipped to UC SIF to be sampled for TN and 15N content.         15N Application Area 4 m !PCAPS 8 m Plot Length  Drip!Line!N x!x!x! x!x!x!1.2 m x = Systematically rotated row soil sampling at 10, 20 and 30 cm from the dripline.  25 2.5 Analysis 2.5.1 Chemical analysis All water, soil and plant data collected were analyzed for total N and 15N at the UC SIF in Davis, CA.  Only twenty-five of sixty leachate sample dates between 2011/2012 and 2012/2013 were analyzed due to budget restraints. All leachate sampled between April 12, 2011 and December 28, 2011 was analyzed.  Three additional dates were also analyzed: April 30, October 31 and November 26, 2012.  The water samples were prepared by bacteria denitrification assay at the UC SIF.  The UC SIF measures isotope ratios of 15N using a ThermoFinnigan GasBench + PreCon trace gas concentration system interfaced to a ThermoScientific Delta V Plus Isotope-ratio mass spectrometer (IRMS). Evaporated water was removed from the sample vials by a double needle sampler into a helium carrier stream at a rate of 25 mL min-1.  After this, a CO2 scrubber trapped N2O in 2 separate traps and concentrated it.  The N2O collected in the first trap and only passed to the second trap when the non-condensing portion of the gas sample has been replaced by the helium carrier.  The second trap was warmed to ambient temperature and the helium carried the N2O to the IRMS via a capillary column at a rate of 1 mL min-1.  This column separates N2O from residual CO2.  A reference N2O peak was used to calculate the isotope ratios from the sample N2O peak.  Finally, the correct 15N values were calculated by adjusting the provisional values using nitrate calibration standards.    The plant and soil samples from each of the four sampling dates were analyzed by the UC SIF using a PDZ Europa ANCA-GSL elemental analyzer interfaced with a continuous flow PDZ Europa 20-20 IRMS. All weekly leaf samples from 2011, and monthly samples from 2012, were analyzed.  The samples were combusted at 1000°C by a reactor.  After this, a reduction reactor removed oxides.  A helium carrier then sent the sample through a water trap to separate N2 and CO2 before entering the IRMS.  The process of combustion and removal of N2 and CO2 was the same for the soil samples except that the soil samples were analyszed using an Elementar Vario EL Cube or Micro Cube elemental analyzer interfaced with a PDZ Europa 20-20 IRMS. The soil N analysis in this study accounted for all forms of N in the soil (inorganic N, ammonium and nitrate, as well as organic N).     26 A quality control program was conducted at the UC SIF laboratory.  All of the analyses included several replicates of different laboratory standards selected to be similar to the samples being analyzed.  The samples preliminary isotope ratio was measured relative to reference gases analyzed with each sample.  The preliminary values were then corrected using the known values of the laboratory standards.  The long-term standard deviation for 15N analysis is 0.3 per mil.   As an extra form of quality control, 10% of the samples sent to UC SIF were duplicated.  The values reported by UC SIF for these duplicates were checked for variability.  All duplicates were found to be within 1 decimal places of per mil values of each other and the majority were the same to the third decimal place.  Therefore the duplicate values were averaged to produce one value for statistical analysis.  2.5.2 Calculations Nitrogen was tracked in the system for a range of sample types: leachate, soil, roots, primocanes, floricanes, laterals, primocane leaves, floricane leaves, fruiting structures and harvest fruit.  For all sample types the following parameters were calculated: Total N (TN in kg N ha-1), nitrogen from fertilizer (NFF in kg N ha-1), percent NFF of applied (%FOA) and percent NFF of total N (%FOT).    The concentration of NFF was calculated using the concentration of 15N in samples reported by UC SIF (Appendix A, equation 1).  The true concentration of 15N from the enriched fertilizer was determined by subtracting the background concentration (0.3663 %) from the reported concentration of 15N.  The concentration of NFF was then calculated as the ratio of true 15N concentration (as a percent) to percent enrichment and the product of the concentration of TN.  The concentration of TN and NFF were converted to a total amount per sub-plot basis.  For leachate, this involved calculating nitrate load in the lysimeter as the product of the concentration in the leachate and leachate volume at each sample time (Appendix A, Equations 2 and 3).  This quantity was converted to a sub-plot (per m2) basis as the product  27 kg N in leachate and the ratio between the area of the sub-plot (4.8 m2) and the leachate catchment area of the PCAPS (0.36 m2).  For soil and roots, this involved calculating the per sub-plot weight of N held in soil or roots as the product of bulk density (1270 kg m-3), sub-plot volume (1.44 m3) and N sample concentration per plot at each sample time (Appendix A, Equations 4 and 5).  For plant components (excluding roots), this involved calculating N content per 6 cane sample as the product of the concentration and dry weight per 6 canes (Appendix A, Equations 6 and 7).  This was then converted to a per sub-plot basis as a product of N determined per 6 canes and the number of canes per sub-plot.   The amounts of TN and NFF per plot were converted to a total amount per hectare of field basis (kg N ha-1).  Each 4.8 m2 sub-plot requires 12 m2 field space due to the alleys which take up 7.2 m2 of extra space.  The fertilizer application rates of 50 and 100 kg N ha-1 were applied on a whole field basis. Therefore, the ratio 10000 m2 ha-1 to 12 m2 of field space was used as the conversion factor (Appendix A, Equation 8).    Cumulative amounts of TN and NFF in leachate between April 11th, 2011 and each sample date were calculated on a per hectare of field basis for each sample date. This involved calculating the sum of N in leachate between April 11th and each subsequent sample date (Appendix A, Equation 9).  Nitrogen removed by way of fruit harvest and pruning was also calculated as a cumulative per field hectare value between July 4 and each subsequent sample date.  For the August 22nd sample date, the removed component included the N per field hectare value of fruit harvested between July 25th and August 19th, 2011.  On January 24th and June 25th, 2012 the removed portion included the N removed via harvest fruit (August 2011 values) as well as N removed by floricanes and laterals (August 2011 values)          Percent fertilizer of applied (%FOA) was calculated as the ratio between NFF (kg N ha-1) and the fertilizer application rate (kg N ha-1) (Appendix I, Equation 10). The ratio between NFF (kg N sub-plot-1) and NFF applied (kg N sub-plot-1) would give the same value.  The fertilizer application rate was either 50 or 100 kg N ha-1 depending on the treatment.  This ratio was then multiplied by 100 %.  Percent fertilizer of total N (%FOT) was calculated as the ratio between NFF (kg N ha-1) and TN (kg N ha-1) (Appendix A, Equation 11).  The ratio  28 between NFF (kg N sub-plot-1) and TN (kg N sub-plot-1) would give the same value.  This ratio was then multiplied by 100 %.  The N budget for all four parameters (TN, NFF, %FOA, %FOT) was calculated at four sampling dates between 2011 and 2012 (July 4th, 2011; August 22nd 2011; January 24th 2012; and June 25th 2012).  This involved accounting for the location of N in the system at each date.  On July 4th, 2011, the budget included the soil and plant N values at that date.  For leachate, this involved the N accumulated in all leachate samples between April 11th and July 4th, 2011.  No N had been removed via harvest or pruning by July 4th, 2011.    On August 22nd, 2011, the budget included the soil and plant N values (corrected for removal) at that date.  The portion removed from the plant included the harvested fruit.  The removed component included the fruit harvested between July 25th and August 19th, 2011. For leachate, this involved the N accumulated in all leachate samples between April 11th and August 22nd, 2011.    On January 24, 2012 the budget included the soil and plant N values at the date.  The removed portion included the N removed in harvest fruit (2011), spent floricanes and laterals (2011).  For leachate, it involved the N accumulated in all leachate samples between April 11th, 2011 and January 24th, 2012.    On June 25th, 2012 the budget included the soil and plant N values at the date.  The removed portion included the N removed in harvest fruit (2011), spent floricanes and laterals (2011).  The leachate portion was incomplete and the cumulative N value was calculated as the sum between April 11, 2011 and December 28, 2011 with the addition of April 30, October 31 and November 26, 2012.         2.5.3 Statistical analysis and considerations It was important to consider whether a general linear model procedure (PROC GLM) or a mixed model procedure (PROC MIXED) should be used to make treatment comparisons for the four sample dates as well as the repeated measures nitrate and leaf tissue data.  The main  29 differences between the two procedures are that PROC GLM treats blocks as fixed whereas PROC MIXED treats them as random. Agricultural RCBD experiments are often analyzed using PROC GLM even though blocks are usually interpreted as being random effects.  It has always been thought that the data provides sufficient information to estimate variance and covariance parameters of random effects with sufficient precision (Yang 2009).  However, recently there have been discussions about whether using PROC MIXED provides more accurate results, especially with agricultural field experiments (Gbur et al. 2012; SAS 2013; UCLA SCG 2013; Yang 2009) in which blocks are random effects.  PROC GLM results may be incorrect if block is considered random.   PROC GLM estimates fixed effects using Ordinary Least Squares (OLS) whereas PROC MIXED was designed for mixed effects models and estimates fixed effects using Generalized Least Squares in a Gaussian error model where estimates are Maximum Likelihood Estimates under normality (SAS 2013; UCLA SCG 2013).  Unlike OLS, which is primarily a descriptive tool used for linear data, MLE chooses values for parameters that maximize the probability of observing dependent values in the sample with the given independent values (Myung 2003). SAS (2013), UCLA SCG (2013) and Yang (2009) all suggest that because PROC GLM is a fixed effect model it may give the wrong denominator for an F-test and the wrong standard error for a treatment when used to analyze data from a RCB experiment.  Although PROC GLM and PROC MIXED have the same statements such as REPEATED and RANDOM, the functions differ greatly (Yarandi 2012). While PROC GLM can be used for both multivariate and univariate repeated measures the results may not be the same as those produced by PROC MIXED.  PROC MIXED does not assume the errors to be identical, allowing the data to exhibit correlation and variability (Yang 2009; Yarandi 2012). Another benefit of PROC MIXED is that when analyzing repeated measures it allows for missing data, covariates within subject and can analyze data in the original form rather than requiring a dimension reducing orthogonal transformation.  Its standard errors also reflect the appropriate covariance structure.  However, one downside to using PROC MIXED for repeated measures is that it can be computationally intensive and may require different runs for different covariance structures whereas PROC GLM is fast and prints all significance tests in one run.  15N studies similar to the this study, such as Banados et al. (2012), used  30 PROC MIXED to analyze tissue data; I therefore chose to treat blocks as random and used PROC MIXED to analyze all the data.  In the case of this study, time course sample data were treated as repeated measures.  Initially an analysis including sample date was performed.  If treatment effect, in the analysis including sample date, was found to be significant (P < 0.05) but the treatment by date interaction was not (P > 0.05) an overall analysis excluding sample date was performed.  The significance of differences between treatments was tested using the P-DIFF command. Treatment comparisons from the model outputs with p values <0.05 were considered significant.     31 Chapter 3 Results  3.1 Nitrogen budget of major system components This section describes treatment effects on the amount of nitrogen (N) in each of the major components of the system (plant, soil, and losses to leaching, harvest and pruning), presented separately as an N budget for each of the four sample dates: July 4, 2011, August 22, 2011, January 24, 2012, and June 25, 2012.    For most combinations of parameters (TN, NFF, %FOA and %FOT) and components (plant tissue, soil, losses to leaching, harvest and pruning) there was a significant date x treatment interaction (Appendix B).  Consequently, the data are summarized by date, with P-values for effect of treatment on each parameter at each date taken from the SLICE option of the PROC Mixed analysis.  The data are summarized this way due to the management perspective of this study for which the fate of NFF is viewed as a balance.   Table 3.1    Effects of N input by irrigation type treatment combinations on total N in soil, plant components, leachate and the entire system on July 4, 2011.  Nitrogen balance on July 4, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  Total Na (kg N ha-1) Pr>Fx Soil 3219 3170 3241 3249 NS NS Plant 85 94 85 89 NS NS Removedb 0 0 0 0 n/a n/a Leachatec 3 8 7 10 NS NS Entire System 3306 3273 3333 3349  NS  NS a Total N includes nitrogen from fertilizer (NFF) b There were no removals of N due to harvest and pruning on July 4th, 2011 c Leachate N is reported as cumulative values since April 11, 2011  3.1.1 Nitrogen budget in growing season: July 4th, 2011 3.1.1.1  Total N July 4th, 2011 The total N in the entire system (soil, leachate and plant components) ranged between 3273 and 3349 kg N ha-1 and did not differ significantly by treatment (Table 3.1).  Treatment had  32 no significant effects on any of the system components.  The amount of N in plant tissue and leachate combined comprised less than 3% of the amount of N in soil.   3.1.1.2 Nitrogen from fertilizer July 4th, 2011 The NFF in the entire system (soil, leachate and plant components) ranged between 41 and 91 kg N ha-1 (Table 3.2).  The high N treatment plots, F-100N and S-100N, held a mean total of 84 and 91 kg N ha-1 NFF respectively whereas the low N treatment plots, F-50N and S-50NF, held approximately 50% less at 39 and 41 kg NFF ha-1 respectively.  At this time, the majority of NFF was found in the soil and plant component whereas little NFF had leached out of the system and none had been removed.    Table 3.2    Effects of N input by irrigation type treatment combinations on nitrogen from fertilizer (NFF) in soil, plant components, leachate and the entire system on July 4, 2011. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on July 4th, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  NFF (kg N ha-1) Pr>Fx Soil 59 a 20 b 65 a 13 b <0.0001 <0.0001a Plant 25 a 17 b 26 a 24 a 0.0024 <0.006 Removedb 0 0 0 0 n/a n/a Leachatec 0 2 0 5 NS NS Entire System 84 a 39 b 91 a 41 b  <0.0001  <0.0001 a P-values reported in the pairwise comparison column are the highest, significant, P-value of all the significant comparisons at that date and for that parameter. b There were no removals of N due to harvest and pruning on July 4th, 2011 c Leachate N reported as cumulative values since April 11, 2011  Nitrogen application rate had an effect on NFF in soil whereas irrigation type did not. Soil in the high N treatment plots, F-100N (59 kg N ha-1) and S-100N (66 kg N ha-1) held more than double the NFF than low N treatment plots, F-50N (20 kg N ha-1) and S-50NF (13 kg N ha-1). The amount of NFF in plant tissue in the high N treatments comprised approximately 41% of the amount of NFF in soil whereas in the low N fixed irrigation treatment it was 85%.  Plant uptake efficiency of NFF was greatest in the S-50NF (fertigation) treatment where the amount of NFF in plant tissue was 185% of the amount of NFF in the soil.  Plant uptake of NFF in treatment S-50NF, was comparable to that of the F-100N treatment and the values  33 were not significantly different.  Additionally, plants in treatment, S-50NF (24 kg N ha-1) as well as both high N treatments, F-100N (25 kg N ha-1) and S-100N (26 kg N ha-1), all held more NFF than treatment F-50N (17 kg N ha-1).  No treatment differences were found in leachate.  3.1.1.3 Percent NFF of applied July 4th, 2011 The %FOA present in the entire system (soil, leachate and plant components) ranged between 78% and 91% and did not differ significantly by treatment (Table 3.3).  As with NFF, irrigation in treatments receiving the same level of nitrogen did not affect %FOA in soil.  The %FOA residing in the soil of high N treatment plots, F-100N (59 %) and S-100N (65 %), was greater than in the low N treatment plots, F-50N (40 %) and S-50NF (25 %).  Plants at this date had taken up 25 to 48% of the fertilizer applied.  Plant uptake of NFF was most efficient in the fertigation treatment (S-50NF), which had a greater percentage of applied N (48 %) than all other treatments.  Plants in treatment F-50N also held more %FOA at 35 % than the high N treatments, F-100N and S-100N, at 25 and 26 %FOA respectively. No treatment differences were found in leachate.   Table 3.3    Effects of N input by irrigation type treatment combinations on percent fertilizer of applied (%FOA) in soil, plant components, leachate and the entire system on July 4, 2011. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on July 4th, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOA Pr>Fx Soil 59 a 40 b 65 a 25 b 0.0011 <0.01a Plant 25 c 35 b 26 c 48 a <0.0001 <0.02 Removedb 0 0 0 0 n/a n/a Leachatec 0  4  0  10 NS NS Entire System 84 78 91 82  NS  NS a P-values reported in the pairwise comparison column are the highest, significant, P-value of all the significant comparisons at that date and for that parameter. b There were no removals of N due to harvest and pruning on July 4th, 2011 c Leachate N reported as cumulative values since April 11, 2011     34 3.1.1.4 Percent NFF of total nitrogen July 4th, 2011 The %FOT present in the entire system (soil, leachate and plant components) ranged between 1% and 3% and was higher in the high N treatments than the low N treatments (Table 3.4).  The proportion of NFF to TN was greatest in the plant and leachate components of the system, ranging from 9% to 29%.  The proportion in soil did not exceed 2%.  The nitrogen application rate had an effect on %FOT in soil resulting in a greater proportion of NFF to TN in soil in the high N treatments, F-100N (1.9%) and S-100N (2.0%), than the two low N treatments, F-50N (0.6%) and S-50NF (0.4%).  Irrigation did not affect %FOT in soil in treatments receiving the same level of nitrogen. The %FOT in the entire plant system was lower in treatment F-50N (19 %) than all other treatments, F-100N (29 %), S-100N (30 %) and S-50NF (27 %) indicating higher non-NFF uptake in the F-50N treatment. The %FOT in the plant system was also higher in S-100N, than in S-50NF.  Plant uptake of NFF in the fertigation treatment, in proportion to TN in the plant, was comparable to treatment F-100N, and better than F-50N even though it had lost the most NFF (proportionately) to leaching.  The proportion of NFF to TN in fertigation treatment plants (27%) was not significantly different from the F-100N treatment (29%) and greater than in F-50N (19%) even though the fertigation treatment lost more %FOT in leachate than all other treatments.      Table 3.4    Effects of N input by irrigation type treatment combinations on percent fertilizer N of total N (%FOT) in soil, plant components, leachate and the entire system on July 4, 2011. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on July 4th, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOT Pr>Fx Soil 1.9 a 0.6 b 2.0 a 0.4 b <0.0001 <0.0001a Plant 29 ab 19 c 30 a 27 b <0.0001 <0.04 Removedb 0 0 0 0 n/a n/a Leachatec 9 b 12 b 4 b 29 a 0.0234 <0.04 Entire System 2.6 a 1.2 b 2.7 a 1.2 b  <.0001  <.0001 a P-values reported in the pairwise comparison column are the highest, significant, P-value of all the significant comparisons at that date and for that parameter. b There were no removals of N due to harvest and pruning on July 4th, 2011 c Leachate N reported as cumulative values since April 11, 2011     35 3.1.2 Nitrogen budget post-harvest: August 22nd, 2011 3.1.2.1 Total N August 22nd, 2011 The total N in the entire system (soil, leachate, plant and removed components) ranged between 2962 and 3194 kg N ha-1 and did not differ significantly by treatment (Table 3.5).  The amount of N in plant tissue and leachate combined, comprised less than 4% of the amount of N in soil.  Nitrogen loss by leaching was greater in F-50N (29 kg N ha-1) than in both scheduled treatments, S-100N (8 kg N ha-1) and S-50NF (13 kg N ha-1).  The amount of TN lost to leaching from the F-100N treatment was not different than any other treatments.  No treatment differences were found for soil, plant tissue, or plant material removed in harvest.   Table 3.5    Effects of N input by irrigation type treatment combinations on total N in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on August 22nd, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  Total Na (kg N ha-1) Pr>Fx Soil 3016 2899 3097 2872 NS NS Plant 60 58 61 52 NS NS Removedb 29 33 28 25 NS NS Leachatec 20 ab 29 a 8 b 13 b 0.0649 <0.04d Entire System 3125 3019 3194 2962  NS  NS a Total N includes NFF b Removed portion is fruit harvested between July 25 and August 19, 2011. Fruit values are not included in the plant portion. c Leachate N reported as cumulative values since April 11, 2011 d The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter.  3.1.2.2  Nitrogen from fertilizer August 22nd, 2014 The NFF in the entire system (soil, leachate, plant and removed components) had decreased by approximately 50% since July 2011 and ranged between 24 and 46 kg N ha-1 (Table 3.6).  Fertilizer application rate had an effect on entire system NFF. The high N treatment plots, F-100N and S-100N, held a mean total of 43 and 46 kg N ha-1 NFF respectively whereas the low N treatment plots, F-50N and S-50NF, held much less at 24 kg N ha-1.  At this time the  36 majority of NFF was found in the soil and plant components. Little additional NFF leached out of the system between July 4 and August 22, 2011 and the portion removed in harvested fruit was less than 10 kg N ha-1.    In the high N treatments, NFF held by the plant components comprised approximately 61% of the amount of NFF in soil whereas in treatment F-50N it was 55%.  Nitrogen from fertilizer in treatment S-50NF plant tissue was 330% of the amount of NFF in the soil.  Soil in the high N treatment plots held more than 7x the NFF (22 kg N ha-1) than the fertigation treatment (3 kg N ha-1).  Irrigation did not affect NFF in soil in treatments receiving the same rate of nitrogen.  Plants in the S-100N treatment held more NFF (15 kg N ha-1) than both F-50N (6 kg N ha-1) and S-50NF (10 kg N ha-1).  The N held in F-100N treatment plants was not significantly greater than the S-50NF plants.  Additionally, N application rates affected the amount of NFF removed by harvest.  Approximately 40% more NFF was removed in high N treatment, harvest fruit than in the low N treatments.  No treatment differences were found in leachate.  Table 3.6    Effects of N input by irrigation type treatment combinations on NFF in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on August 22nd, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  NFF (kg N ha-1) Pr>Fx Soil 22 a 11 ab 22 a 3 b 0.0692 <0.03a Plant 12 ac 6 b 15 a 10 bc <0.0024 <0.03 Removedb 8 a 5 b 8 a 5 b 0.0256 <0.04 Leachatec 1 2 1 6 NS NS Entire System 43 a 24 b 46 a 24 b  0.0247  <0.04 a The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. b Removed portion is fruit harvested between July 25 and August 19, 2011.  Fruit values are not included in the plant portion. c Leachate N reported as cumulative values since April 11, 2011  3.1.2.3 Percent NFF of applied August 22nd, 2014 The %FOA present in the entire system (soil, leachate, plant and removed components) ranged between 43% and 48% and did not differ significantly by treatment (Table 3.7).  No  37 treatment differences were found in the soil component. The irrigation regime had an effect on the proportion of NFF to applied N held in plant components.  Plants in the S-50NF plots held a higher proportion of applied N at 20% than both F-100N and F-50N, which held 12%.  Although the fertigation treatment had more efficient uptake, the proportion of applied N that had leached in that treatment (12%) was much higher than both high N treatments (1%) but not higher than F-50N (5%).   Table 3.7    Effects of N input by irrigation type treatment combinations on %FOA in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on August 22nd, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOA Pr>Fx Soil 22 22 22 6 NS NS Plant 12 b 12 b 15 ab 20 a 0.1115 <0.05a Removedb 8 10 8 9 NS NS Leachatec 1 b 5 ab 1 b 12 a 0.1468 <0.05 Entire System 43 48 46 47  NS  NS a The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. b Removed portion is fruit harvested between July 25 and August 19, 2011. Fruit values are not included in the plant portion. c Leachate N reported as cumulative values since April 11, 2011  A noteworthy loss of NFF from the system was observed between July and August, 2011 that was not accounted for in this study’s N monitoring (Figure 3.1). Approximately 50% of the NFF was lost by an unknown mechanism during this 6-week period. Total NFF values in the entire system stabilized and remained fairly constant after August 2012, which indicates that the majority of the loss occurred between the July and August 2011 samplings.  Irrigation type and fertilizer type appeared to have some effect on total NFF loss, although the losses were not statistically analyzed.  The losses observed in the budget occurred in the cane components (floricanes, primocanes, floricane and primocane leaves, laterals and fruiting structures) and in soil (Table 3.8). Overall, the unaccounted for loss of NFF appeared to be 2x greater in treatments that  38 received high N inputs. The greatest losses were from the soil (9 to 43 %FOA) but also occurred in the floricane (3 to 23 %FOA) and primocane (+4 to 3 %FOA) components.  Losses in soil treatments receiving high N inputs were 4x greater than treatments receiving low N inputs.  In the floricane component, F-50N lost 3x more than both high N treatments and S-50NF lost 7x more than high N treatments.  In the primocane component, scheduled irrigation resulted in a gain of NFF rather than the loss observed under fixed irrigation.  The gain was greatest under fertigation. The combined losses from the cane and soil in high N treatments make up the entire unaccounted for loss whereas in low N treatments between 6 to 8 %FOA was lost somewhere other than from the plant and soil.     !Figure 3.1  The percent fertilizer of applied (%FOA) lost from the entire system by an unknown mechanism at each sampling date. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fertilizer was applied as a split application on April 17 and May 24, 2012.  These data were not statistically analyzed. 0"10"20"30"40"50"60"70"80"90"100"06,Jun,11"26,Jul,11"14,Sep,11"03,Nov,11"23,Dec,11"11,Feb,12"01,Apr,12"21,May,12"10,Jul,12"29,Aug,12"%FOA%Lost%To%Unknown%Mechanism%(%)%F,100N"F,50N"S,100N"S,50NF" 39 Table 3.8    Unaccounted for losses between July 4th and Aug 22nd, 2011 of NFF as a proportion of that applied (%FOA).  These values were not statistically analyzed.   Unaccounted for %FOA lost in entire system %FOA lost from PC Componenta %FOA lost from FC Componentb %FOA lost from Soil %FOA loss in FC, PC and Soil F-100N - 41c - 1.7 - 3.3 - 37 - 42 F-50N - 30 - 3.2 - 9.8 - 9 - 22 S-100N - 45 + 0.13 - 3.2 - 43 - 46 S-50NF - 35 + 4.3 - 23.3 - 10 - 29 a The primocane component includes the primocane and primocane leaf b The floricane component includes the floricane and floricane leaf c A negative value denotes a loss of NFF and a positive value denotes a gain of NFF 3.1.2.4 Percent NFF of total nitrogen August 22nd, 2011 The %FOT present in the entire system (soil, leachate and plant components) ranged between 0.8% and 1.4% and was higher in the high N treatments than the low N treatments (Table 3.9).  The soil in high N treatments still held a greater proportion of NFF to TN than in the fertigation treatment.  In the plant component, the F-50N treatment held less %FOT at 10% than all other treatments.  Conversely, S-50NF (19%) plants only held less %FOT than S-100N (25%), which held the greatest amount of %FOT.  Additionally, the S-100N treatment plants held more %FOT than F-100N (21%).  NFF application rate had an effect on the %FOT held in harvest fruit.  Approximately 40% more %FOT was removed in high N treatments than in low N treatments.  Since April 2011, approximately 6x greater %FOT had accumulated in S-50NF leachate than all other treatments.   3.1.3 Nitrogen budget in mid-winter: January 24th, 2012 3.1.3.1 Total nitrogen January 24, 2012 The total N in the entire system (soil, leachate, plant and removed components) was approximately 3250 kg N ha-1 in all treatments (Table 3.10).  The amount of N in plant tissue and leachate combined comprised less than 4% of the amount of N in soil.  At this date the irrigation regime had an effect on the cumulative N losses due to leaching.  Nitrogen loss due to leaching was 2x greater in the fixed rate irrigation treatments than it was in treatments irrigated based on plant need.  No treatment differences were found among any of the other system components.   40 Table 3.9    Effects of N input by irrigation type treatment combinations on %FOT in soil, plant components, leachate, removed N and the entire system on August 22, 2011. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on August 22nd, 2011  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOT Pr>Fx Soil 0.7 a 0.4 ab 0.7 a 0.1 b 0.0470 <0.02a Plant 21 b 10 c 25 a 19 b <0.0001 <0.02 Removedb 26 a 15 b 26 a 18 b 0.0005 <0.02 Leachatec 6 b 6 b 7 b 37 a 0.0005 <0.0006 Entire System 1.4 a 0.8 b 1.4 a 0.8 b  0.0333  <0.04 a The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. b Removed includes fruit harvested between July 25 and August 19, 2011. Fruit values are not included in the plant portion. c Leachate N reported as cumulative values since April 11, 2011   Table 3.10    Effects of N input by irrigation type treatment combinations on total N in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on January 24th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  Total Na (kg N ha-1) Pr>Fx Soil 3168 3014 3413 3039 NS NS Plant 15 14 13 11 NS NS Removedb 47 51 45 39 NS NS Leachatec 39 a 49 a 23 b 21 b 0.0021 <0.05d Entire System 3269 3128 3494 3110  NS  NS a Total N includes NFF b The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. c Leachate N reported as cumulative values since April 11, 2011 d The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter.  3.1.3.2 Nitrogen from fertilizer January 24, 2012 The NFF in the entire system (soil, leachate, plant and removed components) was very similar to that of August 22, 2011 and ranged between 28 and 46 kg N ha-1 (Table 3.11).  The  41 S-100N treatment plots had retained more NFF at 46 kg N ha-1 than both low N treatment plots.  However, system NFF in the F-100N (34 kg N ha-1) treatment was not significantly higher than either F-50N (28 kg N ha-1) or S-50NF (28 kg N ha-1).  At this time, the majority of NFF was found in the soil and the sum of NFF in the plant, removed and leached portion was approximately equivalent to that found in soil. Little additional NFF had leached out of the system and all plant components had been removed in the form of harvest fruit, spent floricanes and their laterals.    The S-100N treatment (30 kg N ha-1) retained more NFF in soil than S-50NF (14 kg N ha-1).  However, fixed rate irrigation did not result in higher soil NFF values than irrigation based on plant need.  The amount of N removed in high N treatments was approximately 1.75x greater than that in low N treatments.  No treatment differences were found in the plant or leachate component.  Table 3.11    Effects of N input by irrigation type treatment combinations on NFF in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on January 24th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  NFF (kg N ha-1) Pr>Fx Soil 18 ab 17 ab 30 a 14 b 0.2268 <0.05a Plant 2 2 3 2 NS NS Removedb 10 a 6 b 11 a 6 b 0.0093 <0.02 Leachatec 3 3 2 7 NS NS Entire System 34 ab 28 b 46 a 28 b  0.1886  <0.05 a The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. b The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. c Leachate N reported as cumulative values since April 11, 2011  3.1.3.3 Percent NFF of applied January 24, 2012 The %FOA present in the entire system (soil, leachate, plant and removed components) had increased from that accounted for in August 2011 and ranged between 34% and 57% (Table  42 3.12).  Treatment F-100N had retained the least NFF in proportion to that applied at 34% whereas F-50N and S-50NF had the best recovery at 56 and 57% respectively.  The proportion of NFF to applied that had leached from the fertigation treatment (14%) was higher than the S-100N treatment (2%) but not greater than in the high N treatments.  Treatment did not have an effect on %FOA in any other system components.   Table 3.12    Effects of N input by irrigation type treatment combinations on %FOA in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on January 24th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOA Pr>Fx Soil 18 34 30 27 NS NS Plant 2 4 3 3 NS NS Removeda 10 13 11 13 NS NS Leachateb 3 ab 6 ab 2 b 14 a 0.1596 <0.04c Entire System 34 b 56 a 46 ab 57 a  0.1000  <0.04 a The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. b Leachate N reported as cumulative values since April 11, 2011 c The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter.  3.1.3.4  Percent from NFF of total nitrogen January 24, 2012 The %FOT present in the entire system (soil, leachate, plant and removed components) was approximately 1% in all treatments (Table 3.13). At this date, the greatest %FOT was found in the removed component.  Nitrogen from fertilizer was a very small component of TN in soil. The proportion of NFF to TN was approximately 115x greater in the plant, removed and leached portions than in soil.    Soil in treatment S-100N had greater proportion of NFF to TN residing in soil than treatment S-50NF plots at 0.9 and 0.4 % respectively.  At this date the plants were dormant and no significant differences were found among treatments in the plant component. NFF application rate had an effect on the %FOT removed by harvest and pruning.  Approximately  43 30% greater %FOT was removed in the high N treatments, F-100N (58%) and S-100N (61%), than in the low N treatments, F-50N (33%) and S-50NF (44%). Additionally, the removed component from S-50NF plots held greater %FOT than material removed from F-50N plots.   Table 3.13    Effects of N input by irrigation type treatment combinations on %FOT in soil, plant components, leachate, removed N and the entire system on January 24, 2012. Letters (a, b) denote significant pairwise comparisons.  Nitrogen balance on January 24th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOT Pr>Fx Soil 0.6 ab 0.6 ab 0.9 a 0.4 b 0.2130 <0.05a Plant 6 4 7 5 NS NS Removedb 58 a 33 c 61 a 44 b <.0001 <0.0006 Leachatec 7 b 4 b 10 b 30 a 0.0139 <0.02 Entire System 1.0 0.9 1.4 0.9  NS  NS a The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. b The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. c Leachate N reported as cumulative values since April 11, 2011  3.1.4 Nitrogen budget in the following growing season: June 25th, 2012 3.1.4.1 Total nitrogen June 25, 2012 The total N in the entire system (soil, leachate, plant and removed components) ranged between 3212 and 3410 kg N ha-1 and did not differ significantly by treatment (Table 3.14).  The amount of N in plant tissue, in the removed portion and leachate combined, comprised less than 5% of the amount of N in soil.  At this date the irrigation regime had an effect on the cumulative N losses due to leaching, although the dataset is not complete.  Nitrogen loss due to leaching was greater in F-100N (41 kg N ha-1) and F-50N (50 kg N ha-1) than it was in S-100N (25 kg N ha-1) and S-50NF (23 kg N ha-1).  No treatment differences were found among any of the other the system components.    44 Table 3.14    Effects of N input by irrigation type treatment combinations on Total N in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. Nitrogen balance on June 25th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  Total Na (kg N ha-1) Pr>Fx Soil 3348 3080 3277 3100 NS NS Plant 65 64 63 49 NS NS Removedb 47 51 45 39 NS NS Leachatec 41 a 50 a 25 b 23 b 0.0031 <0.05d Entire System 3501 3246 3410 3212  NS  NS a Total N includes NFF b The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit harvested from entire plot was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. c Leachate N reported as cumulative values between April 11, 2011 and Dec 28, 2011 with the addition of N recovered from lysimeters on April 30, 2012.  The N dataset is incomplete, as not all leachate samples between January and June 2012 were analyzed in time for inclusion in this analysis. d The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. 3.1.4.2  Nitrogen from fertilizer June 25, 2012 The NFF in the entire system (soil, leachate, plant and removed components) ranged between 24 and 57 kg N ha-1 (Table 3.15).  The S-100N treatment held more system NFF at 57 kg N ha-1 than both low N treatment plots.  System NFF in the S-50NF treatment was not significantly lower than in F-100N although it had received half the N rate.  At this time, the majority of NFF was found in the soil and the sum of NFF in the plant, removed and leached portion was approximately equivalent to that found in soil, with the exception of S-100N which held more than double the NFF in soil than that in the plant, removed and leached portion.  The S-100N treatment (40 kg N ha-1) had retained more NFF in soil than all other treatments.  As on January 24, 2012, the amount of N removed in high N treatments was approximately 1.75x greater than that in low N treatments.        45 Table 3.15    Effects of N input by irrigation type treatment combinations on NFF in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. Nitrogen balance on June 25th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  NFF (kg N ha-1) Pr>Fx Soil 23 b 12 b 40 a 13 b 0.0055 <0.04a Plant 5 3 5 3 NS NS Removedb 10 a 6 b 11 a 6 b 0.0093 <0.02 Leachatec 3 3 2 7 NS NS Entire System 41 ab 24 b 57 a 29 b 0.0025 n/a <0.03 n/a a The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. b The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit harvested from entire plot was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. c Leachate N reported as cumulative values between April 11, 2011 and Dec 28, 2011 with the addition of N recovered from lysimeters on April 30, 2012.  The N dataset is incomplete, as not all leachate samples between January and June 2012 were analyzed in time for inclusion in this analysis.  3.1.4.3 Percent NFF of applied June 25, 2012 The %FOA present in the entire system (soil, leachate, plant and removed components) increased between January 24th and June 25th, 2012 and ranged between 41% and 58% (Table 3.16). There were no differences among treatments.  The cumulative leachate results between April 2011 and June 25, 2011 are incomplete as not all samples taken between January and June 2012 were analyzed in time for inclusion in this analysis.  However, the proportion of NFF to applied N recovered in the fertigation treatment leachate (14%) was higher than that in the S-100N treatment (3%) but not greater than in the high N treatments.  Treatment did not have an effect on %FOA in any other system components.  3.1.4.4 Percent NFF of total nitrogen June 25, 2012 The %FOT present in the entire system (soil, leachate, plant and removed components) ranged between 0.8% and 1.7% (Table 3.17). The %FOT was found to be highest in the treatment S-100N.  The removed and leached portion of the system held the greatest proportion of %FOT. Proportionately, NFF was still lower in soil at this date as shown by the  46 %FOT being approximately 110x greater in the plant, removed and leached portions than in soil.    Table 3.16    Effects of N input by irrigation type treatment combinations on %FOA in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. Nitrogen balance on June 25th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOA  Pr>Fx Soil 23 24 40 25 NS NS Plant 5 6 5 6 NS NS Removeda 10 13 11 13 NS NS Leachateb 3 ab 6 ab 2 b 14 a 0.1582 <0.04c Entire System 41 48 57 58  NS  NS a The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. b Leachate N reported as cumulative values between April 11, 2011 and Dec 28, 2011 with the addition of N recovered from lysimeters on April 30, 2012.  The N dataset is incomplete, as not all leachate samples between January and June 2012 were analyzed in time for inclusion in this analysis. c The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter.  Soil in treatment S-100N (1.2%) had greater %FOT residing in soil than all other treatments.  At this date the plants were in early stages of growth and no significant differences were found among treatments. NFF application rate had an effect on the %FOT removed by harvest and pruning; treatments F-100N (58%) and S-100N (61%) held a greater proportion of NFF to TN than F-50N (33%) and S-50NF (44%). Additionally, the removed component from S-50NF plots held greater %FOT than material removed from F-50N plots.  The fertigation treatment had leached the greatest proportion of NFF to TN.  3.2 Temporal dynamics 3.2.1 Entire system nitrogen over time The amount of total N in the system was relatively constant through time ranging between 2962 and 3494 kg N ha-1 between July 4, 2011 and June 25, 2012.  Total N in the entire system reported on a per hectare basis did not differ significantly by treatment at any date. At  47 all four dates the amount of TN in soil was orders of magnitude larger than the amount of TN found in all other system components combined.  However, in different components of the system, total N and NFF did change dramatically through time. This section describes the temporal dynamics of N in both the entire system and individual parameters for each of the main system components: leachate, soil, plant tissue, and N removed from the system in prunings and harvested berries.  Table 3.17    Effects of N input by irrigation type treatment combinations on %FOT in soil, plant components, leachate, removed N and the entire system on June 25, 2012. Letters (a, b) denote significant pairwise comparisons. Nitrogen balance on June 25th, 2012  F-100N F-50N S-100N S-50NF Treatment effects Pairwise comparison  %FOT Pr>Fx Soil 0.7 b 0.4 b 1.2 a 0.4 b 0.0087 <0.05a Plant 7 4 7 6 NS NS Removedb 58 a 33 c 61 a 44 b <.0001 <0.0006 Leachatec 7 b 4 b 9 b 27 a 0.0361 <0.04 Entire System 1.2 b 0.8 b 1.7 a 0.9 b  0.0105  0.0101 a The pairwise comparison column reports the highest significant P-value of all the significant comparisons at that date and for that parameter. b The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. c Leachate N reported as cumulative values between April 11, 2011 and Dec 28, 2011 with the addition of N recovered from lysimeters on April 30, 2012.  The N dataset is incomplete, as not all leachate samples between January and June 2012 were analyzed in time for inclusion in this analysis.  The overall percentage of NFF that was recovered varied significantly over time and ranged between 34 and 91 percent depending on treatment and sample date (Figure 3.2).  The main factor effects of treatment, sample date and the interaction between treatment and sample date for total system %FOA were significant and the P-values were 0.2032, <0.0001 and 0.6542 respectively.  Nitrogen from fertilizer reported on a per hectare basis reflected the respective treatment application rates at all sample dates showing higher NFF values in high N treatment plots and lower values in low N treatment plots (Figure 3.3). Nitrogen from fertilizer varied over time and ranged between 24 to 91 kg N ha-1 depending on treatment and sample time.  The main factor effects treatment, sample date and the interaction between  48 treatment and sample date for NFF were significant and the P-values were <0.0001, <0.0001 and 0.0308 respectively.  Similarly, the distribution of NFF contained in the individual system components changed dramatically through time (Figure 3.4 to 3.7).    Figure 3.2  The proportion of nitrogen from fertilizer (NFF) to applied N expressed as a percent (%FOA), grouped by treatment. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for treatment, date and treatment x date interaction are 0.2032, <0.0001 and 0.6542 respectively.   0"10"20"30"40"50"60"70"80"90"100"F-100N" F-50N" S-100N" S-50NF"Propor%on'NFF'to'N'Applied'in'2011'(%)'04-Jul-11"22-Aug-11"10-Jan-12"25-Jun-12" 49  Figure 3.3  The effects of nitrogen input by irrigation type treatment combinations on NFF in the entire system at each sampling date. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fertilizer was applied as a split application on April 17 and May 24, 2012.  P-values for effects of treatment, date and treatment by date are <0.0001, <0.0001 and 0.0308 respectively.  Error bars attached to S-100N points are (+/- 6.33 kg N ha-1) standard error calculated from the pooled variance.   Figure 3.4  The NFF balance on July 4, 2011. Negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components. At this date no plant material had yet been removed from the system.  0"10"20"30"40"50"60"70"80"90"100"6,Jun,11"26,Jul,11"14,Sep,11"3,Nov,11"23,Dec,11"11,Feb,12"1,Apr,12"21,May,12"10,Jul,12"29,Aug,12"NFF#in#System#(kg#N#ha11 #)##F,100N"F,50N"S,100N"S,50NF"!70$!60$!50$!40$!30$!20$!10$0$10$20$30$40$F!100N$ F!50N$ S!100N$ S!50NF$NFF#Distribu+on#.#July#4,#2011#(kg#N#ha.1 #)#Plant$Soil$Leachate$ 50  Figure 3.5  The NFF balance on August 22, 2011. Negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components.  At this date harvest fruit had been removed from the system. The removed portion is cumulative and includes N recovered in fruit harvested between July 25 and Aug 19, 2011.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.    Figure 3.6  The NFF balance on January 24, 2012. Negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components. The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012. Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop. Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in. The decomposing laterals and leaves were not separated from the cane.  !70$!60$!50$!40$!30$!20$!10$0$10$20$30$40$F!100N$ F!50N$ S!100N$ S!50NF$NFF#Distribu+on#.#August#22,#2011#(kg#N#ha.1 #)#Plant$Removed$Soil$Leachate$!70$!60$!50$!40$!30$!20$!10$0$10$20$30$40$F!100N$ F!50N$ S!100N$ S!50NF$NFF#Distribu+on#.#January#24,#2012#(kg#N#ha.1#)#Plant$Removed$Soil$Leachate$ 51  Figure 3.7  Nitrogen from fertilizer balance on June 25, 2012. The negative values denote leachate and soil NFF residing below ground.  Positive values denote NFF residing above ground in the plant components.  The removed portion is cumulative and includes N recovered in harvest fruit and floricane prunings including decomposing laterals and leaves.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane.  3.2.1.1 General leachate dynamics A complete dataset of TN and NFF values in leachate over the entire experimental period (April 2011 to June 2012) is not available because leachate samples taken between January 2012 and June 2012 have not been analyzed at the time of writing of this thesis.  The total mass of nitrate (NO3--TN), composed of both non-NFF and NFF nitrate, lost by leaching between the first fertilizer application on April 13, 2011 and December 29, 2011 are comparable to those values found between 2009 and 2011 (Kuchta, 2012).  Nitrate leaching values from 2009 to 2011 were approximately 20 to 50 kg N ha-1 per year (Kuchta, 2012), the majority of this N being leached between April and December.  Nitrate leaching values found in this study were very similar, ranging 21 to 49 kg N ha-1 between April and December 2011.  It is likely that the nitrate losses between January and April 2012 comprised only a small portion of the losses between April 2011 and April 2012.   !70$!60$!50$!40$!30$!20$!10$0$10$20$30$40$F!100N$ F!50N$ S!100N$ S!50NF$NFF#Distribu+on#.#June#25,#2012#(kg#N#ha.1#)#Plant$Removed$Soil$Leachate$ 52 This study found that the majority of nitrate leached was not derived from NFF (Tables 3.4, 3.9, 3.13 and 3.17) (Figures 3.9 and 3.10).  NFF lost as nitrate (NO3--NFF) comprised a small proportion of the NO3--TN losses which ranged between 21 and 49 kg N ha-1 between April 13, 2011 and December 29, 2011.  Only 5 to 10 kg N ha-1 of the overall nitrate leached came from fertilizer in 2011.  NO3--NFF lost by leaching as a proportion of the total was lower than anticipated; ranging between 3 to 6% in F-100N, F-50N and S-100N, and 14% in S-50NF.  It was possible to estimate the losses of NFF in this study after December 29, 2011 using the winter loss values from the Kuchta (2012) study because the overall nitrate loss results in this study were comparable.  Estimated losses of NFF to leaching between January and June 2012 for F-100N, F-50N, S-100N and S-50NF are 3.2, 2.1, 4.7 and 1.8 kg N ha-1 respectively (Table 3.18). This indicates that S-100N, although it retained NFF during the 2011 growing season, may have a delay in the timing NFF lost to leaching. This is consistent with the findings of Kuchta (2012) which found that the majority of overall nitrate was leached by the end of fall rains and mid-winter rains resulted in far lower nitrate losses, presumably because most had already been lost from the system.   Table 3.18    An estimate of cumulative NO3- -NFF (kg N ha-1) in leachate between January and June 2012.  Estimates were derived using the product of cumulative values of NO3- -TN in leachate between January and June 2012 and %FOT in leachate from April 2012.  April was used as it was the only date leachate was analyzed during that time period. Treatment Jan – June 2012 cumulative  NO3- -TNa in leachate (kg N ha-1) April 2012 %FOT in leachate Estimated 2012  NO3- -NFF (kg N ha-1) F-100N 63 4.99 3.2 F-50N 74 2.91 2.1 S-100N 108 4.40 4.7 S-50NF 81 2.22 1.8 a Cumulative NO3- -TN in leachate values were calculated by Agriculture and Agri-Food in 2013.   53  Figure 3.8  Effects of N input by irrigation type treatment combinations on total cumulative NO3-N in leachate. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 5.48 kg N ha-1) standard error calculated from the pooled variance. Cumulative values are incomplete after Dec 27, 2011 because after that date only a select few sample dates were analyzed.    Figure 3.9  Effects of N input by irrigation type treatment combinations on NFF cumulative NO3-N in leachate. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, 0.0016 and 1.000 respectively.  Error bars represent (+/- 1.90 kg N ha-1) standard error calculated from the pooled variance. Cumulative values are incomplete after Dec 27, 2011 because after that date only a select few sample dates were analyzed.    0"10"20"30"40"50"60"70"80"90"100"26,Feb,11"6,Jun,11" 14,Sep,11"23,Dec,11"1,Apr,12" 10,Jul,12" 18,Oct,12"26,Jan,13"Cumula&ve)Total)NO30N)(kg)N)ha01 ))F,100N"F,50N"S,100N"S,50NF"0"1"2"3"4"5"6"7"8"9"10"26,Feb,11"6,Jun,11" 14,Sep,11"23,Dec,11"1,Apr,12" 10,Jul,12" 18,Oct,12"26,Jan,13"Cumula&ve)NO3-NFF)(kg)N)ha-1 ))F,100N"F,50N"S,100N"S,50NF" 54  3.2.1.2 System changes between preharvest and postharvest The best recovery of NFF occurred on the first sample date, July 4, 2011 at which time %FOA ranged between 78 to 91 % (Figure 3.2).  Overall, the NFF values were the highest in July and the lowest in January (Figure 3.3).  Between July and August 2011, %FOA and NFF values decreased by approximately 46%.  The overall decrease was mainly influenced by the 65% decrease in soil NFF between those two dates (Figure 3.4 and 3.5).  On July 4, 2011, plant N was the only system component aside from soil that held a significant amount of NFF. Between 60 and 66 %FOA remained in the soil of high N treatments, 40% in treatment F-50N and 25% for treatment S-50NF on July 4, 2011.  This decreased to approximately 22% for treatments F-100N, F-50N and S-100N and 6% for S-50NF after harvest. The NFF held in the plant component decreased after harvest as a portion of NFF originally held in the plant was redistributed into the removed component as harvest fruit.   An increase was seen in both the TN and NFF cumulative leaching losses since April 2011 (Figure 3.8 and 3.9).  The amount of TN increased by 4.5x in fixed rate irrigation treatments between July and August, 2011.  There was only a 1.2x increase in treatments irrigated based on plant need.  Nitrate from NFF began to show up in leachate one week after the first application, around April 22, 2011. The increase of NFF in leachate was very slight between April and July 2011.  On July 11, 2011, just after the irrigation was increased from 4 to 6 hours per day in fixed rate plots, NFF began to show up in greater quantity, especially in treatments F-100N, F-50N and S-100N. There was a shift between July and August 2011, from no treatment differences to the fixed irrigation treatment, F-50N, having lost more cumulative TN than both S-100N and S-50NF.  3.2.1.3 System changes into the dormant season Into the dormant season, %FOA values in F-100N decreased by approximately another 21% whereas the low N treatments F-50N and S-50NF increased by approximately 15% (Figure 3.2).  The NFF values in S-100N plots remained constant between August 22, 2011 and January 24, 2012 (Figure 3.3).  As between July and August 2011, some NFF held in the plant in August was added to the removed component as floricane prunings.  Also, the  55 relocation of plant NFF which occurs after the growing season was apparent by the low values of NFF in the plant portion on January 24, 2012.    Another small increase was also seen in both the TN and NFF cumulative leaching losses since April 2011 (Figure 3.8 and 3.9). Cumulative TN in leachate increased by another 2x in fixed rate irrigation treatments between August 2011 and January 2012 whereas there was virtually no change in the amount that had leached from treatments irrigated based on plant need.  By September 19, 2011, both of the fixed irrigation treatments had cumulatively leached more total NO3-N than both of the scheduled treatments.  By September 19, 2011, 28, 40, 8 and 13 kg N ha-1 total NO3-N had leached in the treatments F-100N, F-50N, S-100N and S-50NF respectively.  This trend persisted through to December 28, 2011 even after scheduled treatments had presumably begun leaching.  After the rainy season, both fixed irrigation treatments had cumulatively lost more TN than both S-100N and S-50NF. By December 28, 2011, the total NO3-N leached in all the treatments, F-100N, F-50N, S-100N and S-50NF, was 40, 49, 23 and 21 kg N ha-1.  The increase of NFF in leachate between July and August 2011 was negligible in proportion to TN.   3.2.1.4 System changes into the next growing season  Total N and %FOA in the entire system were no longer different between treatments into the next growing season. However, S-100N showed the best retention of NFF within its system due to little leaching and a high rate of N.  System %FOA and NFF appeared to increase between January 24 and June 25, 2012 in the high N treatments (Figure 3.2), but further statistical analysis is required.  System NFF in F-100N, F-50N and S-50NF remained relatively constant.  An increase was seen in S-100N of 10 kg N ha-1. The NFF held in the soil component increased in both high N treatments and decreased in both low N treatments.  Between 23 to 40 percent of NFF was stored in the top 30 cm of soil into the next growing season.   The proportion of NFF in the soil remained relatively constant during the dormant period and into the spring for fixed irrigation treatments but rebounded in treatments irrigated on plant demand. Plant uptake of the previous year’s NFF from soil was observed in all treatments. The S-100N treatment retained the most Plant NFF in soil into the next growing season and N retained in S-50NF, F-100N and F-50N was comparable.  On June 25, 2012,  56 approximately 94% of N in the plant was from sources other than NFF applied in 2011.  Between January and June 2012, the leachate dataset is incomplete, as not all leachate samples between January and June 2012 were analyzed in time for inclusion in this analysis. Therefore, the cumulative amount between April 2011 and June 25, 2012 could not be reported.   Figure 3.10  Effects of N input by irrigation type treatment combinations on total NO3-N in leachate at each sampling event. Between April and December 2011 all sampling dates are depicted.  After Dec 27, 2012 only three more events in January, April, October and November were statistically analyzed. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 0.81 kg N ha-1) standard error calculated from the pooled variance.   In general, fixed rate irrigation treatments (F-100N and F-50N) leached NO3- -TN and NO3- -NFF during the growing season while irrigation was running whereas the treatments irrigated based on plant need (S-100N and S-50NF) leached during the fall rainy season (Figure 3.10, 3.11). On a cumulative basis the fixed irrigation rate treatments lost more overall TN than the atmometer scheduled treatments.  This was in spite of the fact that the fertigation treatment lost the greatest proportion of NFF to that applied.  Ultimately the NFF values on a per hectare basis were not different among treatments whereas the TN cumulatively leached was greater in the fixed rate irrigation treatments than in treatments irrigated based on plant need. 0"2"4"6"8"10"12"14"16"18"20"26(Feb(11"6(Jun(11" 14(Sep(11"23(Dec(11"1(Apr(12" 10(Jul(12" 18(Oct(12"26(Jan(13"Total&NO3*N&(kg&N&ha*1 )& F(100N"F(50N"S(100N"S(50NF" 57  Figure 3.11  Effects of N input by irrigation type treatment combinations on NO3-NFF in leachate at each sampling event. Between April and December 2011 all sampling dates are depicted.  After Dec 27, 2012 only three more events in January, April, October and November were statistically analyzed. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are 0.2729, <0.0001 and 0.5538 respectively.  Error bars represent (+/- 0.33 kg N ha-1) standard error calculated from the pooled variance.   3.3 Internal Cycling of Plant Nitrogen This section reports on the internal cycling of plant N parameters: Total nitrogen (TN); nitrogen from fertilizer (NFF); NFF as a percent of applied (%FOA) and NFF as a percent of total N (%FOT).  This section will present results found within the plant component which is comprised of the floricane, primocane, floricane leaf, primocane leaf, laterals, root and fruiting structures.  This section will refer to the plant component as a whole as well as the individual components.  The main factor effects of treatment, sample time and the treatment x sample time interaction for all the entire plant system and the components can be found in Appendix B.    3.3.1 Timing of uptake  Nitrogen from fertilizer was present in the plant component of all treatments on the first sampling date, July 4, 2011 (Figure 3.12).  Root uptake of NFF was best in the fertigation treatment whose plants had taken up approximately 48 % FOA.  Root uptake, as shown by 0"0.5"1"1.5"2"2.5"3"3.5"4"4.5"5"26*Feb*11"6*Jun*11" 14*Sep*11"23*Dec*11"1*Apr*12" 10*Jul*12" 18*Oct*12"26*Jan*13"NO3$NFF&(kg&N&ha$1 )& F*100N"F*50N"S*100N"S*50NF" 58 %FOA in the plant component, was slightly less in treatments F-100N (25%), F-50N (35%) and S-100N (26%).  The NFF values in the plant component and the NFF contribution to TN (%FOT) in the plant component were affected by treatment.  Greater %FOT was found in high N treatments than low N treatments on both July and August 2012.  On the post-harvest sampling date, August 22, 2011, all parameter values in the plant component had decreased.  The relocation of NFF within the plant component after the growing season was apparent by the low values of NFF in the plant component and an increase of NFF found in roots and soil between August 22, 2011 and January 24, 2012 (Table 3.19, 3.20, 3.21, 3.22).   Figure 3.12  Effects of N input by irrigation type treatment combinations on NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) at each sampling event. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are 0.0002, <0.0001 and 0.0780 respectively.  Error bars represent (+/- 1.78 kg N ha-1) standard error calculated from the pooled variance. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane.    0"5"10"15"20"25"30"6(Jun(11"26(Jul(11"14(Sep(11"3(Nov(11"23(Dec(11"11(Feb(12"1(Apr(12"21(May(12"10(Jul(12"29(Aug(12"NFF#in#Plant#(kg#N#ha.1 )# F(100N"F(50N"S(100N"S(50NF"b#a#a#a# a#a#b#b# 59 Table 3.19    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on July 4, 2011. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. Sample Type NFFa in Plant Components (kg N ha-1) July 4, 2011 Pr>Fx  F-100N F-50N S-100N S-50NF Treatment Effect Pairwise Comparison Floricane 0.58 0.40 0.54 0.61 0.5674 NS Floricane leaf 9.48 ab 7.53 b 10.96 a 10.01 ab 0.1476 <0.03 Lateral 1.49 1.03 1.52 1.43 0.5131 NS Fruit structure 5.47 b 4.48 b 6.32 ab 8.69 a 0.0180 <0.01 Floricane Componentb 17.02 13.44 19.34 20.74 Not statistically analyzed Primocane 2.23 a 0.81 b 1.37 b 1.27 b 0.0096 <0.03 Primocane leaf 5.54 a 3.09 b 5.03 a 1.85 b 0.0005 <0.03 Primocane Componentc 7.77 3.9 6.4 3.12 Not statistically analyzed Roots 0.002 ab 0.002 ab 0.003 a 0.001 b 0.1982 <0.05 Total Plant 24.80 a 17.35 b 25.74 a 23.85 a 0.0069 <0.01 a Plant NFF in plant components was calculated using plant component dry weights including floricane, floricane leaf, lateral, fruiting structure, primocane, primocane leaf and root components. b Floricane component includes canes, leaves, fruiting structures and laterals c Primocane component includes canes and leaves                       60 Table 3.20    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on August 22, 2011. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. Sample Type NFFa in Plant Components (kg N ha-1) August 22, 2011 Pr>Fx  F-100N F-50N S-100N S-50NF Treatment Effect Pairwise Comparison Floricane 0.64 ab 0.39 b 0.81 a 0.59 ab 0.0907 <0.01 Floricane leaf 3.59 2.38 5.46 2.62 0.5425 NS Lateral 1.87 ab 0.97 b 2.19 a 1.29 b 0.0682 <0.05 Harvest fruitb 7.64 a 4.81 b 7.65 a 4.56 b 0.0256 <0.03 Floricane Componentc 13.74 8.55 16.11 9.06 Not statistically analyzed Primocane 1.83 a 0.60 b 1.09 ab 1.56 a 0.0220 <0.02 Primocane leaf 4.27 ac 1.71 b 5.47 a 3.70 c 0.0015 <0.05 Primocane  Componentd 6.1 2.31 6.56 5.26 Not statistically analyzed Roots 0.002 0.002 0.002 0.001 0.3620 NS Total Plant 19.84 a 10.86 b 22.66 a 14.32b <.0001 <0.03 a Plant NFF in plant components was calculated using plant component dry weights including floricane, floricane leaf, lateral, fruiting structure, primocane, primocane leaf and root components. b A composite sample of fruit harvested between July 25, 2011 and Aug 19, 2011 was used to calculate total cumulative fruit biomass and NFF. Fruit was harvested from entire plot 5 times between those dates and was removed from the field.  No fruit was left to drop.   c Floricane component includes canes, leaves, fruiting structures and laterals d Primocane component includes canes and leaves                    61 Table 3.21    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on January 24, 2012. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. Sample Type NFFa in Plant Component (kg N ha-1) Jan 24, 2012 Pr>Fx  F-100N F-50N S-100N S-50NF Treatment Effect Pairwise Comparison Floricane 0.80 0.54 0.77 0.49 0.3854 NS Floricane leaf 0 0 0 0 n/a n/a Lateral 0 0 0 0 n/a n/a Fruit structure 0 0 0 0 n/a n/a Floricane Componentb, c 0.80 0.54 0.77 0.49 Not statistically analyzed Primocane 2.47 a 1.85 ab 2.58 a 1.57 b 0.0533 <0.03 Primocane leaf 0 0 0 0 n/a n/a Primocane  Componentd 2.47 1.85 2.58 1.57 Not statistically analyzed Roots 0.0049 a 0.0034 ab 0.0034 ab 0.0026 b 0.0757 <0.01 Total Plant  3.28 2.39 3.35 2.07 0.9415 NS a Plant NFF in plant components was calculated using plant component dry weights including floricane, floricane leaf, lateral, fruiting structure, primocane, primocane leaf and root components. b Floricane component includes canes, leaves, fruiting structures and laterals c The spent floricanes were pruned out on January 24, 2012. Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. d Primocane component includes canes and leaves                  62 Table 3.22    The effects of N input by irrigation type treatment combinations on the NFF in the plant component (floricanes, primocanes, floricane leaves, primocane leaves, fruiting structures, laterals and roots) on June 25, 2012. Letters (a, b) denote significant pairwise comparisons.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. Sample Type NFF in Plant Components (kg N ha-1) June 25, 2012 Pr>Fx  F-100N F-50N S-100N S-50NF Treatment Effect Pairwise Comparison Floricane 0.99 a 0.44 b 0.98 a 0.73 ab 0.0039 <0.001 Floricane leaf 1.88 1.08 1.59 1.07 0.9345 NS Lateral 0.46 0.18 0.36 0.20 0.9404 NS Fruit structure 0.67 0.39 0.57 0.43 0.9961 NS Floricane  Componentb,c 4.0 2.09 3.5 2.43 Not statistically analyzed Primocane 0.23 0.30 0.40 0.26 0.9754 NS Primocane leaf 0.46 0.40 0.68 0.23 0.9630 NS Primocane  Componentd 0.69 0.7 1.08 0.49 Not statistically analyzed Roots 0.001 b 0.002 b 0.005 a 0.002 b 0.0006 <0.0003 Total Plant  4.68 2.79 4.59 2.99 0.8060 NS a Plant NFF in plant components was calculated using plant component dry weights including floricane, floricane leaf, lateral, fruiting structure, primocane, primocane leaf and root components. b Floricane component includes canes, leaves, fruiting structures and laterals and are the primocanes from 2011. c The spent floricanes were pruned out on January 24, 2012. Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. d Primocane component includes canes and leaves  Nitrogen uptake, as shown by NFF concentration in the floricane leaves, behaved similarly for all treatments (Figure 3.13).  The increase of NFF in floricane leaves but overall decrease of TN in floricane leaves up to May 2011 suggest an increased need for NFF as availability of N in the stored plant pool declined, especially as the primocanes began to form (Figure 3.14).  The uptake of NFF by the plant component remained constant after May 2011, whereas TN in the plant component continued to decrease.  Initially the TN and NFF values are much higher in primocane leaves than the floricane leaves.  The primocanes used NFF more readily than the floricanes.  This trend was not seen in the fertigation treatment, S-50NF, probably due to its greater initial NO3- loss resulting in a lower available NFF pool in the soil. A drop of NFF is seen in floricanes close to harvest and an increase of NFF in primocane leaves in the early fall.  This dynamic may be due to the relocation of NFF from the floricane to the primocane or a loss from the floricane by an unknown mechanism that returns NFF to the soil, which is then taken up by the primocane.  63  Figure 3.13  Effects of N input by irrigation type treatment combinations on NFF concentration in floricane leaves at each sampling event.  15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 0.5 !g N mg-1 sample) standard error calculated from the pooled variance. The decomposing leaves were left to drop to the ground.    Figure 3.14  Effects of N input by irrigation type treatment combinations on NFF concentration in primocane leaves at each sampling event. 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  P-values for effects of treatment by date, treatment and date are <0.0001, <0.0001 and <0.0001 respectively.  Error bars represent (+/- 0.72 !g N mg-1 sample) standard error calculated from the pooled variance. The decomposing leaves were left to drop to the ground.   0"2"4"6"8"10"12"14"16"18"20"26(Feb(11"6(Jun(11" 14(Sep(11"23(Dec(11"1(Apr(12" 10(Jul(12" 18(Oct(12"NFF#in#Floricane#Leaves#(μg#N#mg31 #sample)#F(100N"F(50N"S(100N"S(50NF"0"2"4"6"8"10"12"14"16"18"20"26(Feb(11"6(Jun(11" 14(Sep(11"23(Dec(11"1(Apr(12" 10(Jul(12" 18(Oct(12"26(Jan(13"NFF#in#Primocane#Leaves#(μg#N#mg31 #sample)#F(100N"F(50N"S(100N"S(50NF" 64 3.3.2 Partitioning of N Total N in the whole plant as well its individual components showed very few differences between treatments (Tables 3.1, 3.5, 3.10 and 3.14).  However, NFF values among individual components did show treatment differences (Table 3.19, 3.20, 3.21, 3.22).  Between July 4, 2011 and June 25, 2012, NFF in the whole plant ranged between 2 and 26 kg N ha-1.  NFF values were highest in the whole plant on the first sampling, July 4, 2011, and slowly decreased over the course of the year with the lowest values occurring during the dormant period on January 10, 2012.  The NFF residing in the whole plant had decreased slightly across all treatment types between July and August, 2011, with the fertigation treatment decreasing the most (by 19 %FOA). The NFF in the whole plant t had increased slightly by June 25, 2012.    Table 3.23    Yield (kg ha-1) of Saanich raspberry as influenced by N and irrigation management in 2010 and 2011.  No treatment differences were found.  Treatment 2010 Yielda (kg ha-1) 2011 Yieldb (kg ha-1) F-100N 11412 10434 F-50N 10307 12274 S-100N 9397 10618 S-50NF 10447 10204 a In 2010 total yield measured using 12 canes per plot (all plots standardized to 48 canes) b A composite sample of fruit harvested between July 25, 2011 and Aug 19, 2011 was used to calculate total cumulative fruit biomass per plot. Plots were did not have a standardized number of canes. Fruit was harvested from entire plot 5 times between those dates and was removed from the field.  No fruit was left to drop.   No differences were found in fruit and fruit structure biomass at any sample date (Table 3.23).  On the first two sampling dates there were no differences in whole plant biomass.  However on January 24, 2012, F-50N had greater biomass than S-100N in both the roots and the whole plant.  Into the next growing season whole plant and root biomass was greater in both F-50NF and S-100N than F-100N and F-50NF.  The majority %FOA in the whole plant  65 was found in the annual components whereas the least was found in roots and canes at all sample dates except January 24, 2012 (Figures 3.15, 3.16, 3.17, 3.18). During the dormant season the majority of NFF was held in the primocanes, the spent floricanes which had not been pruned out and the roots (Figure 3.17).   Into the next growing season, an increase in NFF from 2011 is seen among the individual plant parts, although at far lower values than in 2011 (Figure 3.18).       Figure 3.15  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on July 4th, 2011 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.   0"10"20"30"40"50"F)100N" F)50N" S)100N" S)50NF"NFF#of#Applied#N#within#Plant#Component#(%)#Frui/ng"Structure"Laterals"Floricane"Leaf"Floricane"Primocane"Leaf"Primocane"Roots" 66  Figure 3.16  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on August 22, 2011 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011.  Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane.   Figure 3.17  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on January 24, 2012 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane. 0"10"20"30"40"50"F)100N" F)50N" S)100N" S)50NF"NFF#to#Applied#N#within#Plant#Components#(%)#Fruit"Laterals"Floricane"Leaf"Floricane"Primocane"Leaf"Primocane"Roots"0"10"20"30"40"50"F)100N" F)50N" S)100N" S)50NF"NFF#of#Applied#N#within#Plant#Component#(%)#Floricane"Primocane"Roots" 67   Figure 3.18  Nitrogen from fertilizer partitioning in N input by irrigation type treatment combinations on June 25, 2012 (F-100N, F-50N, S-100N and S-50NF). 15N enriched fertilizer was applied as a split application on April 13 and May 12, 2011. Fruit was harvested between July 25 and Aug 19, 2011 and the floricanes were pruned out on January 24, 2012.  Fruit was harvested from entire plot and was removed from the field.  No fruit was left to drop.  Six pruned floricanes were kept for analysis and the rest were discarded in alleyways and tilled in.  The decomposing laterals and leaves were not separated from the cane.  3.4 Environmental Conditions During the 6-week period between the July and August 2011 sample dates, 89.9 mm of precipitation fell (Table 3.24) (Environment Canada 2011, 2012).  After the first urea application on April 13, 2011 and the first sampling date on July 4, 2011, 279.3 mm fell.  During the six weeks of fertigation application (April 11 to May 22, 2011), 256.2 mm of precipitation occurred.  The soils in the experimental field are moist year round due to irrigation and environmental conditions and have a 2:1 water:soil pH of 6.0 for the 26 cm deep Ap horizon (Kuchta 2012).    0"10"20"30"40"50"F)100N" F)50N" S)100N" S)50NF"NFF#of#Applied#N#within#Plant#Components#(%)#Frui/ng"Structure"Laterals"Floricane"Leaf"Floricane"Primocane"Leaf"Primocane"Roots" 68 Table 3.24    Monthly precipitation and mean monthly air temperatures at the Abbotsford Airport (~1 km from the experimental site) 2011 – 2013 (Environment Canada, 2011, 2012)  Monthly Precipitation Totals (mm)a Mean Monthly Air Temperature (°C)a Month 2011 2012 2013 2011 2012 2013 January 328 181 136 3.7 2.5 2.2 February 88 153 103 3.1 5.3 5.1 March 151 206 215 7.3 5.8 7.4 April 164 141 144 7.7 10.4 9.4 May 135 71 103 11.8 13.0 13.8 June 49 139 81 15.6 15.3  16.4 July 75 53 4 17.4 17.6  18.9 August 16 4 26 18.5 18.9  19.4 September 91 7 123 17.2 15.9  16.1 October 107 262 63 10.5 10.9  10.1 November 166 160 218 4.9 6.7  5.8 December 92 161 118 3.3 3.4  1.8 Annual Total 1463 1538 1332.6     a Data from the Environment Canada National Climate Data and Information Archive for Abbotsford A.      69 Chapter 4 Discussion The goal of this study was to determine how variation in irrigation and nitrogen (N) inputs affect the fate of N from one year’s fertilizer application, with an emphasis on leaching losses, in order to help identify practices that maximize crop acquisition of N and minimize the risks of nitrate leaching.  Specific practices that were compared include the N inputs (50 vs. 100 kg N ha-1), mode of N application (fertigation vs broadcast) and irrigation method (fixed schedule vs based on plant need).  To determine the influence of different management practices, the fate of N from fertilizer (NFF) was traced by applying 15N labelled fertilizer and then analyzing the movement of 15N into plant tissues, leachate, and soil at four critical times during the subsequent year.    Plant uptake, leaching losses and soil retention are linked responses of the system to management practices. Consequently, this chapter is organized by management practices, with the major fates of NFF discussed in tandem.  In addition, improved understanding of temporal dynamics of NFF within the plant and losses from the system as a whole can provide insight on other means to improve NFF acquisition and retention in the system. Thus, this chapter is organized as follows:   1. N application rate  (50 vs. 100 kg N ha-1) 2. Mode of application (fertigation vs. broadcast) 3. Irrigation regime (fixed rate vs. based on plant need) 4. Timing of uptake and internal redistribution of NFF in the plant component 5. Possible fates of unaccounted-NFF.  4.1  N application rate   Efficient fertilizer management regimes should match N inputs and total N supply to meet plant demand as well as to reduce nitrate leaching.  This study demonstrated that reducing N inputs from 100 to 50 kg N ha-1 did not affect berry yield or N (both TN and NFF) ending up in leachate but did affect NFF acquisition by the plant as well as residual soil NFF into the following growing year.  Higher N inputs resulted in greater soil NFF values three months after application. However, the influence of application rate disappeared in the next growing  70 season (Table 3.2, 3.6, 3.11, 3.15). The proportion of NFF to TN (%FOT) was also influenced by N inputs.  The whole plant held a greater %FOT in treatments with higher N inputs (Figure 3.12).  High N inputs resulted in greater NFF being removed from the system by berry harvest and pruning (Table 3.6) (Figure 3.5 and 3.6).  Low N input resulted in berries holding less NFF but these treatments produced yields comparable to that of treatments receiving higher N inputs (Table 3.20).  Similarly, the findings of Rempel et al. (2004) and Kuchta (2012) suggest N application rates between 40 to 50 kg N ha-1 are sufficient to maintain raspberry production.  Rempel et al. (2004) also found that N application rates as low as 0 kg N ha-1 did not have a significant effect on plant dry weight or TN content.   4.2 Mode of application Alternative crop management strategies such as fertigation aim to match N supply to plant requirements.  In my study, the mode of application had a strong influence on the fate of NFF and fertigation proved to be more efficient than all other management strategies.  Fertigation improved plant acquisition of NFF, resulting in reduced leaching of total NO3-N and improved plant N uptake relative to broadcast application of fertilizer.  The fertigation treatment resulted in the least cumulative nitrate leaching overall, although it lost more NFF to leaching than other treatments. This anomaly is likely the result of fertilizer type, calcium nitrate (Ca(NO3)2), and excess water from precipitation.  The calcium nitrate would have been vulnerable to leaching immediately as it was applied. In contrast, N from granular urea, which was used for the broadcast treatments, is not immediately vulnerable to leaching. The urea N must first be hydrolyzed by urease and then the ammonium ions subjected to nitrification before it would have been vulnerable to leaching. Consequently, any excess precipitation that may have occurred during the 6 week fertigation period would likely have leached more nitrate from the fertigated plots than from the broadcast urea-treated plots.  Water applied to deliver the fertigated N would also have retained higher soil moisture contents than in the broadcast treatments, making it more likely that any given precipitation event would drive leachate past the root zone.  The environmental conditions in the 12 weeks after the first application were indeed wet (Table 3.24), with 279 mm of precipitation falling during that time; the NFF applied as Ca(NO3)2 was more susceptible to leaching than that  71 applied to the other three treatments because it was already in the form of nitrate.  Neilsen and Neilsen (2002) found in apple orchards that excess water as a result of fertigation can leach N below the root zone.  The fertilizer type used by Neilsen and Neilsen (2002) was also Ca(NO3)2.  Under fertigation management, berry yields (Table 3.23), plant component biomass, and TN and NFF in the plant components were comparable to those of all other treatments (Table 3.1, 3.2, 3.5, 3.6, 3.10, 3.11, 3.14, 3.15) although it was the most conservative treatment in terms of N and overall water inputs. These findings agree with previous studies, which found that fertigation management can optimize yield (Gurovich 2008) while reducing N losses to leaching (Neilsen and Neilsen 2002; Neilsen et al. 2008).  The reduced losses to leaching under fertigation management in this study agree with the findings of Kuchta (2012), who demonstrated that the combination of reduced water inputs and appropriate N application rates (50 kg N ha-1) minimize leaching losses overall.  Less overall NO3--TN was lost by leaching despite apparent greater cumulative NO3--NFF losses to leaching.  At the post-harvest sampling, %FOA values had all dropped and there were no differences between F-100N, F-50N and S-100N which all held 22 percent NFF of that applied (%FOA) whereas S-50NF was much lower at 6 %FOA.  Into the dormant season the %FOA in soil under fertigation management rebounded and the influence of mode of application disappeared.  This rebound was not due to a change in the form of N as the method of soil analysis used in this study accounted for all forms of N in the soil.  The rebound may have been the result of NFF from leaf litter and prunings re-entering the soil pool after decomposing.  Leaf litter generally drops to the ground around the base of the plants whereas prunings are cut up and placed in the alley where they are eventually tilled in.         72 4.3  Irrigation regime Irrigation regime had a strong influence on N losses by leaching, and reduced water inputs were not found to negatively affect productivity nor yield. The irrigation regime based on plant demand was found to be more efficient in retaining and using NFF (Tables 3.3, 3.7, 3.12, 3.16). In my study, reduced water inputs under plant demand based irrigation did not influence harvest biomass. The 2011 yield results agree with previous years’ results (2009-2010), which also showed reduced irrigation inputs did not adversely affect crops (Kuchta 2012) (Table 3.23).  Treatments with irrigation based on plant need lost significantly less total NO3-N and fertilizer-derived NO3-N by leaching than the fixed rate irrigation treatments by January 2012 (Table 3.9).  These finding agree with two studies on drip irrigated apples which showed greater overall N losses to leaching under fixed rate irrigation compared to irrigation based on plant need (Neilsen and Neilsen 2002 and Neilsen et al. 2008)  In 2009 and 2010, Kuchta (2012) found that drainage volume and nitrate losses from the root zone moved in tandem and were positively correlated.  He also found that irrigation based on plant need reduced water inputs by 50% whereas under fixed irrigation in 2010, growing season losses were sensitive to irrigation management. In this study, irrigation treatments showed the same timing of nitrate loss as found by Kuchta (2012) in 2010 (Figure 3.10 and 3.11).  This study also found that between April 12, 2011 and December 29, 2011, treatments receiving irrigation based on plant need leached 50% less nitrate overall than those with fixed rate irrigation (Figure 3.8).   The fixed rate treatments lost nitrate during the growing season whereas those irrigated on plant need lost N in the autumn.  The trend of higher growing season leaching in fixed irrigation likely persisted into 2011 because the fixed irrigation rates were kept the same as those from 2010.  Kuchta (2012) explained these treatment differences by examining the water inputs and duration of irrigation application under fixed rate irrigation as well as the resulting soil moisture status.  He found that under the fixed rate irrigation regime total water inputs (precipitation plus irrigation) were 148 to 122 mm in excess of expected plant water use.  In addition to this, Kuchta (2012) found that soil moisture at 30 cm depth was consistently higher (3-6% VWC) and that soil water potential at a depth of 55 cm under fixed rate irrigation was more saturated than under irrigation regimes  73 based on plant need.  As in the Kuchta (2012) study, the higher soil moisture content in fixed rate irrigation treatments in this study likely increased the unsaturated hydraulic conductivity of the soil, resulting in an increased flow of water out of the root zone and increased N losses by leaching.   In general, the soil N pool was very large as shown by NFF comprising a small portion of the TN in soil (Tables 3.4, 3.9, 3.13, 3.17).  TN in soil did not vary among treatments and sample times.  The trend in soil NFF was opposite to that observed in the plant.  The small proportion of NFF in soil remained relatively constant during the dormant period and into the spring for fixed schedule irrigation treatments but rebounded in treatments irrigated on plant demand.  4.4  Timing of uptake and internal redistribution of NFF in the plant  This study demonstrated that the timing of N application is an important consideration in regard to plant uptake and internal cycling of NFF. Total N in plant material was not different among treatments at any sampling (Tables 3.1, 3.5, 3.10, 3.14) indicating all treatments supplied equal total N to the plants.  Several observations indicate that floricanes use stored N (in soil) early in the season, between April and June.  For example, leaf NFF content between April and June 2011 indicated that primocanes, which form in May/June after the floricanes have formed leaves, used NFF more readily than floricanes (Figures 3.13 and 3.14). Furthermore, between July 2011 and June 2012, TN and NFF in roots was very low suggesting N uptake early on comes mostly from soil, not from N stored in the root system.  This is contrary to the findings of other studies, such as Rempel et al. (2004) which indicate that early uptake of N comes from that stored in roots. On July 4th, 2011 floricane and its components (laterals, fruiting structures and leaves), collectively, held more NFF than the primocane and its components (leaves) (Table 3.19). By July 4th, 2011, root uptake of NFF from soil was already occurring in both primocanes and floricanes, regardless of treatment, as shown by the presence of NFF in the plant component.  Preferential partitioning of NFF to new growth was found in a similar study on blackberry (Mohadjer et al. 2001). If NFF uptake by the floricane is less than thought between April and June it indicates that N  74 fertilizers could be applied at a later date to coincide with primocane need and prevent loss from the system.   TN and NFF decreased in soil by the post-harvest sampling on August 22, 2011, likely due to N being pulled from the soil pool by the plant and leaching due to water inputs (Table 3.5 and 3.6).  Total plant NFF decreased in all treatments, between July and August 2011.  Similarly, NFF had decreased in the floricane component during this time period and the majority of NFF was held in the primocane component (Table 3.19 and 3.20).  The decrease in floricane NFF content agrees with previous studies, which found N uptake by floricanes to be rapid in spring and early summer (April to June) with the maximum floricane N content occurring between June and July (Kowalenko 1994; Rempel et al. 2004). The greatest portion of NFF being held in the primocane in August agrees with a study on the uptake, partitioning and storage of NFF in red raspberry, which found that the majority of NFF from the second split N application was in primocanes by mid-harvest (Rempel et al. 2004). Three other studies on NFF partitioning in blackberry found the majority of N to be in both the roots and the primocanes (Malik et al. 1991; Mohadjer et al. 2001; Naraguma et al. 1999). Interestingly, the NFF content held in the primocane and its leaves, under a fixed rate irrigation regime, decreased between July and August 2011 (Table 3.19 and 3.20).  This is contrary to the findings of previous studies which found N uptake by the primocanes has been found to be rapid in July and August with the maximum content occurring in late summer and early fall.  However, under an irrigation regime based on plant need, the trend agreed with studies and the NFF content in the primocane and its leaves increased between July and August 2011.  It is likely that irrigation based on plant need resulted in less loss from the system and positively influenced the availability of N for the plant.  NFF behaved as expected after harvest and towards the end of the growing season.  Leaf NFF concentration decreased rapidly in floricanes after harvest until floricane leaf drop in September.  Primocane vegetation continued to grow towards the end of the season, between September and November 2011, although leaf NFF concentration also decreased slowly after harvest (Figure 3.14).  The slow decrease of NFF in primocane leaves after harvest indicates remobilization of NFF from the leaf to the cane.  In general, over the course of the growing season and into the next year, the least plant N, both TN and NFF, was found in roots and  75 canes, but rather in annual components (Table 3.19, 3.20, 3.21, 3.22).  Unlike other studies, NFF concentration in roots in this study was very low, perhaps due to sampling error.   4.5 Unaccounted for fates of NFF Despite several years of varying N input, the total N in the system did not differ among treatments and except for a single decline between July and August it remained relatively constant into the next growing season.  A loss of NFF from the system was observed between July and August, 2011 that was not accounted for in this study’s N monitoring (Figure 3.1) (Table 3.8). Approximately 50% of the NFF was lost by an unknown mechanism during this 6-week period. The NFF values in the entire system stabilized and remained fairly constant after this loss during the growing season. Irrigation type, N inputs and fertilizer type appeared to have an effect on NFF loss suggesting the loss mechanism may not be the same for all treatments.   Overall, the unaccounted-for loss appeared to be 2x greater in treatments that received high N inputs. The greatest losses were from the soil but floricane and primocane components also lost N.  Interestingly, NFF was lost from the floricane component in the low N treatments whereas the high N treatments lost NFF preferentially from the soil. This suggests that the total loss of NFF from each treatment may be the result of more than one loss mechanism, which operated to different extents in different treatments.   The mechanisms of N removal that were monitored in this study were harvest removals, pruning removals and nitrate leaching.  Leaching was expected to account for the greatest NFF losses to the environment. Loss mechanisms that were not monitored in this study include; nitrous oxide (N2O) losses as a result of nitrification and denitrification, as well as ammonia volatilization and gaseous losses from the plant. The soil N analysis in this study accounted for all forms of N in the soil (inorganic N, ammonium and nitrate, as well as organic N), which therefore excludes immobilization as an option for the unknown loss mechanism.  The processes, that were not monitored in this study, are discussed in the following sections.    76 4.5.1 Losses from Soil Approximately 90% of global N2O emissions are thought to come from soils.  The soil processes that produce N2O emissions are nitrification and denitrification (Freney 1997).  Denitrification generally occurs under anaerobic conditions when nitrate is available along with labile sources of carbon. Nitrification generally occurs rapidly in agricultural soils whenever ammonium substrate is available to the nitrifying bacteria under aerobic conditions. The default IPCC values for the proportion of applied N that is lost as nitrous oxide by denitrification is up to 4 to 5% (Jassal et al. 2008).  The observed 52 to 57 % loss of applied N in treatments by August 2011 is therefore likely too large to have been all lost as a result of denitrification. The environmental conditions during the 6 week period also did not favour denitrification as the loss mechanism (Table 3.8).  N application as nitrate under the fertigation treatment added additional water and nitrate nitrogen inputs to the system prior to July 2011 and, with the addition of rain during that time, conditions may have been favourable for denitrification.  However, the unknown losses occurred after that time period.   Nitrification is pervasive in Fraser Valley soils during the growing season (Dean et al. 2000). Recent greenhouse gas studies at the Summerland AAFC research facility strongly suggest that early spring N2O losses are mainly due to denitrification but that during fertilizer application and into the summer nitrification is a contributing factor (Denise Neilsen personal communication, 2015).  Given that the decline in total system NFF occurred during the summer, when soil was not saturated-anaerobic and conducive to denitrification, I speculate that N2O losses during nitrification was the primary cause of the decline in total system N, particularly in the 100 kg N ha-1 treatments which would have had more residual ammonium substrate in the soil.  Enhanced nitrification would likely have been stimulated shortly after the April and May, 2011 applications of urea.  If this was the case nitrification may have continued from spring into and throughout the summer.  Furthermore, Freney (1997) concluded that N2O emissions from soil are affected by N inputs but not by fertilizer type, suggesting that both urea and calcium nitrate may result in N2O emissions.  This conclusion agrees with the overall finding in this study that low N input treatments lost less NFF than high N input treatments.     77 Nitrification may not explain the losses observed under fertigation because the applied N was already in the form of nitrate. Also, the losses in that treatment occurred mostly from the floricane component (Table 3.8). This may have been because the fertigation treatment had better uptake of NFF leaving proportionally less to be lost from the soil and more to be lost from floricanes.  The greatest gain of NFF in the primocane component was observed in fertigation treatment suggesting that NFF may have been remobilized from the floricane to the primocane.   Inadequate sampling of root systems could have contributed to the low NFF values for the root component, and high sampling error for the root component could have contributed to the apparent temporal changes in total system N, especially for the fertigated treatments, which had better uptake of NFF and thus presumably held more NFF in roots through the dormant season.     Volatilization of ammonia gas into the atmosphere generally occurs in warm, moist soils that have a higher pH, often closely following a surface application of urea or ammonium-based fertilizer.  Hydrolysis of urea by urease, liberating ammonium, is generally very rapid in agricultural soils.  Conditions for volatilization occurs when ammonium reacts with hydroxide (OH-) to form water and ammonia gas, but this only happens at high pH values. Soil pH values were below 6.0 (Kuchta 2012), and it therefore seems unlikely that ammonia volatilization was a significant cause of N loss from the soil.    4.5.2 Losses from the plant The 3 to 23 %FOA losses observed in the cane component may have been a result of early leaf senescence in floricanes that was not adequately accounted for in the leaf sampling program.  However ammonia emissions from the plant component may be another possible loss mechanism.  The likelihood that crops will emit ammonia increases if fertilizer applications substantially exceed crop need (Holtan-Hartwig and Bockman 1994).  Several studies have found that between 5 to 50 kg N ha-1, or up to 17 %FOA, may be lost from various crops as ammonia and that the greatest losses occur during rapid growth periods with peak emissions at; leaf senescence, times of N remobilization and at temperatures between  78 30-35 degrees (Schepers and Raun, 2008; Schjoerring et al. 1998).  These conclusions agree with the findings in this study, which found losses occurred during a period of rapid and peak growth, that occurred during summer temperatures.  Shifts in temperature have also been found to cause plants to become ammonia sources rather than sinks (Schjoerring et al. 1998; Husted and Schjoerring 1996).  Higher air concentrations of ammonia have been shown to result in lower plant emissions.  This experiment was located in close proximity to a poultry barn and in an area of intense agriculture so air concentrations may have been high.  Some other factors that influence plant ammonia emissions are the plant type, wind, humidity, and light.  More research is needed on ammonia emission from raspberry crops during the growing season.      79 Chapter 5 Conclusion  The management of N and irrigation in agriculture, in order to increase the efficiency of N-fertilizer use by crops, is important to reduce impacts, such as nitrate leaching, on the environment. Raspberry production is very intensive in the Fraser Valley in particular and thus its management is important due to the potential impacts it could have on the Abbotsford Sumas aquifer.  This study has contributed an improved understanding of how to minimize and mitigate land use impacts of raspberry production on groundwater while optimizing yield and minimizing environmental damage to water resources.  This study assessed four management strategies and showed some important findings relating to irrigation management, nutrient management and plant use of fertilizer nitrogen    This study found that irrigation regime had a strong influence on N losses by leaching, as did the prior study at the site (Kuchta 2012).  Reduced water inputs reduced NO3- leaching and did not negatively affect productivity. Similarly, irrigation regimes based on evapotranspiration were found to be more efficient in retaining and using NFF and also resulted in less TN being lost to leaching.  The findings of my study and those of Kuchta (2012), indicate that excessive drainage through the root zone in the coarse soils above the Abbotsford Sumas aquifer can result in increased NO3--TN transport from the active soil zone to the aquifer.  However, transport of NFF between April 2011 and January 2012 was minimal and a longer-term study is needed to determine the quantity of one year’s NFF ending up in leachate.  Additional samples collected during this study, but not analyzed, may provide further insight.  Nitrogen management affected nitrogen use efficiency.  Overall, fertigation proved to be a more efficient management method than all other treatments as it improved plant acquisition of NFF, resulted in reduced leaching of NO3--TN and improved plant N uptake while harvest yields were comparable to that of all other treatments.  In general, crops receiving low N inputs produced harvest yields comparable to those with higher N inputs while high N inputs resulted in greater NFF being removed from the system by berry harvest and pruning. The higher losses of NO3- -NFF to leaching observed under fertigation early on were due to the  80 type of fertilizer used, Ca(NO3)2. It is likely that if urea had been used instead of Ca(NO3)2, fertigation may also have had the least NO3- -NFF losses. In 2013, the fertigation treatment was switched from Ca(NO3)2 to urea in the experiment used in this study.  An additional fate of NFF study could be conducted to assess whether fertigation NO3- -NFF losses may be reduced with dissolved urea.    Tracing the location of fertilizer nitrogen in the plant system provided insight on the timing of fertilizer application.  Primocane leaf NFF concentrations between April and July indicate that the primocane component may be the primary user of NFF early on. Peak plant uptake of NFF occurred in July 2011 with the maximum NFF held in the floricane component at that time. By August, NFF content in both the floricane and primocane components had decreased and the primocane component had relocated NFF from the leaves to the cane.  Current practice of applying a split application in April and again in May may be poorly timed, thus exceeding plant requirements (as indicated by leachate losses).  Furthermore, primocanes may be taking up NFF preferentially early on so application timing may be better matched to primocane sprouting rather than floricane leaf budding, which happens earlier.  More substantial research is needed to determine NFF held in the entire plant between the first fertilizer application and July.  Rather than calculating leaf concentration I recommend that kg N ha-1 of NFF held in the individual plant components be determined on a weekly basis.  This study was unable to account for all the nitrogen applied, with an unaccounted for loss of 50 % of the applied fertilizer observed between July and August, 2011.  More research on possible NFF loss mechanisms in Fraser Valley raspberry production is needed. Some of these losses may have been a result of early leaf senescence in floricanes suggesting the field sampling strategy should be improved.  A greater number of sampling dates are suggested, particularly between fertilizer applications and July, one mid growing season sampling, and an additional autumn sampling. Mechanisms of N loss to monitor in the future are: ammonia and N2O/NOx emissions from soil as a result of ammonia volatilization, nitrification, and denitrification and; ammonia emission from the plant component.    81 Bibliography  Banados, M. P., B.C. Strik, D.R. Bryla and T. Righetti. 2012. Response of highbush blueberry to nitrogen fertilizer during field establishment, I: accumulation and allocation of fertilizer nitrogen and biomass. HortScience,47(5), 648-655.  British Columbia Ministry of Agriculture and Lands. 2009. Berry Production Guide 2009-2010. Beneficial management practices for berry growers in British Columbia, Lower Mainland Horticulture Improvement Association, Victoria, BC.  British Columbia Ministry of Agriculture 2012. BC Agriculture: Interactive production guides. “2012 Berry Production Guide: Raspberries.” http://productionguide.agrifoodbc.ca/guides/14 (accessed January 2015).  Boll J., Selker, J.S., Nijssen B. M., Steenhuis T. S., van Winkle J., E. Jolles. 1991. Water quality sampling under preferential flow conditions. ASCE, New York, NY, (USA). 290-298  Boll J., Steenhuis, T.S,  and J. S. Selker. 1992. Fiberglass wicks for sampling of water and solutes in the vadose zone. Soil Science Society of America Journal. 56 (3): 701-707  Brandi-Dohrn, F.M., Hess, M., Selker, J.S. and R.P. Dick. 1996. Field evaluation of passive capillary samplers. Soil Science Society of America Journal, 60 (6): 1705-1713    Burt T.P., Howden, N.J.K., Worrall, F and M.J. Whelan. 2008. Importance of long-term monitoring for detecting environmental change: Lessons from a lowland river in south east England. Biogeosciences 5 (6):1529–1535.   Chesnaux, R. and D.M. Allen. 2008. Stimulating nitrate leaching profiles in a highly permeable vadose zone. Environmental Modeling and Assessment 13 (4): 527-539.   Chesnaux, R., Allen, D.M., and G. Graham. 2007. Assessment of the impact of nutrient management practices on nitrate contamination in the Abbotsford-Sumas aquifer. Environmental Science & Technology 41 (21): 7229-7234.  Dale, A. 1989. Productivity in red raspberries. Horticulture Review 11:185-267.  Dean, D. M., Zebarth, B. J., Kowalenko C. G., Paul J. W. and K. Chipperfield. 2000. Poultry manure effects on soil nitrogen processes and nitrogen accumulation in red raspberry.   82 Canadian Journal of Plant Science 80 (4): 849-860  Denise Neilsen personal communication. (2015) Nitrification and N2O losses in the spring and summer.  Drew, M.C. and E.J. Sisworo  1977. Early effects of flooding on nitrogen deficiency and leaf chlorosis in barley. New Phytologist 79 (3): 567-571  Environment Canada. 2011. National Climate Data and Information Archive. http://www.climate.weatheroffice.gc.ca/Welcome_e.html. Abbotsford A. ID 1100030. (Accessed February 2015) Environment Canada. 2012. National Climate Data and Information Archive. http://www.climate.weatheroffice.gc.ca/Welcome_e.html. Abbotsford A. ID 1100030. (Accessed February 2015)  Freney, J.R. 1997. Emission of nitrous oxide from soils used for agriculture. Nutrient Cycling in Agroecosystems. 49(1-3): 1-6  Galloway, J. N., Dentener, F.J., Capone, D.G., Boyer, E.W., Howarth, R.W., Seitzinger, S.P., Asner, G.P., Cleveland, C.C., Green, P.A. Holland, E.A., Karl, D.M., Michaels, A.F., Porter, J.H., Townsend A.R. and C.J. Vöosmarty. 2004. Nitrogen cycles: past, present, and future. Biogeochemistry 70 (2): 153-226.  Gbur, E. E., Stroup, W. W., McCarter, K. S., Durham S., Young L. J., Christman, M., West, M. and M. Kramer. 2012. Analysis of Generalized Linear Mixed Models in the Agricultural and Natural Resources Sciences.  American Society of Agronomy. Madison, Wisconsin.  Gurovich, Luis A. 2005. A model to define fertigation strategies for raspberries, integrating soil water and nutrient availability to cropping objectives. In IX International Rubus and Ribes Symposium 777, pp. 411-422. 2005.  Hart J., Strik B. and H. Rempel. 2006. Caneberries. Nutrient Management Guide, published by Oregon State University and the US Department of Agriculture  Holtan-Hartwig, L., and O.C. Bockman. 1994. Ammonia exchange between crops and air. Norwegian Journal of Agricultural Science supplement 14: 52-61.  Hughes-Games, G., and B. Zebarth 1999. Soil Factsheet. Nitrogen recommendation for raspberries for south coastal British Columbia., Ministry of Agriculture and Food, Abbotsford, BC.   83 Husted, S. and J.K. Schjoerring. 1996. Ammonia flux between oilseed rape plants and the atmosphere in response to changes in leaf temperature, light intensity, and air humidity (interactions with leaf conductance and apoplastic NH4+ and H+ concentrations). Plant Physiology 112 (1): 67-74.  Jassal, R. S., Black, T.A., Chen, B., Roy, R., Nesic, Z., Spittlehouse, D.L. and J.A. Trofymow. 2008. N2O emissions and carbon sequestration in a nitrogen fertilized Douglas fir stand. Journal of Geophysical Research: Biogeosciences 113(G4): 1-10  Jeffries, M., Hughes-Games, G., Sweeney, M. and C. Mouritzen. 2008. Sustainable nitrogen management in British Columbia raspberry crops, in: P. Banados and A. Dale (Eds.), Proceedings of the IXth International Rubus and Ribes Symposium. Acta Hort. 777, ISHS 2008. pp. 435-438.  Knutson, H. and J. S. Selker. 1994. Unsaturated hydraulic conductivities of fiberglass wicks and designing capillary wick pore-water samplers. Soil Science Society of America Journal 58 (3): 721-729.   Knutson, H. and J. S. Selker. 1996. Fiberglass wick effects to solute travel time and dispersion for application in monitoring vadose zone chemical transport. Soil Science Society of America Journal 60: 420-424.  Kowalenko, C. G. 1989. The fate of applied nitrogen in a Fraser Valley soil using 15N in field microplots. Canadian Journal of Soil Science 69 (4): 825-833.  Kowalenko, C.G. 1994. Growing season dry matter and macroelement accumulations in Willamette red raspberry and related soil-extractable macroelement measurements. Canadian Journal of Plant Science 74(3): 565-571.  Kowalenko, C. G., Schmidt, O., and G. Hughes-Games. 2007. "Fraser Valley soil nutrient study 2005." A survey of the nitrogen, phosphorus and potassium contents of Lower Fraser Valley agricultural soils in relation to environmental and agronomic concerns. Report to British Columbia Agricultural Council, Victoria, BC  February 2007 http://www.agf.gov.bc.ca/resmgmt/EnviroFarmPlanning/FV_SoilNutrientStudy/_FVSNS-CombinedReport_Feb28_2007_for_Release.pdf (Accessed July 2014)  Kuchta, S. 2012. The effect of nitrogen, water and alley management strategies on nitrate and water loss from the root zone in perennial red raspberry. M.Sc. Thesis. University of British Columbia, Okanagan. In cIRcle [database on-line]; available from https://circle.ubc.ca/bitstream/handle/2429/43534/ubc_2013_spring_kuchta_shawn.pdf?sequence=1 (accessed November 2012)  84  Luttmerding, H. A. Soils of the Langley-Vancouver map area. Province of British Columbia, Ministry of Environment, 1984.  Malik, H., Archbold, D.D. and C. T. MacKown. 1991. Nitrogen partitioning by ‘Chester Thornless' Blackberry in Pot Culture. HortScience 26 (12): 1492-1494.  Mitchell, R.J., Babcock, R.S., Gelinas S., Nanus L. and D.E. Stasney. 2003. Nitrate distributions and source identification in the Abbotsford-Sumas Aquifer, northwestern Washington State. Journal of Environmental Quality 32 (3): 789-800. Mohadjer, P., Strik, B.C., Zebarth, B.J. and T.L Righetti. 2001. Nitrogen uptake, partitioning and remobilization in ‘Kotata’ blackberry in alternative-year production. Journal of Horticultural Science & Biotechnology 76 (6): 700-708  Myung, I.J. 2003. Tutorial on maximum likelihood estimation. Journal of Mathematical Psychology 47 (1): 90-100.  Naraguma, J., Clark, J.R., Norman, R.J. and R. W. McNew. 1999. Nitrogen uptake and allocation by field grown ‘Arapaho’ thornless blackberry. Journal of Plant Nutrition 22 (4-5): 753-768.  Neilsen, D., and G.H. Neilsen. 2002. Efficient Use of Nitrogen and Water in High-density Apple Orchards. HortTechnology 12:19-25.  Neilsen, D., Neilsen G.H., Gregory D., Forge T., and B. Zebarth. 2008. Drainage losses of water, N and P from micro-irrigation systems in a young, high density apple planting. , in: I. Goodwin and M. G. O’Connell (Eds.), Proceedings of the Vth International Symposium on Irrigation of Horticultural Crops. Acta Hort 792, ISHS 2008. pp. 482-490.  Neilsen, G.H., Neilsen, D. and L.C. Hambert. 2009. Nitrogen fertigation concentration and timing of application affect nutrition, yield, firmness, and color of apples grown at high density. HortScience 44 (5): 1425-1431  Neilsen, D., Millard, P., Herbert, L.C., Neilsen, G.H., Hogue, E.J., Parchomchuk, P. and B.J. Zebarth. 2001. Remobilization and uptake of N by newly planted apple (Malus domestica) trees in response to irrigation method and timing of N application. Tree Physiology 21 (8):513- 521.  85  Novoa, R. and R.S. Loomis. 1981. Nitrogen and Plant Production. Plant and Soil 58 (1-3): 177-204  Portela, S.I., Andriulo, A.E., Sasal, M.C., Mary, B. and E.G. Jobbágy. 2006. Fertilizer vs. organic matter contributions to nitrogen leaching in cropping systems of the Pampas: 15N application in field lysimeters. Plant and Soil 289 (1-2): 265-277.  Rempel, H. G., Stik B. C. and T. L. Righetti. 2004. Uptake, partitioning, and storage of fertilizer nitrogen in red raspberry as affected by rate and timing of application. Journal of the American Society for Horticultural Science, 129(3): 439-448    Schepers, J.S. and W.Raun, eds. Nitrogen in agricultural systems. Vol. 49. ASA-CSSA-SSSA, 2008.  Sebilo, M, Mayer, B. Nicolardot, B., Pinay, G. and A. Mariotti. 2004. Long-term fate of nitrate fertilizer in agricultural soils. Proceedings of the National Academy of Sciences 110(45): 18185–18189.  Schjoerring, J. K., Husted S. and M. Mattsson. 1998. Physiological parameters controlling plant–atmosphere ammonia exchange. Atmospheric Environment 32(3): 491-498.  Spalding, R.F., Watts, D.G, Schepers, J.S., Burbach, M.E., Exner, M.E., Poreda R.J. and G.E. Martin. 2001. Controlling nitrate leaching in irrigated agriculture. Journal of Environmental Quality 30 (4):1184-1194.  Statistics Canada. 2011 Table 2. Estimate of fruit area in Canada, by province, 2011. http://www.statcan.gc.ca/pub/22-003-x/2011001/t030-eng.htm (Accessed: June 17, 2012)  Strik, B.C. 2008. A review of nitrogen nutrition of Rubus, in: P. Banados and A. Dale (Eds.), Proceedings of the IXth International Rubus and Ribes Symposium. Acta Hort. 777, ISHS 2008. pp. 403-410.   86 Strik, B., Righetti, T. and H. Rempel. 2006. Black plastic mulch improved the uptake of 15Nitrogen from inorganic fertilizer and organic prunings in summer-bearing red raspberry. HortScience 41 (1): 272-274  SAS Support. 2013. Chapter 1: A setting for mixed models applications: Randomized block designs. http://support.sas.com/publishing/pubcat/chaps/55235.pdf (Accessed September 2011)  Tilman, D., Cassman, K. G., Matson, P. A., Naylor, R. and S. Polasky. 2002. Agricultural sustainability and intensive production practices. Nature 418 (6898): 671-677.  UCLA Statistical Consulting Group. 2013. A brief comparison of PROC GLM and PROC MIXED of SAS. http://statistics.ats.ucla.edu/stat/sas/library/GLMvsMIXED_os.htm  (Accessed September 2011)   Van der Gulik, T.W. 1999. BC Trickle Irrigation Manual BC Ministry of Agriculture and Food Resource Management Branch and Irrigation Industry Association of British Columbia, Abbotsford, BC.   Wassenaar, L.I. 1995. Evaluation of the origin and fate of nitrate in the Abbotsford Aquifer using the isotopes of 15N and 18O in NO3. Applied Geochemistry 10 (4): 391-405.  Wassenaar, L.I., Hendry, M.J. and N. Harrington. 2006. Decadal geochemical and isotopic trends for nitrate in a transboundary aquifer and implications for agricultural beneficial management practices. Environmental Science & Technology 40 (15): 4626–4632  Yang, R. C. 2009. Mixed model analysis of agricultural experiments: When some effects are random. (CSA Statistics Symposium). http://www.agronomycanada.com/09stats_Yang.pdf  Yarandi, H. 2003. Comparison of PROC MIXED and PROC GLM  for analysis of repeated measures data. http://analytics.ncsu.edu/sesug/2003/SD06-Yarandi.pdf (Accessed September 2012)  Yasmin, K., Cadisch G. and E.M. Baggs. 2006. Comparing 15N-labelling techniques for enriching above- and below-ground components of the plant-soil system. Soil Biology & Biochemistry 38 (2): 397-400  Zebarth, B. J., Dean, D. M., Kowalenko, C. G., Paul, J. W. and K. Chipperfield. 2002. Spatial and temporal variation in soil organic N concentration, and soil test P and K, in red raspberry fields and implications for soil strategies.  Canadian Journal of Soil Science  87 82 (2): 355-364.  Zebarth, B.J., Hii, B., Liebscher, H., Chipperfield, K., Paul, J.W., Grove, G. and S. Y. Szeto. 1998. Agricultural land use practices and contamination in the Abbotsford Aquifer, British Columbia, Canada.  Agriculture, Ecosystems and Environment, 69 (2): 99-112  Zebarth, B.J., Kowalenko C.G. and B. Harding. 2007. Soil inorganic nitrogen content and indices of red raspberry yield, vigor, and nitrogen status as affected by rate and source of nitrogen fertilizer. Communications in Soil Science and Plant Analysis 38 (5-6): 637-660.  Zebarth, B.J., Kowalenko, C.G., Harding, B., and D. Chapple 1997. Development of a test-based nitrogen recommendation system for raspberries to reduce nitrate leaching to the Abbotsford Aquifer. Technical report 97-01., Pacific Agri-Food Research Centre, Summerland, BC.      88 Appendices Appendix A: Equations  The following equations were used to obtain the parameters discussed in Chapter 2, 3, and 4:  Nitrogen from fertilizer concentration [NFF] in mg N L-1 sample (leachate) and µg N mg-1 sample (solids) was calculated for each interval (i) in all samples using:   !"" !"#$%&,! = !!" !"#$%&,!!!" !!" !"#$%&!!% *[UCTN]i    (Equation 1)  Where,  Sample = Leachate, soil, roots, primocanes, floricanes, laterals, primocane leaves, floricane leaves, fruiting structures and harvest fruit 15NSample,i = 15N atom % reported in this study’s samples collected over the ith sampling interval BG15NSample = Naturally occurring background 15N atom % (0.3663 %) UCTN,i != Total N sampled over the ith sampling interval. In leachate (mg N from NO3 + NO2 L-1 sample) in solid samples (µg 15N mg-1 sample)  Total nitrogen (TN) is calculated per experimental sub-plot.  In leachate, TNLeachate,i (kg N sub-plot-1) was calculated for each interval using:   TNLEACHATE, i = ! [!"#$]!"#$!!"#,! ∗ !!!"!"!! " ∗ !!!"#$!!"#,! ∗ ( !"#$!"#$!"#$!"#!$)  (Equation 2)  Where, UCTNLeachate,i != Total N leached over the ith sampling interval (mg N from NO3 + NO2 L-1 sample) VLeachate,i = Volume of leachate captured by the PCAPS over the ith sampling interval (L) AreaPLOT = Experimental sub-plot area to which enriched fertilizer was applied and samples were taken from (4.8 m2)  89 AreaPCAPS = Catchment area of the row PCAPS  (0.36 m2)  Nitrogen from fertilizer (NFF) is calculated per experimental sub-plot.  In leachate, NFFLeachate,i (kg N sub-plot-1) was calculated for each interval using:   NFFLEACHATE, i = [!""]!"#$!!"#,! !∗ !!!"!"!! " ∗ !!!"#$!!"#,! ∗ ( !"#$!"#$!"#$!"#!$)    (Equation 3)  Where, [!""]!"#$!!"#,! = The concentration of NFF leached over the ith sampling interval (mg N from NO3 + NO2 L-1 sample) VLeachate,i = Volume of leachate captured by the PCAPS over the ith sampling interval (L) AreaPLOT = Experimental sub-plot area to which enriched fertilizer was applied and samples were taken from (4.8 m2) AreaPCAPS = Catchment area of the row PCAPS (0.36 m2)  Total nitrogen (TN) is calculated per experimental sub-plot.  For both soil and roots, TNSoil or Root,i (kg N sub-plot-1) was calculated for each interval using:   TNSoil or root, i = !!" ∗ !!"#$ ∗ [!"#$]!"#$!!"!!""#,! ∗ !"!! "!!"!!!"!!!"!!"!! !!(Equation 4)  Where, BD = Bulk density of soil in experimental field (1270 kg m-3 or 1.27 g m-3) VPLOT = Volume of the plot (1.44 m3) UCTNSoil or Root,i != Total N collected over the ith sampling interval (µg N mg-1 sample)  Nitrogen from fertilizer (NFF) is calculated per experimental sub-plot.  For both soil and roots, NFFSoil or Root,i (kg N sub-plot-1) was calculated for each interval using:   NFFSoil or Root, i = !!" ∗ !!"#$ ∗ [!""]!"#$!!"!!""#,! ∗ !"!! "!!"!!!"!!!"!!"!! !!(Equation 5)  90  Where, BD = Bulk density of soil in experimental field (1270 kg m-3  or1.27 g m-3)  VPLOT = Volume of the sub-plot (1.44 m3) [!""]!"#$!!"!!""#,! = The concentration of NFF in the soil or root samples collected over the ith sampling interval (µg N mg-1 sample)   Total nitrogen (TN) is calculated per experimental sub-plot.  In plant components, TNPlant,i (kg N sub-plot-1) was calculated for each interval using:   TNPlant, i = !!"!"#$%,!!!!"#$% ∗ !!"#$ ∗ [!"#$]!"#$%,! ∗ !"!! "!!"!!!"!!!"!!"!! !  (Equation 6)  Where, Plant = Primocanes, floricanes, primocane leaves, floricane leaves, laterals, fruiting structures and harvest fruit DWPlant,i  = Dry weight of plant component per 6 cane composite sample collected over the ith sampling interval (kg) CPLOT = Number of primocanes or floricanes per plot (depending on plant component type). This value was unique for each plot.  [UCTN]Plant,i != Total N collected over the ith sampling interval (µg N mg-1 sample)  Nitrogen from fertilizer (NFF) is calculated per experimental sub-plot.  In plant components, NFFPlant,i (kg N sub-plot-1) was calculated for each interval using:   NFFPlant, i = ! !"" !"#$%,! ∗ !"!"#$%,!!!!"#$! ∗ !!"#$ ∗ !"!! "!!"!!!"!!!"!!"!! !(Equation 7)  Where, Plant = Primocanes, floricanes, primocane leaves, floricane leaves, laterals, fruiting structures and harvest fruit  91 [!""]!"#$%,! = The concentration of NFF in plant samples collected over the ith sampling interval (µg N mg-1 sample) DWPlant,i  = Dry weight of plant component per 6 cane composite sample collected over the ith sampling interval (kg) CPLOT = Number of primocanes or floricanes per plot (depending on plant component type). This value was unique for each plot.  TNFIELD and NFFFIELD is calculated per field hectare (kg N ha-1) using:   TN!"#$%!or!NFF!"#$% = TN!"#$ !or!NFF!"#$ ! ∗ 10,000! 2ℎ!−112! 2   (Equation 8) Where, TN!"#$ =  TN per experimental sub-plot (kg N plot-1) NFF!"#$ !!= NFF per experimental sub-plot (kg N plot-1)  The amounts of TN and NFF per plot were converted to a total amount per hectare of field basis (kg N ha-1).  Each 4.8 m2 sub-plot requires 12 m2 field space due to the alleys which require take up 7.2 m2 of space. The fertilizer application rates of 50 and 100 kg N ha-1 were applied on a whole field basis.  Therefore the conversion factor shown above can be used to calculate N applied to the crop on a per hectare basis.  Nitrogen leached or removed over the growing year (NCumulative) was calculated for TN, NFF, %FOA and %FOA using:   !!"#"$%&'() = ! !!!!   (Equation 9)  Where, NCumulative = The N leached or removed between April 2011 to December 2011 (kg N ha-1) Ni = The N leached or removed over the ith sample collection interval (kg N ha-1) n = The number of sampling intervals between April 2011 to November 2012.   92 The percent NFF of applied (%FOA) was calculated for all sample types using:   %FOA = ! NFF!"#$%!!""#$%!&$'( ∗ 100%  (Equation 10)  Where, NFFFIELD= N calculated per field hectare (kg N ha-1) RAPPLICATION  = The treatment application rate per field hectare (kg N ha-1)  The percent NFF of TN (%FOT) was calculated for all sample types using:   %FOT = ! !""!"#$%!"!"#$% ∗ 100%  (Equation 11)  Where, NFFFIELD= The concentration of NFF in samples collected over the ith sampling interval (kg N ha-1) TNFIELD= The concentration of TN in samples collected over the ith sampling interval (kg N ha-1)     93 Appendix B: Treatment Comparisons  ENTIRE SYSTEM BALANCE B.1 Entire System: Total Nitrogen                                                    Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                     3      45       1.05    0.3778                                          samptime           3      45       1.55    0.2152                                          samptime*trt       9      45       0.20    0.9925                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 1       3       45       0.03    0.9931                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 1    5      6       33.4034      273.11       45       0.12      0.9032                           samptime 1    5      7      -26.2970      273.11       45      -0.10      0.9237                           samptime 1    5      8      -42.5074      273.11       45      -0.16      0.8770                           samptime 1    6      7      -59.7003      273.11       45      -0.22      0.8280                           samptime 1    6      8      -75.9107      273.11       45      -0.28      0.7823                           samptime 1    7      8      -16.2104      273.11       45      -0.06      0.9529                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 2       3       45       0.29    0.8307                                         Simple Differences of samptime*trt Least Squares Means  94                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 2    5      6        106.15      273.11       45       0.39      0.6993                           samptime 2    5      7      -69.2165      273.11       45      -0.25      0.8011                           samptime 2    5      8        163.34      273.11       45       0.60      0.5528                           samptime 2    6      7       -175.37      273.11       45      -0.64      0.5241                           samptime 2    6      8       57.1837      273.11       45       0.21      0.8351                           samptime 2    7      8        232.55      273.11       45       0.85      0.3990                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 3       3       45       0.84    0.4775                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 3    5      6        140.98      273.11       45       0.52      0.6083                           samptime 3    5      7       -224.76      273.11       45      -0.82      0.4149                           samptime 3    5      8        159.28      273.11       45       0.58      0.5627                           samptime 3    6      7       -365.74      273.11       45      -1.34      0.1872                           samptime 3    6      8       18.3013      273.11       45       0.07      0.9469                           samptime 3    7      8        384.04      273.11       45       1.41      0.1665                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 4       3       45       0.50    0.6835                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 4    5      6        254.85      273.11       45       0.93      0.3557                           samptime 4    5      7       90.5091      273.11       45       0.33      0.7419  95                           samptime 4    5      8        288.99      273.11       45       1.06      0.2956                           samptime 4    6      7       -164.34      273.11       45      -0.60      0.5504                           samptime 4    6      8       34.1370      273.11       45       0.12      0.9011                           samptime 4    7      8        198.48      273.11       45       0.73      0.4712  B.2 Entire System: Fertilizer Nitrogen                                                     Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                       3      45      25.28     <.0001                                          samptime           3      45      22.38    <.0001                                          samptime*trt       9      45       2.32    0.0308                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 1       3       45      20.57    <.0001                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 1    5      6       45.1964      8.6130       45       5.25      <.0001                           samptime 1    5      7       -6.8184      8.6130       45      -0.79      0.4327                           samptime 1    5      8       43.1321      8.6130       45       5.01      <.0001                           samptime 1    6      7      -52.0148      8.6130       45      -6.04      <.0001                           samptime 1    6      8       -2.0644      8.6130       45      -0.24      0.8117                           samptime 1    7      8       49.9504      8.6130       45       5.80      <.0001                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 2       3       45       3.86    0.0153                                         Simple Differences of samptime*trt Least Squares Means  96                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 2    5      6       19.2176      8.6130       45       2.23      0.0307                           samptime 2    5      7       -2.3093      8.6130       45      -0.27      0.7898                           samptime 2    5      8       19.8023      8.6130       45       2.30      0.0262                           samptime 2    6      7      -21.5269      8.6130       45      -2.50      0.0162                           samptime 2    6      8        0.5848      8.6130       45       0.07      0.9462                           samptime 2    7      8       22.1117      8.6130       45       2.57      0.0136                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 3       3       45       1.85    0.1525                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 3    5      6        6.1510      8.6130       45       0.71      0.4788                           samptime 3    5      7      -11.6349      8.6130       45      -1.35      0.1835                           samptime 3    5      8        5.6739      8.6130       45       0.66      0.5134                           samptime 3    6      7      -17.7860      8.6130       45      -2.07      0.0447                           samptime 3    6      8       -0.4771      8.6130       45      -0.06      0.9561                           samptime 3    7      8       17.3088      8.6130       45       2.01      0.0505                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 4       3       45       5.95    0.0016                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 4    5      6       16.7951      8.6130       45       1.95      0.0574                           samptime 4    5      7      -16.5966      8.6130       45      -1.93      0.0603  97                           samptime 4    5      8       11.9032      8.6130       45       1.38      0.1738                           samptime 4    6      7      -33.3917      8.6130       45      -3.88      0.0003                           samptime 4    6      8       -4.8919      8.6130       45      -0.57      0.5729                           samptime 4    7      8       28.4998      8.6130       45       3.31      0.0018   B.3 Entire System: Percent Fertilizer Nitrogen of Applied                                                     Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                3      45       1.60    0.2032                                          samptime           3      45      23.37    <.0001                                          samptime*trt       9      45       0.76    0.6542                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 1       3       45       0.53    0.6669                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 1    5      6        6.0558     10.4765       45       0.58      0.5661                           samptime 1    5      7       -6.8184     10.4765       45      -0.65      0.5185                           samptime 1    5      8        1.9271     10.4765       45       0.18      0.8549                           samptime 1    6      7      -12.8741     10.4765       45      -1.23      0.2255                           samptime 1    6      8       -4.1287     10.4765       45      -0.39      0.6954                           samptime 1    7      8        8.7454     10.4765       45       0.83      0.4083                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 2       3       45       0.08    0.9680   98                                        Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 2    5      6       -4.9954     10.4765       45      -0.48      0.6358                           samptime 2    5      7       -2.3093     10.4765       45      -0.22      0.8265                           samptime 2    5      8       -3.8283     10.4765       45      -0.37      0.7165                           samptime 2    6      7        2.6861     10.4765       45       0.26      0.7988                           samptime 2    6      8        1.1671     10.4765       45       0.11      0.9118                           samptime 2    7      8       -1.5190     10.4765       45      -0.14      0.8854                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 3       3       45       2.04    0.1212                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 3    5      6      -21.7182     10.4765       45      -2.07      0.0439                           samptime 3    5      7      -11.6349     10.4765       45      -1.11      0.2727                           samptime 3    5      8      -22.6750     10.4765       45      -2.16      0.0358                           samptime 3    6      7       10.0833     10.4765       45       0.96      0.3410                           samptime 3    6      8       -0.9568     10.4765       45      -0.09      0.9276                           samptime 3    7      8      -11.0401     10.4765       45      -1.05      0.2976                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 4       3       45       1.22    0.3136                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 4    5      6       -7.2993     10.4765       45      -0.70      0.4896  99                           samptime 4    5      7      -16.5966     10.4765       45      -1.58      0.1202                           samptime 4    5      8      -17.0856     10.4765       45      -1.63      0.1099                           samptime 4    6      7       -9.2973     10.4765       45      -0.89      0.3796                           samptime 4    6      8       -9.7863     10.4765       45      -0.93      0.3552                           samptime 4    7      8       -0.4890     10.4765       45      -0.05      0.9630  B.4 Entire System: Percent Fertilizer Nitrogen of Total Nitrogen                                                     Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                      3      45      20.25    <.0001                                          samptime           3      45      19.49    <.0001                                          samptime*trt       9      45       2.40    0.0256                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 1       3       45      18.87    <.0001                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 1    5      6        1.3541      0.2684       45       5.05      <.0001                           samptime 1    5      7       -0.1624      0.2684       45      -0.61      0.5481                           samptime 1    5      8        1.3298      0.2684       45       4.95      <.0001                           samptime 1    6      7       -1.5166      0.2684       45      -5.65      <.0001                           samptime 1    6      8      -0.02434      0.2684       45      -0.09      0.9281                           samptime 1    7      8        1.4922      0.2684       45       5.56      <.0001                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 2       3       45       3.17    0.0333   100                                        Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 2    5      6        0.5656      0.2684       45       2.11      0.0407                           samptime 2    5      7      -0.02517      0.2684       45      -0.09      0.9257                           samptime 2    5      8        0.5786      0.2684       45       2.16      0.0365                           samptime 2    6      7       -0.5907      0.2684       45      -2.20      0.0329                           samptime 2    6      8       0.01306      0.2684       45       0.05      0.9614                           samptime 2    7      8        0.6038      0.2684       45       2.25      0.0294                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 3       3       45       1.21    0.3161                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 3    5      6        0.1650      0.2684       45       0.61      0.5417                           samptime 3    5      7       -0.2964      0.2684       45      -1.10      0.2752                           samptime 3    5      8        0.1261      0.2684       45       0.47      0.6407                           samptime 3    6      7       -0.4615      0.2684       45      -1.72      0.0924                           samptime 3    6      8      -0.03892      0.2684       45      -0.15      0.8853                           samptime 3    7      8        0.4226      0.2684       45       1.57      0.1224                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 4       3       45       4.20    0.0105                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 4    5      6        0.4292      0.2684       45       1.60      0.1168  101                           samptime 4    5      7       -0.4651      0.2684       45      -1.73      0.0900                           samptime 4    5      8        0.2558      0.2684       45       0.95      0.3457                           samptime 4    6      7       -0.8943      0.2684       45      -3.33      0.0017                           samptime 4    6      8       -0.1735      0.2684       45      -0.65      0.5213                           samptime 4    7      8        0.7208      0.2684       45       2.69      0.0101   B.5 Soil: Total Nitrogen                                                     Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                3      45       1.07    0.3726                                          samptime           3      45       1.33    0.2774                                          samptime*trt       9      45       0.18    0.9953                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 1       3       45       0.03    0.9922                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 1    5      6       48.1677      279.78       45       0.17      0.8641                           samptime 1    5      7      -22.1626      279.78       45      -0.08      0.9372                           samptime 1    5      8      -30.5974      279.78       45      -0.11      0.9134                           samptime 1    6      7      -70.3303      279.78       45      -0.25      0.8027                           samptime 1    6      8      -78.7651      279.78       45      -0.28      0.7796                           samptime 1    7      8       -8.4348      279.78       45      -0.03      0.9761                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 2       3       45       0.28    0.8394  102                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 2    5      6        116.62      279.78       45       0.42      0.6788                           samptime 2    5      7      -81.3927      279.78       45      -0.29      0.7725                           samptime 2    5      8        143.82      279.78       45       0.51      0.6097                           samptime 2    6      7       -198.02      279.78       45      -0.71      0.4827                           samptime 2    6      8       27.1975      279.78       45       0.10      0.9230                           samptime 2    7      8        225.21      279.78       45       0.80      0.4251                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 3       3       45       0.85    0.4738                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 3    5      6        154.03      279.78       45       0.55      0.5847                           samptime 3    5      7       -244.35      279.78       45      -0.87      0.3871                           samptime 3    5      8        129.55      279.78       45       0.46      0.6456                           samptime 3    6      7       -398.38      279.78       45      -1.42      0.1614                           samptime 3    6      8      -24.4779      279.78       45      -0.09      0.9307                           samptime 3    7      8        373.90      279.78       45       1.34      0.1881                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 4       3       45       0.44    0.7229                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|   103                           samptime 4    5      6        267.63      279.78       45       0.96      0.3439                           samptime 4    5      7       71.1611      279.78       45       0.25      0.8004                           samptime 4    5      8        247.95      279.78       45       0.89      0.3802                           samptime 4    6      7       -196.46      279.78       45      -0.70      0.4862                           samptime 4    6      8      -19.6728      279.78       45      -0.07      0.9443                           samptime 4    7      8        176.79      279.78       45       0.63      0.5307  B.6 Soil: Fertilizer Nitrogen                                                     Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                       3      45      20.69    <.0001                                          samptime           3      45      12.89    <.0001                                          samptime*trt       9      45       2.82    0.0102                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 1       3       45      20.33    <.0001                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 1    5      6       39.2313      8.4003       45       4.67      <.0001                           samptime 1    5      7       -5.9001      8.4003       45      -0.70      0.4861                           samptime 1    5      8       46.6657      8.4003       45       5.56      <.0001                           samptime 1    6      7      -45.1314      8.4003       45      -5.37      <.0001                           samptime 1    6      8        7.4345      8.4003       45       0.89      0.3808                           samptime 1    7      8       52.5658      8.4003       45       6.26      <.0001                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 2       3       45       2.53    0.0692  104                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 2    5      6       11.5741      8.4003       45       1.38      0.1751                           samptime 2    5      7      -0.05934      8.4003       45      -0.01      0.9944                           samptime 2    5      8       19.2408      8.4003       45       2.29      0.0267                           samptime 2    6      7      -11.6335      8.4003       45      -1.38      0.1729                           samptime 2    6      8        7.6667      8.4003       45       0.91      0.3663                           samptime 2    7      8       19.3002      8.4003       45       2.30      0.0263                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 3       3       45       1.50    0.2268                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 3    5      6        1.4449      8.4003       45       0.17      0.8642                           samptime 3    5      7      -11.9878      8.4003       45      -1.43      0.1605                           samptime 3    5      8        4.6727      8.4003       45       0.56      0.5808                           samptime 3    6      7      -13.4326      8.4003       45      -1.60      0.1168                           samptime 3    6      8        3.2278      8.4003       45       0.38      0.7026                           samptime 3    7      8       16.6605      8.4003       45       1.98      0.0535                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 4       3       45       4.80    0.0055                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|   105                           samptime 4    5      6       10.7696      8.4003       45       1.28      0.2064                           samptime 4    5      7      -17.1438      8.4003       45      -2.04      0.0472                           samptime 4    5      8       10.0709      8.4003       45       1.20      0.2369                           samptime 4    6      7      -27.9134      8.4003       45      -3.32      0.0018                           samptime 4    6      8       -0.6987      8.4003       45      -0.08      0.9341                           samptime 4    7      8       27.2147      8.4003       45       3.24      0.0023  B.7 Soil: Percent Fertilizer Nitrogen of Applied                                                     Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                3      45       4.30    0.0094                                          samptime           3      45      11.41    <.0001                                          samptime*trt       9      45       1.76    0.1036                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 1       3       45       6.35    0.0011                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 1    5      6       19.2209     10.2685       45       1.87      0.0677                           samptime 1    5      7       -5.9001     10.2685       45      -0.57      0.5684                           samptime 1    5      8       34.0898     10.2685       45       3.32      0.0018                           samptime 1    6      7      -25.1210     10.2685       45      -2.45      0.0184                           samptime 1    6      8       14.8689     10.2685       45       1.45      0.1545                           samptime 1    7      8       39.9899     10.2685       45       3.89      0.0003                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 2       3       45       1.19    0.3241  106                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 2    5      6        0.7140     10.2685       45       0.07      0.9449                           samptime 2    5      7      -0.05934     10.2685       45      -0.01      0.9954                           samptime 2    5      8       16.0473     10.2685       45       1.56      0.1251                           samptime 2    6      7       -0.7733     10.2685       45      -0.08      0.9403                           samptime 2    6      8       15.3334     10.2685       45       1.49      0.1424                           samptime 2    7      8       16.1067     10.2685       45       1.57      0.1238                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 3       3       45       0.82    0.4884                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|                            samptime 3    5      6      -15.3347     10.2685       45      -1.49      0.1423                           samptime 3    5      7      -11.9878     10.2685       45      -1.17      0.2492                           samptime 3    5      8       -8.8791     10.2685       45      -0.86      0.3918                           samptime 3    6      7        3.3470     10.2685       45       0.33      0.7460                           samptime 3    6      8        6.4556     10.2685       45       0.63      0.5327                           samptime 3    7      8        3.1087     10.2685       45       0.30      0.7635                                                     F Test for samptime*trt Least                                                         Squares Means Slice                                                            Num      Den                                           Slice           DF       DF    F Value    Pr > F                                            samptime 4       3       45       1.21    0.3160                                         Simple Differences of samptime*trt Least Squares Means                                                                     Standard                           Slice         trt    trt    Estimate       Error       DF    t Value    Pr > |t|   107                           samptime 4    5      6       -1.2477     10.2685       45      -0.12      0.9038                           samptime 4    5      7      -17.1438     10.2685       45      -1.67      0.1020                           samptime 4    5      8       -2.6452     10.2685       45      -0.26      0.7979                           samptime 4    6      7      -15.8960     10.2685       45      -1.55      0.1286                           samptime 4    6      8       -1.3975     10.2685       45      -0.14      0.8924                           samptime 4    7      8       14.4986     10.2685       45       1.41      0.1648  B.8 Soil: Percent Fertilizer Nitrogen of Total Nitrogen                                                     Type 3 Tests of Fixed Effects                                                             Num     Den                                          Effect            DF      DF    F Value    Pr > F                                           trt                3