Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Phosphorus dynamics under different nutrient management regimes in the Sumas Prairie, British Columbia,… Oka, Gladys Azaria 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_oka_gladys.pdf [ 3.94MB ]
Metadata
JSON: 24-1.0166274.json
JSON-LD: 24-1.0166274-ld.json
RDF/XML (Pretty): 24-1.0166274-rdf.xml
RDF/JSON: 24-1.0166274-rdf.json
Turtle: 24-1.0166274-turtle.txt
N-Triples: 24-1.0166274-rdf-ntriples.txt
Original Record: 24-1.0166274-source.json
Full Text
24-1.0166274-fulltext.txt
Citation
24-1.0166274.ris

Full Text

   PHOSPHORUS DYNAMICS UNDER DIFFERENT NUTRIENT MANAGEMENT REGIMES IN THE SUMAS PRAIRIE, BRITISH COLUMBIA, CANADA  by  Gladys Azaria Oka   BSc., The University of British Columbia, 2012       A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE   in   The Faculty of Graduate and Postdoctoral Studies  (Soil Science)        THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)   May 2015   © Gladys Azaria Oka, 2015ii  Abstract  Phosphorus (P) is essential to life. Depletion of phosphate rock may occur within 50-400 years, yet P application to soils continues and has resulted in over-application of manure and fertilizer contributing to contamination of water systems.  The Sumas Prairie in British Columbia demonstrates the convergence of P deficiency and excess in alluvial soils with high organic matter (SOM) content and serpentinitic contributions. The objective of this study was to evaluate how nutrient management at four sites (i.e. dairy farm, turf farm, mixed manure and fertilizer, and uncultivated) influence P compounds, their solubility, and (de)sorption reactions occurring in surface soils and adjacent ditch sediments. Samples were collected at 0-20 cm, 20-50 cm, 50+ cm depths in soil at each site, and in ditch sediments. P was vertically mobile from the surface to 20-50 cm depth in soil, and accumulation in ditch sediments was evident.  P concentrations declined in the following order: mixed > dairy ~ uncultivated > turf. P fractionation indicated the following relative composition of total P (Aqua regia): 55-80% stable (acid ammonium oxalate), 12-46% labile (HCl), 2-5% Bray, 2-7% Mehlich, and 10-50% organic P (NaOH-EDTA). Mineral P sources were attributed to apatite, P adsorped on amorphous aluminum oxide as a stable source of P, and calcium phosphate precipitate as a labile source of P. 31P NMR analysis of organic extracts indicated that Pi,(orthophosphate and pyrophosphate fractions) accounted for 69-100% of extracted P, and Po  (predominantly phosphate monoesters) accounted for 0-31%. SOM was the most important factor determining P concentrations and sorption sites for the retention of P. Clay content and SOM were positively associated with increased degrees of P saturation.  P content in soils and ditch sediments were below maximum adsorption. Processes of adsorption and desorption in these soils are influenced by competition of Po and Pi forms particularly in the dairy, mixed, and uncultivated sites where SOM content is high. There is potential for P loss and subsequent environmental contamination from topsoil and ditch sediments at the mixed site according to DPSox, and from topsoil and ditch sediments at all sites according to M3-PSR I and M3-PSR II.   iii  Preface  This dissertation is original, unpublished, independent work by the author, G.A. Oka.   iv  Table of Contents Abstract .................................................................................................................................................. ii Preface ................................................................................................................................................... iii Table of Contents .................................................................................................................................. iv List of Tables ........................................................................................................................................ vii List of Figures ....................................................................................................................................... ix Nomenclature ........................................................................................................................................ xi Acknowledgements ............................................................................................................................. xiii Dedication ........................................................................................................................................... xiv 1. Introduction and overview .................................................................................................................. 1 1.1 Phosphorus in the environment: natural flows, human impacts, and food security .................... 2 1.2 Inorganic and organic phosphorus in the soil .............................................................................. 3 1.3 Phosphorus as a nutrient amendment .......................................................................................... 5 1.4 Selected soil factors contributing to phosphorus solubility ......................................................... 6 1.5 Environmental contamination of phosphorus .............................................................................. 8 1.6 The Sumas Prairie ..................................................................................................................... 10 1.7 Thesis objectives and overview ................................................................................................. 12 2. Evaluating immediate and long-term phosphorus pools and the influence of selected soil factors on phosphorus release................................................................................................................................ 13 2.1 Synopsis .................................................................................................................................... 13 2.2 Introduction ............................................................................................................................... 14 2.3 Methods ..................................................................................................................................... 16 2.3.1 Site description and field sampling ..................................................................................... 16 2.3.2 pH, electrical conductivity (EC) and soil organic matter (SOM) ........................................ 21 2.3.3 Particle size distribution ...................................................................................................... 21 2.3.4 Chemical extractions for inorganic and organic phosphorus .............................................. 21 v  2.3.5 Calculations for degrees of phosphorus saturation (%) using the AAO and Mehlich-III extractions..................................................................................................................................... 23 2.3.6 Soil organic carbon and total nitrogen ................................................................................. 23 2.3.7 Statistical analysis ............................................................................................................... 24 2.4 Results ....................................................................................................................................... 24 2.4.1 Soil factors influencing P solubility: pH, soil organic matter, and clay content ................. 24 2.4.2 Accumulation of P and associated elements in field soil and ditch sediments .................... 27 2.4.3 Environmental P saturation as measured by the AAO and Mehlich-III extractions ........... 35 2.4.4 Carbon, nitrogen, and organic phosphorus concentrations and nutrient ratios .................... 38 2.5 Discussion ................................................................................................................................. 39 2.5.1 Status of soil phosphorus in dairy, turf, mixed, and uncultivated sites ............................... 39 2.5.2 Influence of selected soil factors on P accumulation and saturation ................................... 41 3. Characterizing inorganic and organic phosphorus sources, and evaluating surface properties relevant for phosphorus retention ......................................................................................................... 43 3.1 Synopsis .................................................................................................................................... 43 3.2 Introduction ............................................................................................................................... 43 3.3 Methods ..................................................................................................................................... 46 3.3.1 Sample collection ................................................................................................................ 46 3.3.2 Soil mineralogy: X-ray diffraction (XRD) .......................................................................... 46 3.3.3 Inorganic and organic phosphorus compounds: Solid and liquid nuclear magnetic resonance (NMR) ......................................................................................................................... 47 3.3.4 Zeta potential: zero point of charge ..................................................................................... 52 3.3.5 Phosphorus adsorption capacity: Linear, Langmuir and Freundlich models ....................... 53 3.4 Results ....................................................................................................................................... 55 3.4.1 Soil mineralogy and 31P-NMR inorganic compounds ......................................................... 55 3.4.2 31P-NMR compounds from organic soil extracts ................................................................. 59 3.4.3 Surface charge of soil .......................................................................................................... 65 3.4.4 Phosphorus adsorption capacity: Linear, Langmuir and Freundlich models ....................... 66 vi  3.5 Discussion ................................................................................................................................. 71 3.5.1 Inorganic and organic phosphorus compounds ................................................................... 71 3.5.2 Dynamics of phosphorus adsorption ................................................................................... 73 4. Conclusions ...................................................................................................................................... 75 Bibliography ......................................................................................................................................... 77   vii  List of Tables  Table 2.1 Description of study site characteristics. .............................................................................. 18 Table 2.2 Mean values of active and exchangeable pH in water and 0.01 M CaCl2 respectively, and SOM content at the study sites. ............................................................................................................ 25 Table 2.3 Mean values of the concentration and standard error of the mean (SEM) for total phosphorus and associated elements Ca, Mg, Fe, and Al from the AR extraction. .............................. 28 Table 2.4 Mean values of the concentration and standard error of the mean for stable phosphorus and associated elements Mg, Fe, and Al from the AAO extraction. ........................................................... 28 Table 2.5 Mean values of the concentration and standard error of the mean for labile phosphorus and associated elements Ca, Mg, Fe, and Al from the 0.1M HCl extraction. ............................................. 29 Table 2.6 Mean values of the concentration and standard error of the mean for available phosphorus and associated elements Ca, Mg, Fe, and Al from the Bray and Kurtz P-1 extraction. ....................... 29 Table 2.7 Mean values of the concentration and standard error of the mean for available phosphorus and associated elements Ca, Mg, Fe, and Al from the Mehlich-III extraction. .................................... 30 Table 2.8 Mean values of the concentration and standard error of the mean for organic phosphorus from Bowman and Moir’s (1993) NaOH and EDTA extraction. ......................................................... 30 Table 2.9 Summary of minimum and maximum concentrations for total, stable, labile, plant-available and organic P across the study sites...................................................................................................... 31 Table 2.10 Indices for environmental P saturation measured using the AAO (DPSox) and Mehlich-III (M3-PSR I and M3-PSR II) extractions. .............................................................................................. 36 Table 2.11 Concentrations and ratios of organic carbon, total nitrogen, and organic phosphorus at the study sites. ............................................................................................................................................ 38 Table 3.1 Mean concentrations and standard error of the mean (when applicable), along with proportions of phosphate, phosphate monoesters and other compounds, and pyrophosphate in organic P extracts of field soil and ditch sediments at the study sites. .............................................................. 63 viii  Table 3.2 Parameters of the linear, Langmuir, and Freundlich models for phosphorus adsorption in field soils and ditch sediments at the study sites. ................................................................................. 68 Table 3.3 Buffering capacity indices of field soils and ditch sediments at the study sites, including: maximum buffering capacity (MBC), buffering index (BI), phosphorus buffering capacity (PBC), and phosphorus sorption index (PSI). ......................................................................................................... 68    ix  List of Figures  Figure 2.1 Inorganic and organic P sources and pools in the soil (Stewart & Sharpley, 1987). .......... 14 Figure 2.2 The location of the Sumas Prairie in the Lower Fraser Valley (adapted from Boyle et al., 1997). .................................................................................................................................................... 19 Figure 2.3 The locations of the study sites within the Sumas Prairie relative to the Sumas River and Swift Creek Landslide in the delineation of the Sumas Watershed. ..................................................... 19 Figure 2.4 Conditions of the four study sites in the Sumas Prairie, depicting the varying degrees of ground cover, cultivation type, and nutrient management regimes. ..................................................... 20 Figure 2.5 The ditch sediments from the treatment sites, depicting location relative to the main field, water saturation, and degree of vegetative cover at the time of sampling. ........................................... 20 Figure 2.6 Change in mean SOM content with depth (0-50+ cm) in the soil profile at the study sites. .............................................................................................................................................................. 26 Figure 2.7 The proportions of sand, silt and clay, and corresponding texture class of: a) topsoil (0-20 cm) and b) soil at 50+ cm depth at the study sites; and c) ditch sediments (0-20 cm) at the treatment sites. ...................................................................................................................................................... 26 Figure 2.8 Mean concentrations of a) total P, b) stable P, c) labile P, d) available P (Bray), e) available P (Mehlich), and f) organic P at the study sites in field soil at depths of 0-50+ cm, and in surface (0-20 cm) ditch sediments. Error bars are standard error of the mean. .................................... 33 Figure 2.9 Proportions of inorganic (stable, labile, and available) and organic P in field soil and ditch sediments compared to total P at the a) dairy farm, b) turf farm, c) mixed site, and d) uncultivated site. .............................................................................................................................................................. 34 Figure 2.10 Linear relationships between selected soil factors and AAO and Mehlich-III P saturation indices: soil organic matter content vs. a) DPSox and b) M3-PSR I; clay content vs. c) DPSox and d) M3-PSR I.. ............................................................................................................................................ 34 Figure 3.1 Integration of 31P peaks in samples using either of the two methods a) integration of distinct peak(s) of phosphate, phosphate monoesters and other compounds, and pyrophosphate; and b) x  deconvolution of peak overlap using Lorentzian fit followed by integration of the two distinct sections of the spectrum. Integral proportions were normalized to sum to a value of 1. ..................... 51 Figure 3.2 The linear relationship between solutions of known concentrations of NH4H2PO4 and their respective 31P NMR peak area. ............................................................................................................. 52 Figure 3.3 XRD spectra for a) topsoil (0-20 cm) from the dairy farm with no chemical pre-treatment performed; and b) topsoil (0-20 cm) from the dairy farm with an AAO pre-treatment. ...................... 56 Figure 3.4 XRD spectra demonstrating the presence of struvite in topsoil (0-20 cm) at the dairy farm. .............................................................................................................................................................. 57 Figure 3.5 The solid-state 31P- NMR spectra of topsoil (0-20 cm) from the treatment sites, and ditch sediments (0-20 cm) from the mixed site compared with mineral peaks of calcium phosphate and synthetic struvite. Inset (left): The struvite peak relative to the broad, asymmetric inorganic P peak in dairy field soil. ...................................................................................................................................... 58 Figure 3.6 Selected 31P NMR spectra of topsoil (0-20 cm) from the a) dairy farm, b) turf farm, c) mixed site, and d) uncultivated site. ..................................................................................................... 60 Figure 3.7 Selected 31P NMR spectra of ditch sediments (0-20 cm) from the treatment sites: a) dairy farm, b) turf farm, and c) mixed site..................................................................................................... 61 Figure 3.8 31P scaled NMR spectra of field soils from the treatment sites (dairy, turf, and mixed) at depths of) 20-50 cm and b) 50+ cm. ..................................................................................................... 62 Figure 3.9 Concentrations (mean values when applicable) of types of P compounds (phosphate, phosphate monoesters and other compounds, and pyrophosphate) in organic extracts of field soil and ditch sediments from the treatment sites: a) dairy, b) turf, and c) mixed. ............................................ 64 Figure 3.10 The change in zeta potential of topsoil (0-20 cm) within a pH range of 4-8 at the a) dairy farm, b) turf farm, c) mixed site, and d) uncultivated site. ................................................................... 65 Figure 3.11 Selected Langmuir isotherms for: a) topsoil (0-20 cm), b) soil at 20-50 cm depth, and c) soil at 50+ cm depth at the study sites; and d) ditch sediments (0-20 cm) at the treatment sites. ........ 69 Figure 3.12 Selected Freundlich isotherms for: a) topsoil (0-20 cm), b) soil at 20-50 cm depth, and c) soil at 50+ cm depth at the study sites; and d) ditch sediments (0-20 cm) at the treatment sites. ........ 70xi  Nomenclature δ Chemical shift  Δ Variation of a variable or function  µ Micro AAO  Acid ammonium oxalate  AEC  Anion exchange capacity  Al Aluminum AR Aqua regia  ATP Adenosine triphosphate  BC Buffering capacity  BI  Buffering index C Carbon ˚C Celsius Ca Calcium CaHPO4 • 2H2O  Dicalcium phosphate CEC Cation exchange capacity  cm Centimetre DAP Diammonium phosphate  DCP Dicalcium phosphate DNA  Deoxyribonucleic acid DPS Degrees of phosphorus saturation DPSox Pox/ (Feox + Alox) EC Electrical conductivity  EDTA Ethylenediaminetetraacetic acid Fe Iron g Gram G6P Glucose-6-phosphate   ha Hectare H2PO4-  Dihydrogen phosphate HCl Hydrochloric acid HMP Hexametaphosphate HNO3 Nitric acid HPO42- Hydrogen phosphate Hz Hertz IP6 Myo-inositol hexaphosphate  K Kelvin kg kilogram km2 Kilometre squared KW Kruskal-Wallis   L  Litre LiCl Lithium chloride M-3 Element concentrations from the Mehlich-III extraction  M3-PSR I PM-3/(FeM-3 + AlM-3) M3-PSR II  PM-3/AlM-3 MAP Monoammonium phosphate  MBC Maximum buffering capacity  MCP Monocalcium phosphate M  Molar  m Milli xii  Mg Magnesium MgHPO4 • 3H20 Magnesium hydrogen phosphate  min Minute mmho  Milli mho mmt Million metric tonne Mn  Manganese MPP Monopotassium phosphate mV  Millivolt n  Sample size N Nitrogen NaOH Sodium hydroxide NH4H2PO4  Ammonium phosphate monobasic NH4MgPO4 Struvite NMR Nuclear magnetic resonance ox Element concentrations from the acid ammonium oxalate extraction  P Phosphorus PBC Phosphorus buffering capacity  Pi  Inorganic phosphorus Po Organic phosphorus  PO43 Orthophosphate ppm Parts per million PR  Phosphate rock PSI  Phosphorus sorption index  s Second SEM Standard error of the mean SOM Soil organic matter TSP Triple superphosphate XRD X-ray diffraction ZPC Zero point of charge    xiii  Acknowledgements To Les Lavkulich: thank you for letting me ask as many questions as I needed to, and for not giving away all the answers. Your humour, critique, and kindness were invaluable. A sincere thanks to Hans Schreier for always thinking of the bigger picture for why this research matters. I appreciate your candor and continual support. Thank you to Ken Ashley, a member of my supervisory committee, for the perspective you provided on this project. The analysis of this research would not have been possible without the exceptional insight provided by Dr. Andrew Lewis and Maureen Soon. Thank you for opening your doors to me.  Many thanks to Martin Hilmer, Jenny Lai, Jophat Engwayu, and Sally Finora for providing technical support.  Thank you to my friends and colleagues in the Soil Science Department for your encouragement, and for sharing the load of graduate school with me.  To end, I’d like to express my gratitude to my family and friends who have spent these past years supporting me through the highs and lows of completing this research. The work is finally done.     xiv  Dedication        For Mama.   For making everything possible.      1  1. Introduction and overview Phosphorus is essential to life. It is the basis of terrestrial food webs because it is fundamental to biochemical reactions such as photosynthesis, and DNA and protein synthesis (Ruttenberg, 2003; Smil, 2000). Agricultural activity has interfered with the P cycle in several ways: i) accelerated erosion and runoff due to conversion of forests to farmland; ii) intensive recycling of organic waste from livestock cultivation; and iv) application of inorganic fertilizers (Cordell et al., 2009; Johnston et al., 2014; Smil, 2000). While the Sanitation Revolution (late 19th century) converted civilization from a P-recycling society to a P-throughput society, the Green Revolution (1940-late 1960s) established a dependence on mined phosphate rock (PR) in replacement of organic sources (Ashley et al., 2011).  It is well established that PR, similar to oil and other non-renewable resources, will follow a peak production curve, commonly termed “peak phosphorus” (Cordell et al., 2009; Neset & Cordell, 2012; Van Kauwenbergh, 2010). At current rates of extraction, it is predicted that economically viable reserves will reach production peak by 2033-2034, and will be exhausted in 50-100 years. More conservative measures suggest that exploitable reserves for fertilizer production may last 300-400 years (Van Kauwenbergh, 2010). While over 30 countries extract PR, in 2011 the top 12 producers account for 95% of the total, and the top three (United States, China and Morocco) for 66% (Cordell et al., 2009; Smil, 2000; Van Kauwenbergh, 2010). The skewed distribution of PR reserves holds meaningful implications for the geopolitics of food security.  Agronomic P imbalances occur because of disparities between the amount nutrients applied and removed by harvested crops (MacDonald et al., 2011; Smil, 2000). Since plants can only take up P in the soil solution, it is necessary to apply more than is removed by the harvested crop to achieve an acceptable yield  (Johnston et al., 2014; Syers et al. , 2008).  This is done by applying fertilizer or manure to raise the plant-available P in soil to a critical level for the crop and soil type, and then maintaining it at this level by replacing the P removed from a harvested crop.  P surpluses in 30% of the global cropland area are primarily driven by excessive manure and fertilizer applications (MacDonald et al., 2011). Schreier et al. (2003) reported that in the Lower Fraser Valley, P application on agricultural land exceed the crop requirements by a significant margin due to limited soil P availability to plants. A conservative calculation of the global P budget measures an average annual average gain of 1.5 kg/ha (Smil, 2000). Nutrient budget calculations in the Sumas Prairie have shown an annual surplus of application reaching 50 kg/ha/year for phosphorus (Schreier et al., 2003). Land management can affect the dominant P compounds present in the soil and their relative mobility and bioavailability (Abdala et al., 2015; Pant et al., 2002; Smith et al., 1998). Long term application 2  of manure was found to result in greater P mobility due to the higher proportion of organic P (Frossard et al., 1989) and potentially due to the formation of less stable and therefore more soluble Mg-phosphates (Güngör et al., 2007; Miyittah et al., 2012) Biochemical reactions among inorganic and organic forms of phosphorus, the chemical solubility of these compounds, and (de)sorption reactions occurring at the soil particle surface determine the potential for phosphorus mobility and transport to surrounding water systems, through surface runoff or subsurface flow. Eutrophication results in excessive production of algae and cyanobacteria and subsequently hypoxia or anoxia, which can produce negative environmental, social, and economic impacts including fish kills, unpalatable drinking water, and unpleasant aesthetics for recreation (Correll et al., 1998; Daniel et al., 1998; Sharpley et al., 2003).  The study, focused on the Sumas Prairie, British Columbia is an ideal case study because: 1) it has been intensively cultivated with dairy, hog, poultry and produce farms for nearly 100 years with continuous applications of fertilizer and manure (IRC Integrated Resource Consultants Inc., 1994); serpentinitic contributions from the Swift Creek landslide have enhanced the geochemical environment by significantly increasing the proportion of Mg to Ca in soil (Baugé et al., 2013a and 2013b); and 3) a high proportion of fields in West Sumas (82%) and Sumas (89%) have P levels warranting high environmental risk (Kowalenko et al., 2007). As such, the Sumas demonstrates the convergence of P deficiency and excess; thus, presenting an opportunity for investigation of P dynamics in the soil to achieve a balance between soil fertility and environmental contamination.  1.1 Phosphorus in the environment: natural flows, human impacts, and food security Phosphorus is the eleventh most abundant elements in the earth’s lithosphere at 1180 ppm (0.1%) (Smil, 2000). In terrestrial systems, P is found in bedrock, soil, or living organisms. The P biogeochemical cycle is comprised of four components: i) tectonic uplift and exposure and weathering of P-bearing rocks; ii) physical erosion and chemical weathering of rocks to form soil; iii) release of dissolved and particulate P to rivers; and iv) sedimentation of P associated with the burial of organic and mineral matter in sediments (Ruttenberg, 2003; Smil, 2000). The complete cycle occurs on the timescale of 107 to 108 years (Smil, 2000). However, secondary land- and water- based cycling of organic P has a turnover time of 10-2 to 100 years. This occurs when plant litter, dead microorganisms, and other biomass are mineralized and P becomes available for autotrophic production (Ruttenberg, 2003; Smil, 2000).  3  Mining of PR began in the middle of the 19th century, and was substantially expanded after 1950 (Cordell et al., 2009; Reijnders, 2014; Smil, 2000). Extraction rates are estimated to be 13-16 Mt P/year. Approximately 85% of mined PR is used for agriculture, with 80% applied to land as fertilizer, and 5-10% as supplements in animal feed (Johnston et al., 2014; Smil, 2000). Smil (2000) estimates that at the beginning of the 19th century, crops assimilated ~ 1 Mt P/year compared to 12 Mt P measured in 2000. Anthropogenic erosion and runoff were ~5 Mt P/year in excess of the natural denudation rate, while in 2000 it measured about twice as much as the natural denudation rate, providing context for the recent increase in environmental contamination and responsible for the global eutrophication epidemic. While there is uncertainty in the timing of the production peak, the quality of P fertilizers is declining as prices continue to increase, resulting in economic hardship for producers and consumers alike (Ashley et al., 2011; Cordell et al., 2009). The significance of the peak is not when 100% of the reserves are depleted but when the quality of remaining reserves is lower and harder to access, making mining and processing uneconomical.  The IFDC estimate of remaining and potential PR reserves ranges from 290,000 to 460,000 mmt (Van Kauwenbergh, 2010) or 2358 Mt P8 in remaining reserves (Cordell et al., 2009).  Food production is expected to increase to meet the rise of global populations. Since P cannot be substituted in food production, or synthesized in a laboratory, countries with PR reserves have shifted towards reducing exports to maintain domestic supply, leaving countries with P deficient soils vulnerable to food insecurity (Cordell et al., 2009; Neset & Cordell, 2012). Further, globalized trade of food crops have altered the P cycle because application and uptake no longer coincides with where P is recycled and mineralized. This emphasizes the need for more efficient use and recycling of P to close the cycle of supply and demand.  1.2 Inorganic and organic phosphorus in the soil In the soil environment P can be broadly categorized into inorganic (Pi) and organic (Po) forms.  Pi generally accounts for 35-70% and Po for 30-65% of total P in soils (Shen et al., 2011). Plant uptake of P by roots occurs mainly as H2PO4- and to a lesser degree as HPO42-; at pH 7.2, H2PO4- and HPO42- are present in the same proportions (Harrison, 1982; Quiquampoix and Mousain, 2005). Within the range of pH 4 and 5, H2PO4 is the dominant form; HPO42- in the near-neutral pH range; and PO43- under alkaline conditions.  4  While apatite is the only primary mineral with substantial P content, secondary minerals include oxides and carbonates which have P chemisorbed to their surfaces, along with minerals with P as a structural component (Sharpley, 1995; Smeck, 1985). The latter include variscite (an Al phosphate), strengite (an Fe phosphate), and brushite, monetite, and octocalcium phosphate (Ca phosphates). In acidic soils (<5.8), P is dominantly adsorbed to Fe/Al oxides and hydroxides such as gibbsite, hematite and goethite. In neutral to basic soils (pH > 5.8), precipitation reactions determine P retention; P can be adsorped on the surface of calcium carbonates or precipitated as dicalcium phosphate (available to plants) and then transformed into more stable forms such as octocalcium phosphate and hydroxyapatite (less available to plants). However, all species (Fe, Al, Ca, Mg phosphates) can coexist within a wide pH range (Lindsay et al., 1989). Due to compatibility of coordination numbers in the mineral structures, P is most strongly fixed by Fe (pH<3), followed by Al (pH 4-6), Ca (pH >7.5), and Mg; thus P is most soluble at pH 6-7. The hydrolysis of primary minerals is typically too slow to meet crop needs (Sharpley, 1995; Smeck, 1985). In contrast, the reactivity of secondary minerals is highly variable because they are comprised of crystalline and non-labile adsorped P, where they often dictate long-term dissolution of P and movement towards soil equilibria.  Phosphate esters of biological importance in soil can be categorized into inositol phosphates, phospholipids, and nucleic acids (Quiquampoix & Mousain, 2005; Turner et al., 2002). The majority of Po (up to 90%) is chemically (Turner et al., 2002) and physically protected from degradation, and very slowly contribute to the labile pool. Inositol phosphates (phytic acid being the most common form in soils) are typically the dominant form of Po in the soil, comprising up to 80% (Quiquampoix & Mousain, 2005; Turner et al., 2002). These compounds are resistant to mineralization, and are often found in complex forms associated with clay minerals, fulvic and humic acids, proteins, and metallic ions. In acidic soils, inositol phosphates form insoluble salts and in alkaline soils, precipitate with calcium. In contrast, phospholipids rapidly degrade in the soil (Bowman & Cole, 1978; Turner et al., 2002). They can originate from plants, animals, and microbes, and comprise approximately 0.5-7% of soil Po (averaging at 1%). Nucleic acids, making up 3% of soil Po, are derived from the decomposition of living matter and can be readily mineralized, incorporated into microbial biomass, or combined with other soil constituents (Harrison, 1982; Quiquampoix & Mousain, 2005).  If not mineralized, the fate of Po includes: losses through leaching, runoff or transfer to water sources; adsorption to Fe-Al oxides and clay particles; incorporation in long-term pools of organic matter; and complexation and precipitation resulting in mineral dissolution (Harrison, 1982). 5  Mineralization of Po has a significant influence on the bioavailability of P in soil, and is mediated by soil organisms and plant roots through secretion of phosphatases, organic acids, as well as inorganic ions which are generally triggered by low P concentrations in soil (Gahoonia et al., 1992; Li et al., 1997; Shen et al., 2011). This activity is generally concentrated in the rhizosphere and is influenced by soil moisture, temperature, surface physical-chemical properties, and soil pH and Eh.   1.3 Phosphorus as a nutrient amendment Management of solution P is necessary to ensure plant productivity, and has motivated much of the research to date. During the 19th to early 20th century, P was understood to be permanently fixed in soils (Johnston et al., 2014; Syers et al., 2008). By the 1950s-1960s, thermodynamic models were developed to assess P (de)sorption behaviour (Syers et al., 2008). Currently, the understanding of P-plant interactions is based on long-term reversible release of P from inner and outer sphere coordination complexes with other soil constituents (Johnston et al., 2014; Shen et al., 2011; Syers et al., 2008). Operational definitions of P forms have been developed based on chemical extractions theoretically targeting stable or labile P fractions (Bjorkman & Reiners, 2013; Hedley et al., 1982). Thus, correlative empirical relationships have been developed for soil P and plant uptake of P. Until recently, with the application of nuclear magnetic resonance analysis for characterization of Po, much of the understanding of P dynamics in soil have focused on Pi.  Concentrations of P in the soil solution can range from 10-4M (very high) to 10-8M (very low) (Johnston et al., 2014). Crop requirements for P range from 0.3-0.5 kg P/ha /day during stages of rapid growth.  Crop recovery of P has been reported to range from 10-25%, (Johnston et al., 2014); however, Smil (2010) reports that recovery up to 50-60% is not uncommon. The remainder of plant P uptake comes from reserves in the soil, which indicates that P residues are available for use by subsequent crops.  P transport to plant roots occurs as a product of two processes:  relatively fast, reversible adsorption, and relatively slow, practically irreversible, diffusion-limited precipitation (McDowell & Sharpely, 2003; Sjoerd et al., 1988). Since orthophosphate is an anion and agricultural soils tend to be negatively charged and have variable charges due to high organic matter content, steady state in solute concentration is rarely achieved. P availability to plants is believed to depend on three factors: P intensity (Q) estimated by the concentration of P in the soil solution; P quantity (I) which is the portion of P in solid phase that 6  equilibrates with the soil solution; and P buffering capacity (PBC), which is the ability of the soil to resist changes in solution P (Bridgham et al., 2001; Shirvani et al., 2005).  Fertilizer is commonly applied as monocalcium phosphate (MCP), dicalcium phosphate (DCP), triple superphosphate (TSP), monoammonium phosphate (MAP), diammonium phosphate (DAP), and monopotassium phosphate (MPP) (Smil, 2000). Application of MCP has a significant impact on soil physicochemical properties. After application, MCP undergoes a wetting process releasing large amounts of protons, phosphate, and DCP and eventually forms a P-saturated zone (Benbi & Gilkes, 1987). Three concurrent Pi reactions occur at this time: precipitation, adsorption, and direct reaction (pH 1-1.6) which solubilizes metals ions that react with Pi to form amorphous Fe and Al phosphates that can be partially available to plants.  Manure have been found to be enriched in stable and labile forms of Po such as, phytase, DNA, pyrophosphate, and polyphosphates (He et al., 2009; Pagliari & Laboski, 2012). Poultry manure has the greatest enrichment of P, followed by hog and dairy (Sharpley et al., 2003). Application of manure can result in exudation of organic acids to facilitate mineralization of Po and indirectly solubilize Ca-phosphates, as well as decrease the overall adsorption of soil P through competition with humic acids which contain negatively-charged carboxyl and hydroxyl groups (Gahoonia et al., 1992; Shen et al., 2011). While continual input of fertilizer and manure may result in annual P surpluses, without additions most agricultural soils would be P deficient.    1.4 Selected soil factors contributing to phosphorus solubility P solubility is determined by adsorption-desorption processes influenced by the interaction of soil factors such as pH, redox state, clay content, organic matter (SOM) content, mineralogy, status of other soil nutrients, and agricultural management. Mildly weathered silicate clays and high SOM content characteristic of soils in the Lower Fraser Valley are conducive to the development of variable-charged (v-c) soils susceptible to protonation and deprotonation of surface hydroxyl groups due to pH changes (Sollins et al., 1988). Protonation produces an anion-exchange capacity (AEC), while deprotonation can produce a cation-exchange capacity (CEC) which can enhance P mobility through ionic repulsion.  In an investigation of sandy loam soils Eghball et al. (1996) found movement of P to deeper soil was greater with manure application (1.8 m) than with fertilizer application (0.9 m). Combined manure and fertilizer application resulted in the greatest labile P concentrations, likely because the amount of P applied was double the P applied in plots receiving 7  only manure or fertilizer. Since manure application is guided by crop requirements for N rather than P, and manures tend to have a narrower N:P ratio than most crops require (i.e. 7-11:1 for crops compared to 2-6:1 in manures), prolonged use can produce large P reserves in the soil, beyond its sorption capacity (McDowell & Sharpley, 2001; Smith et al., 1998).  Generally speaking, pH governs the adsorption properties of major P-fixing minerals as well as the solubility and dissolution kinetics of P containing minerals (Lindsay 1989). In studies across eight soil orders in the US (Tiessen et al., 1984), as well as in long-term cultivated paddy fields in Cixi County (Huang, et al., 2014) high base saturation was found to maintain P as primarily Ca-phosphates.  With increasing soil development, pH and base saturation declined, resulting in accumulation of organic matter and secondary minerals. High clay content, and to a lesser degree silt, is associated with reduced inorganic P solubility through direct surface sorption of orthophosphate, or association with Fe, Mn, and Al oxides with sorbed P (Lair et al., 2009; Tiessen et al., 1984). In old paddy fields (300-1000 years), Huang et al. (2014) found increased P solubility due to losses of clay, Fe and Al, and CaCO3 from long-term eluviation as a result of artificial irrigation.  The interaction of pH and clay is evident in the increased release of P colloids (1 nm-1 μm diameter) within a pH range of 1.4-6 due to the dissolution of Al oxides and clay minerals (Liang et al., 2010). P colloids can account for 13-95% of total P in solution, thereby contributing significantly to losses via leaching (Shand et al., 2000). Gustafsson et al. (2012) also found sorption of P in Swedish soils expressed pH dependence: soils with high clay content had minimal P solubility between pH 6 and 7, while sandy soils did not express minimum solubility.  They hypothesized that hydroxyl Al and Fe interlayers of clay minerals may be important adsorbents in clay soils and the occluded P associated with these layers are released during acidification, and bound up under alkaline conditions.  In addition to increasing P content, application of fertilizer and manure can enhance the soil chemical environment. Long term fertilizer application has been found to increase concentrations of dissolved Pi and Ca resulting in precipitation of Ca-phosphates and facilitation of adsorption to clay minerals under neutral to alkaline conditions (Devau et al., 2011; Parton et al.,1988). Abdala et al. (2015) proposed that long-term application of swine and dairy manure increased soluble organic matter (i.e. low molecular weight organic acids such as citric, oxalic, and malic acids) promoting the formation of amorphous Fe and Al minerals which increase P sorption capacity in soils. Organic P species (ATP, choline phosphate, and glucose-6-phosphate) have also been demonstrated to be less strongly sorbed to soil colloids and more prone to leaching, compared to inorganic P (Frossard et al., 1989). The quality of organic fertilizer along with the type of parent material can influence the likelihood for 8  SOM to increase or decrease P sorption. Yu et al. (2013) found that both mineralized and non-decomposed manure decreased sorption of basaltic parent material due to competition with organic acids, and the enveloping effects of organic fertilizers. In alluvial soils, addition of mineralized manure was found to reduce P sorption due to competition for adsorption sites, while addition of non-decomposed manure increased P sorption due to an overall increase in sorption sites. The impact of manure application can be long lasting as Pant et al. (2002) demonstrates in a study of P solubility in abandoned-intensive (34 years in dairy production  followed by 8 years in forage production) vs. active-intensive (20 years in dairy production) agricultural production.  Anaerobic conditions in the A horizon of the active site were found to promote P mobility through Fe reduction, while aerobic conditions in the B horizon of the abandoned site promoted P mobility through near neutral pH (~6.5) at which P solubility is greatest. Enhanced P mobility in the B horizon may also be attributed to higher SOM content in the A horizon which promotes adsorption of positively charged Po molecules, limiting their mobility (Frossard et al., 1989). P release due to fluctuation in redox state can be meaningful when Fe is abundant and soil water tables are impacted by groundwater or poor drainage. In the Kleine Nete catchment in Northern Belgium, Baken et al. (2015) found that reduction of Fe, typically during the winter, solubilized P in the sediments of drainage ditches.  While the interaction of P with Ca, Fe and Al minerals have been well investigated, the role of Mg-phosphates in promoting P solubility is of current interest. In laboratory studies, Jackman & Black (1952) found that Mg-phytates (compared to Fe, Al, and Ca phytates) had the widest pH range of solubility, being insoluble only above pH 9.7 in the presence of excess Mg. Further, struvite, a Mg-phosphate (NH4MgPO4) has been identified in raw and anaerobically digested manure samples, and demonstrated to be meaningful for the slow release of P in soils (Güngör et al., 2007; Miyittah et al., 2012). In field studies, Pant et al. (2002) found Mg-associated P contributed to increased P release under aerobic and anaerobic conditions, since it is not as stable as Ca-associated P, thus requiring maintenance of equilibrium by each P compound in solution.   1.5 Environmental contamination of phosphorus  Phosphorus from agricultural runoff is an important component of nonpoint-source pollution to freshwater ecosystems where P tends to be a limiting nutrient (Correll, 1998; Daniel et al., 1998; Sharpley et al., 2003). Nutrient loading gained widespread attention in Canada in the 1960s and 1970s because of algal blooms in the Great Lakes which prompted legislation limiting phosphate levels in 9  laundry detergents and sewage effluents (Environment Canada, 2013). Since then, urbanization and agricultural intensification has led to eutrophication of freshwater systems across Canada, particularly the Fraser River, South Saskatchewan River (Lake Winnipeg), the Great Lakes and the St. Lawrence River (Environment Canada, 2013).  Managing eutrophication is challenging because it is not of economic important to farmers as the P loss comprises only 1-2 percent of applied P (Correll, 1998; Sharpley et al., 2003). Critical surface water concentrations of P, above which eutrophication is accelerated, range between 0.01 and 0.02 mg/L, which is an order of magnitude lower than P concentrations in the soil solution critical for plant growth (Canadian Council of Ministers of the Environment, 2004; Sharpley et al., 2003).   ‘Agricultural runoff’ describes field processes of surface runoff and subsurface flow. Subsurface flow encompasses leaching through the soil matrix and preferential flow through drains and macropores (Heathwaite & Dils, 2000; McDowell & Sharpley, 2001). Transported P occurs in particulate and dissolved forms (Correll, 1998; Daniel et al., 1998). Particulate P is associated with soil particles and organic matter eroded during flow events, and comprises 80% of transported P in surface runoff from most cultivated land (Sharpley et al., 1992). Sediment P can be a long-term source of P for aquatic biota. Alternatively, dissolved P tends to originate from grass or forest land or non-cultivated soils, and account for 80% of P loss (Sharpley et al., 2003). Dissolved P is immediately available for biological assimilation. In the aquatic system, P only occurs in the pentavalent form as orthophosphate, pyrophosphate, longer-chain polyphosphates, organic phosphate esters and phosphodiesters, and organic phosphonates (Correll, 1998).  Catchment studies in Denmark have shown that amounts and forms of P losses vary throughout the year, though much of the annual loss occurs in relatively short periods (Grant et al., 1996). Organic P was found to comprise 16-23% and inorganic P 13-47% of total exported P. Heathflow and Dils (2010) found surface runoff to be dominated by dissolved P, while preferential flow was an important pathway for particulate P transport in UK grasslands. Additionally, they suggest that matrix flow is unlikely to result in substantial P transport, though there is some concern that Po, being more mobile than Pi, can leach further down in the soil profile (McDowell & Sharpley, 2001). Contributions of P from drainage tiles to watershed export can be significant, estimated to equal up to have half of annual watershed discharge (King et al., 2014).  Of particular interest to agricultural systems is the role of ditches to act as conduits for concentrated contaminant flow, as they can carry eroded sediments during high flow, and serve as preferential 10  pathways for dissolved P during low flow (Sharpley et al., 2007). Liu et al., (2013) and Sharpley et al. (2007) found ditch sediments to be a source of dissolved P, and to have greater capacity to sorb P due to a greater proportion of organic matter, clay-sized material, and amorphous Fe and Al. In addition to geochemical interactions, biotic processes contributed 40% to P released from ditches (Sharpley et al., 2007). Sampling in the priority basins of the Lake Okeechobee Basin indicated that total P in ditch sediments was related to soil OM and metal content, specifically Ca for ditches draining dairy fields (Dunne et al., 2007). Degrees of phosphorus saturation measured 36%, above the threshold 25% (Breeuwsma et al., 1995) in dairy ditch sediments indicating high risk of P loss, which may be enhanced during flooding when ferric forms of Fe are reduced and release previously bound P (Dunne et al., 2007).   1.6 The Sumas Prairie  The Sumas watershed spans 277 km2 across the City of Abbotsford, the City of Chilliwack, and Whatcom County (Washington, USA) (IRC Integrated Resource Consultants Inc., 1994; Shead, 2004). Approximately 48% of the area lies in Canada and 52% in the United States. Located in the Coastal Western Hemlock biogeoclimatic zone, the watershed is characterized by mild winters and cool summers. Mean annual temperature is 10.4°C (1981-2010) with an average daily minimum and maximum of 5.8˚C and 15.1˚C, respectively (Government of Canada, 2015).  Mean annual precipitation is 1538 mm (1981-2010), the highest average rainfall of the BC’s 14 biogeoclimatic zones. The headwaters of the Sumas River are located in Whatcom County, and flows northeast into the Fraser Valley, crossing a low-lying floodplain located between the Vedder and Sumas Mountains, the Sumas Prairie (Shead, 2004). The geology of the central and eastern sections of the valley is characterized by silty aeolian capping over post-glacial lacustrine deposits (Luttmerding, 1981; Piteau Associates, 2005). Soils of the region are typically 8000 years old, are within the Gleysolic and Brunisolic orders, and vary across 11 soil classifications and 7 management groups (Luttmerding, 1981 and 1984). The valley encompasses 10,000 acres (40 km2) which have been agriculturally developed since 1927, when the Sumas Lake was drained and dyked for agricultural and flood control purposes (IRC Integrated Resource Consultants Inc., 1994; Shead, 2004). Highest flows in the watershed occur in November to January, and lowest flows in July to September. 11  Agriculture is essential to Abbotsford and has supported the long term growth of the community. Total gross farm receipts in 2014 were nearly $700M, with the industry employing over 7000 people (City of Abbotsford, 2014a). Nearly 5700 hectares in the Sumas Prairie are used for agricultural purposes (IRC Integrated Resource Consultants Inc., 1994); currently, the majority of the landbase is occupied by dairy farms, and a minority by poultry farms, and produce and nursery farms. In the past several decades, intensification of livestock operations on a limited landbase has been identified as a crucial reason for excessive nutrient loading in surface waters (Kowalenko et al., 2007; Schreier et al., 1998; Shead, 2004). Berka et al. (2001) found that from 1986 to 1996, while cattle populations remained constant, hog and chicken populations increased by 59% and 165%, respectively. In more recent years (1996-2011), hog populations have drastically declined (-93%), dairy have decreased and stabilized (-22%), and poultry have increased intermittently (+2%) (Statistics Canada, 2015). Nutrient loading is exacerbated by application of manure in the fall when runoff is most likely to occur (Schreier et al., 1998; Shead, 2004). Approximately 1,238,360 L of dairy/hog/poultry manure is produced per day in the Sumas watershed, with an overall loading rate of 262 L/ha/day on land utilized by livestock producers (4728 ha) (IRC Integrated Resource Consultants Inc., 1994).  Dairy producers generate 65% of the manure. In a farm survey of the Sumas Prairie conducted in 1993, chemical fertilizer was applied on 66% of surveyed area by dairy producers and 17% by produce farms and nurseries. The chemistry of agricultural soils in the Sumas is enhanced by contributions of serpentinitic material from the Swift Creek landslide in Washington, USA (Baugé et al., 2013a and 2013b; Whatcom County Public Works, 2012). The confluence of Swift Creek with the Sumas River covers approximately 495 acres. The geologic deposit contains naturally occurring asbestos along with trace metals of cadmium, cobalt, manganese, and nickel. Serpentine is an ultramafic rock (contains >70% mafic or ferromagnesian materials), which is essentially a magnesium iron silicate formed by metamorphosis from peridotite (Kazakou et al., 2008; Whittaker, 1954). Characteristic of soils developing from serpentine are i) enrichment of magnesium relative to calcium, ii) deficiency of soil macronutrients, iii) high levels of phytotoxic heavy metals, and iv) low organic matter content. This may provide context for the persistent application of nutritional amendments to facilitate agricultural production. Baugé et al. (2014) have found evidence that the formation of soluble magnesium phosphates may enhance nutrient mobility in the region.  12  1.7 Thesis objectives and overview It is evident that there has been extensive research on the application of phosphorus as a nutrient amendment and its subsequent potential as an environmental pollutant.  Nonetheless, investigation of P in field soils and ditch sediments under different nutrient management regimes in the Sumas Prairie can contribute knowledge towards:  1) P dynamics in mildly weathered alluvial soils with high organic matter content and serpentinitic contributions; 2) the role of long-term nutrient management and specific soil factors to enhance P concentrations and solubility; and 3) the role of agricultural ditches for accumulation and transport of P.  Additionally, the use of precise and relatively novel techniques, such as nuclear magnetic spectroscopy (NMR), along with well-established extractions and thermodynamic models to capture the complexity of P soil interactions should provide insight into P mobility from the field to surrounding water bodies.  The main objective of this research is to evaluate how nutrient management regimes, including manure, fertilizer, mixed (fertilizer and manure), and uncultivated influence P compounds, their solubility, and (de)sorption reactions occurring in soils and adjacent ditch sediments in the Sumas Prairie. The specific objectives are as follows: 1. To determine the relative concentrations of inorganic and organic P pools in field soil and ditch sediments and how they vary with depth. 2. To evaluate the influence of specific soil factors (pH, clay content, and organic matter content) on P solubility. 3. To characterize inorganic and organic sources of P, and specifically investigate for the presence of struvite.  4. To evaluate surface properties of soil particles and their influence on P retention.     13  2. Evaluating immediate and long-term phosphorus pools and the influence of selected soil factors on phosphorus release   2.1 Synopsis Phosphorus contamination is an environmental concern in the Sumas Prairie particularly as the majority of surveyed fields have accumulated P in topsoil (0-15 cm depth), indicating immediate risk to surface runoff (Kowalenko et al., 2007). Furthermore, fish kills along with lowered hatching success and increased deformity rates of amphibians have been associated with agricultural runoff in the region (de Solla et al., 2002; Shead, 2004). In the Sumas River, total P concentrations have been found to exceed the provincial criterion of <15 µg/L for the protection of aquatic life, with highest concentrations reaching a mean of 265 µg/L in the Arnold Slough (IRC Integrated Resource Consultants Inc., 1994). P concentrations in topsoil of grass and forage corn production constituting high to very high risk of environmental contamination ranged from 50 to >100 mg P/kg using the Kelowna extraction, which is equivalent to 70 to >140 mg P/kg for the Mehlich-III extraction (Kowalenko et al., 2007). Since Sumas soils are coarse-textured and thus prone to fast leaching, transport of P in subsoil through subsurface pathways can result in accumulation of P in agricultural ditches, which are extensively used for drainage in the Sumas Prairie (City of Abbotsford, 2014b).  To date, there has been limited research conducted on British Columbia soils to establish methods of analyses to evaluate environmental risk from increasing P contents (Kowalenko et al., 2007; Schendel et al., 2004; Yuan & Lavkulich, 1995). Application and optimization of universal soil tests, such as Mehlich-III or acid ammonium oxalate, have not been well investigated, instead opting for use of the localized Kelowna extraction.  In the soil system, P can be understood as having a “trigger” range, similar to the framework proposed by the Canadian Council of Ministers of the Environment (2004) for eutrophication, where there is a desired range of concentrations and exceedance of the upper value indicates a potential environmental problem.  The purpose of this work was to assess the status of P in Sumas soils under different nutrient management regimes to better understand their capacity for environmental contamination. This was done through direct measurement of P concentrations in immediate and long term pools in topsoil, subsoil and ditch sediments; calculation of indices of P saturation; and evaluation of the influence of selected soil factors on P release.   14  Figure 2.1 Inorganic and organic P sources and pools in the soil (Stewart & Sharpley, 1987).  2.2 Introduction Intermediary P pools where transformations between inorganic and organic P sources occur, and P can either be taken up by plants, retained within the soil matrix, or lost to the environment are operationally described as the following ( Shand et al., 2000; Smeck, 1985; Walker & Syers, 1976) (Figure 2.1):  1. Soluble P (HPO42-and H2PO4): present in the soil solution and immediately available for plant uptake. 2. Labile P (Pi and Po): exchangeable on adsorption sites and in equilibrium with the soil solution; will be released to the soil solution if soluble P concentrations decline. 3. Moderately labile P (Pi and Po): strongly bonded to or adsorped within the matrices of soil components; can become plant-available over timescales from months to years. 4. Occluded P: has been precipitated as slightly soluble P compounds, is part of the soil mineral complex, or is physically encapsulated by minerals limiting interactions with more reactive P forms; is only very slowly available (over periods of many years) for plant uptake.    15  With soil development, increased weathering along with decreasing pH, favours the formation of secondary mineral (amorphous Fe and Al phosphates) resulting in decreasing pools of soluble and labile P (Tiessen et al., 1984; Walker & Syers, 1976; Yang & Post, 2011). The occluded P sink is formed by secondary P minerals encapsulated by other Fe and Al oxides or others soil minerals. When the final steady state is reached, occluded P and Po are the major forms present, where C/ Po is high and P availability is very low (Walker & Syers, 1976). A complete transformation from primary to occluded P requires hundreds to thousands of years. Since Sumas soils are young, proportions of occluded P will likely be low and Po high from manure enrichment. Further, the geology and limited weathering of Sumas soils indicates a tendency towards circumneutral pH at which Ca, Mg, Fe, and Al will all partially contribute to P solubility (Lindsay et al., 1989). pH and clay interactions are not favoured as low clay content is likely to increase overall P sorption.  P pools can be quantified by using established targeted chemical extractions, such as acid ammonium oxalate to extract P bound to amorphous Fe and Al oxides (McKeague & Day, 1966), or Bray and Kurtz P1 (Bray & Kurtz, 1945) and Mehlich-III  (Tran & Simard, 1993) to extract exchangeable P. Depending on the pH and mineralogy of soils, certain soil tests will be more appropriate than others. The advantage of using a standardized soil test is the ability to compare across geographical regions. Phosphorus saturation indices (i.e. DPSox, M3-PSR I and II) are viewed as good environmental indicators of P loss potential to runoff and leachate, and can be calculated by dividing the amount of P by the P sorption capacity of the soil (Beauchemin & Simard, 1999; Pellerin et al., 2006). With the intensive nutrient management, interactions of SOM with P can be complex: Singh & Jones, (1976) found organic residues having P contents < 0.3% tend to increase P sorption, whereas residues having P contents > 0.3% tend to decrease P sorption. However, it is generally accepted that higher SOM will contribute higher proportion of Po which has been demonstrated to be more mobile within the soil matrix (Frossard et al., 1989; Smith et al., 1998). Characterizing the carbon and nitrogen content of SOM specifies the quality of organic matter and status of mineralization and assimilation of nutrients in the soil, particularly since C, N, P stocks are affected by land management (Groppo et al., 2015; Manzoni et al., 2010). C, N, and P concentrations are related because nitrogen in the forms of amino acids and nucleic acids are needed for protein synthesis and the production of phosphohydrolases and other oxidative enzymes that catalyze 1) the mineralization of carbon substrates, and 2) the mineralization and immobilization of phosphate in cells of plants and microorganisms (Dao & Schwartz, 2010). C/P or N/P act as indicators of biological cycling of P in soil: widening of the ratios can result in increased mineralization and accumulation of Pi, whereas cycling limited by C or N can result in accumulation of Po (Dao & Schwartz, 2010; Smeck, 1985). 16  However, in a P-limited environment, Dao and Schwartz (2010) found that changes in N did not significantly influence C mineralization. In modelling carbon-phosphorus interactions, Parton et al. (1988) postulate that the active P fraction (i.e. microbial biomass and metabolites) had C:P ratios ranging from 30-80; the passive fraction (i.e. most stable organic P forms) had ratios of 20-200; and the intermediate forms in the slow soil fraction have ratios of 90-200. Thus, nutrient ratios can infer the potential for SOM to provide immediate and long-term release of P.   2.3 Methods   2.3.1 Site description and field sampling The Sumas Prairie is located in the Lower Fraser Valley of British Columbia, southeast of Metro Vancouver and just east of Abbotsford (Figure 2.2). Collection of field soil and ditch sediment samples took place on April 30, 2013 prior to seeding or turnover of soil. All study sites were located on the Canadian side of the Prairie, north of the Swift Creek Landslide in Whatcom County and all have been influenced by serpentinitic material carried and deposited by the Sumas River (Figure 2.3). The four study sites were managed for 20-40 years in their present day land uses and associated nutrient management regimes (personal communication, H. Schreier, March 2013) : these included i) a dairy farm managed solely with manure application; ii) a turf farm managed solely with fertilizer application; iii) a corn field with mixed management of manure and fertilizer additions; and iv) an uncultivated site with no agricultural inputs in the past two decades (Figure 2.4). Ditches and site depressions within the treatment sites (i.e. dairy, turf, and mixed sites) were also investigated (Figure 2.5). Specific site characteristics are summarized in Table 2.1. The soils at all study sites were originally classified as part of the Buckerfield Soil Series; however, through long-term drainage have been modified to be more characteristic of Buckerfield/Bates Soil Series (Luttmerding, 1981). According to the Canadian System of Soil Classification, these soils are Gleyed Eluviated Melanic Brunisol. The parent material is medium to moderately fine-textured lacustrine deposits, underlain by sandy material below 1-2 m depth. Topographically, these soils range from slightly depressional to gently undulating with slope gradients less than 10%. Elevations are 5-15 m above sea level. As such, Buckerfield/Bates soils are imperfectly drained and moderately pervious. Water holding capacity is high and surface runoff is slow. Watertables are high and during winter can result in surface ponding which limits the rooting depth to approximately 60 cm.  17  Municipal ditches are managed by the Diking, Drainage and Irrigation Department of the City of Abbotsford, and are cleaned every year with an excavator. During cleaning, a minimal amount of material is removed: only underlying vegetation to increase the oxygen concentration, and reduce the risk of flooding in winter (personal communication, Abbotsford City Official, November 17, 2014). The depth of ditches varies from field to field, but is typically 2 m. At the dairy farm, the personal ditch sampled had not been cleaned for the past five years. Municipal ditches are classified as either a “wet” or “dry” depending on how many months of the year they remain saturated (City of Abbotsford, 2014b). At the turf farm, a dry ditch was sampled, which was likely cleaned 8-10 months previous to sampling.  No ditch was found in close proximity to the soil transect at the mixed site. Since this site was undulating, a proxy for the ditch sample was taken from the deepest depression on the field, approximately 100 m away from the soil transect, where visible ponding of water occurred. A stratified-random design was used for field soil sampling along a randomly oriented 20 m transect selected in a random position 10 m from the nearest road to the field, with the limitation that the transect not be located near edges or corners of the field where agricultural management may not be representative of the overall site (Table 2.1) . Soil pits were dug for comprehensive collection of soil material. Five samples at three depths were taken in each field at 0-20 cm, 20-50 cm, and 50+ cm to evaluate the potential for vertical mobility of phosphorus and associated elements.  These depths also correspond to changes in texture, organic matter content, and saturation of the soil. Two randomly selected surface (0-20 cm) ditch sediment samples were collected from the treatment sites to assess lateral mobility of P. The sample size of each treatment site was 17, and 15 for the uncultivated site, to total 66 samples overall.     18  Table 2.1 Description of study site characteristics.  Site Cultivation Nutrient Amendment Site Characteristics Dairy  Rotations of corn and forage grass Dairy manure is applied yearly and lime every second year (Bauge et al. 2013a and b). Located close to the Canada-United States border, along a channel of the Sumas River (Figure 2.2). The site was flat and uniformly covered in a forage crop with a ditch located approximately 35 m from the soil transect. Turf  Turf No application of manure. Several applications of 40-0-0 (urea and ammonium sulfate) and 23-3-23 (urea, muriate of potash, and mono-ammonium phosphate) fertilizers and lime are applied yearly (Bauge et al. 2013a and b).  At the turf farm, the field had been levelled and sampling took place where turf had recently been removed; ditch samples were taken 15 m away from the soil transect. Mixed Rotations of corn and forage grass Dairy manure and lime is applied yearly, along with applications of 16-16-16 (monoammonium phosphate, ammonium nitrate, potassium magnesium sulfate, ferrous sulfate and zinc sulfate) fertilizer on the corn crop. Irregular ground cover was exhibited at the mixed site as a cover crop had been recently removed and the stalk of corn from the previous harvest remained.      Uncultivated  Grass None Located in close proximity to the manure site. Grass growing on the site was approximately 20 cm high, and sampling took place near a grove of trees away from roots.     19       Figure 2.2 The location of the Sumas Prairie in the Lower Fraser Valley (adapted from Boyle et al., 1997). . Figure 2.3 The locations of the study sites within the Sumas Prairie relative to the Sumas River and Swift Creek Landslide in the delineation of the Sumas Watershed.  20        Figure 2.5 The ditch sediments from the treatment sites, depicting location relative to the main field, water saturation, and degree of vegetative cover at the time of sampling. Figure 2.4 Conditions of the four study sites in the Sumas Prairie, depicting the varying degrees of ground cover, cultivation type, and nutrient management regimes. 21  2.3.2 pH, electrical conductivity (EC) and soil organic matter (SOM) Soils were air-dried for 48 hours prior to sieving through a 2 mm sieve. Approximately 500 g of individual samples for each depth at each site were collected for individual soil and ditch samples. The remaining soil material collected was used to form composite samples for each depth at each site.  Unless specified, individual samples were used for analysis. pH measurements were conducted on fresh, non-sieved samples in a 1:2 distilled water solution, and a 1:3 0.01M CaCl2 solution to determine both active and exchangeable acidity (Hendershot et al.,1993). Soluble salts were was assessed by electrical conductivity measurements in a 1: 5 distilled water solution according to Rayment & Higginson (1992). Soil organic matter content was estimated through loss on ignition measurements at 350°C and 550°C (Atkinson et al., 1958).  2.3.3 Particle size distribution The rapid method for particle size determination (Kettler et al., 2001) was used to characterize the soil texture.  Due to the relatively coarse textured nature of Sumas soils, this method was deemed sufficient to characterize the dominant sand and silt fractions. All surface (0-20 cm) replicates from the field soil and ditch sediments, and three subsoil replicates at each depth (20-50 cm and 50+ cm) were analyzed. Twenty grams of soil was heated at 70˚C while 30 % H2O2 was added at intervals to digest the organic matter. After two hours, the samples were dried in an oven overnight at 50˚C. This was followed by, 15 g of the digested samples were sieved using a 2 mm sieve and mixed with 45 mL of 3 % hexametaphosphate (HMP). The samples were then shaken for two hours. Following deflocculation of the soil by HMP, soils were wet-sieved using a 53 µm sieve and ~600 mL of distilled water to isolate the sand fraction and suspend the clay and silt fraction. The sand fraction was collected and dried in an oven at 55˚C overnight. The silt and clay fractions were stirred thoroughly to suspend all particles and left to settle over three hours at room temperature (~22-24˚C).  After this time, the clay fraction still in suspension was decanted and discarded. The silt fraction was collected and dried overnight at 105˚C. The clay (%) fraction was determined by subtracting the sand (%) and silt (%) fraction from 100%.  2.3.4 Chemical extractions for inorganic and organic phosphorus Five chemical extractions were selected to quantify the inorganic fractions of phosphorus. The Aqua regia (AR) extraction was used to approximate total element concentrations, targeting soluble, labile, 22  moderately labile, and occluded P fractions. The AR extraction approximates all recoverable elements within the sample with the exception of those bound in aluminosilicate compounds (Cheng & Ma, 2001). This procedure was performed on 1 g of soil with 15 mL of 3 HCl: 1 HNO3 solution. The mixture was heated at 70˚C to dryness; this was followed by washing of residual material using 1 M HCl and filtration using Whatman 44 filter paper (3 µm particle size). The solution was then made to volume (25 mL) with 10% HNO3. Acid ammonium oxalate (AAO) was used to measure stable or P bound to amorphous Fe and Al oxides (McKeague & Day, 1966), targeting moderately labile and occluded P fractions.  The AAO solution is made from a mixture of 0.2 M ammonium oxalate and 0.2M oxalic acid at an acid pH of 3 in a dark environment.  The extraction was completed on 1 g of soil with 40 mL of AAO solution. The mixture was shaken for four hours and centrifuged for ten minutes at 2000 rpm; the supernatant was then decanted. Selective filtering using Whatman 44 was completed when the supernatant retained high suspended sediment content; this was also done for the Bray-P1, Mehlich-III, and organic P extractions.  Hydrochloric acid (HCl, 0.1 M) was used as a general acid extraction for soluble element concentrations expected under slightly acidic soil conditions in this region, targeting labile and moderately labile P fractions associated with calcium (Snape et al., 2004). Approximately 5 g of soil was reacted with 50 mL of 0.1M HCl, shaken for one hour, and left to equilibriate overnight. The samples were then washed with 0.1M HCl, filtered with Whatman 44, and made to volume (100mL) using 10% HNO3. The Bray-P1 and Mehlich-III extractions were used as agronomic measures of plant-available P which encompass the soluble and labile P fractions, particularly exchangeable fractions associated with organic matter, amorphous material, or soil colloids. Twenty milliliters of the Bray and Kurtz P-1 extraction solution (0.03 M NH4F + 0.025 M HCl) was added to 2 g of soil (Bray & Kurtz, 1945). Thirty millilitres of the Mehlich-III solution (1.5M NH4F + 0.1M EDTA, NH4NO3, CH3COOH, 10% HNO3) was added to 3 g of soil (Tran & Simard, 1993). Following the extractions, the Bray-P1 samples were shaken for 5 minutes, and the Mehlich-III samples for 15 minutes; both were centrifuged at 2000 rpm for ten minutes and the supernatant was decanted. Organic P was extracted using Bowman & Moir's (1993) 0.25 M NaOH + 0.05 M EDTA solution which targets the labile, moderately labile, and occluded P fractions associated with carbon compounds. The sodium hydroxide neutralizes organic acids, while EDTA chelates phosphates 23  associated with the organic matter. Twenty-five milliliters of solution was added to 1 g of soil, shaken for 15 seconds, and heated at 85˚C for two hours. Samples were then cooled, shaken for one hour, and centrifuged at 2000 rpm for 10 minutes and the supernatant was decanted. Soil extracts were stored at 4˚C prior to analysis. Element concentrations for all chemical extractions were analyzed using an inductively coupled plasma atomic emission spectrometer (ICP-AES) from the Earth, Ocean, and Atmospheric Sciences Department at the University of British Columbia. The dominant element wavelength with the best linear correlation was used consistently used for each extraction. To account for instrumentation error, a replicate sample was tested for every fifteen samples analyzed.    2.3.5 Calculations for degrees of phosphorus saturation (%) using the AAO and Mehlich-III extractions  Degrees of P saturation (DPS) are indices based on standardized soil tests that theoretically calculate the ratio of sorbed P to the total number of sorption sites (Beauchemin & Simard, 1999; Pellerin et al., 2006). In the AAO equation, Fe and Al are the dominant sources of sorption sites due to the formation of Fe-Al oxides in slightly acidic soils (Beauchemin & Simard, 1999). DPSox = Pox/ (Feox + Alox) (1)      Since AAO is not a standard laboratory soil test, DPS equations have been developed for the Mehlich-III extraction which measure Fe and Al as sources of sorption sites (Wang et al., 2010). M3-PSR I = PM-3/(FeM-3 + AlM-3) (2) M3-PSR II = PM-3/AlM-3 (3)    2.3.6 Soil organic carbon and total nitrogen  Determination of soil organic carbon and total nitrogen was conducted at the Analytical Chemistry Laboratory (BC Ministry of Environment) in Victoria, British Columbia. Three grams of field soil and ditch sediment samples, along with 20% randomly selected replicates were used. Initially, the samples were ground through a 100 µm mesh using a Rocklabs “ring grinder”; sample weights for 24  standards and soil samples ranged between 35-50 mg (personal communication, C. Dawson, June 26, 2014). For samples high in organic matter, samples weights ranged from 9-12 mg. Combustion by elemental analysis was conducted using a Thermo Flash 2000. Separate sub-samples were oven-dried at 105˚C to allow N and C results to be reported on a standardized 100% dry weight basis.   2.3.7 Statistical analysis All statistical analysis was completed on R statistical software version 3.0.2. (64 bit) using RStudio version 0.98.501. Prior to conducting statistical tests, normal probability plots were generated of the dataset using Microsoft Excel (Windows 7), in which sorted variables plotted against quantiles demonstrated a linear trend. Thus, correlation tests were measured using Pearson’s correlation coefficient. Due to the limited sample size, tests for significance were done using Kruskal-Wallis (KW), a non-parametric univariate analysis of variance by ranks, followed by pairwise comparisons at p < 0.05 significance. The R packages ‘pgirmess’ was used for KW pairwise comparisons.  2.4 Results  2.4.1 Soil factors influencing P solubility: pH, soil organic matter, and clay content  In the topsoil (0-20 cm) mean active pH ranged narrowly from 5.9 in the dairy and turf farms to 6.7 in the uncultivated field; the mixed site having an intermediate pH of 6.4 (Table 2.2). The range of observed pH measurements is representative of the Buckerfield/Bates soil series (Luttmerding, 1981). In the dairy, turf, and uncultivated sites pH increased marginally up to 0.5 units with depth; this was not observed at the mixed site. Exchangeable acidity expressed by pH in CaCl2 was only narrowly lower than active acidity, indicating that the contribution of labile Al3+ to alter soil pH is minimal. Overall, only pH at the uncultivated site was found to be significantly different from the three treatment sites. Salinity is not a concern as EC values were low in field soils and ditch sediments, ranging from 0.05 to 0.2 mmho/cm.  Mean soil organic matter (SOM) content was > 5% in the dairy, mixed, and uncultivated sites (Table 2.2). Consistent with the removal of turf and the top layer of soil retaining plant roots, the topsoil at the turf farm contained only 2.2 % SOM. Figure 2.6 depicts the change in SOM within the soil profile: the turf and the uncultivated sites having greater ΔSOM between the surface and the 20-50 cm depth; and the dairy and mixed sites having greater ΔSOM between the 20-50 cm and 50+ cm 25  Site Sample (depth) Sample Size (n)pH (H2O) pH (CaCl2)SOM 375˚C (%) Field soil (0-20cm) 5 5.9 5.8 6.3Field soil (20-50cm) 5 6.2 5.9 4.1Field soil (50+cm) 5 6.4 6.2 1.4Ditch sediment (0-20cm) 2 5.9 5.5 5.6Field soil (0-20cm) 5 5.9 5.6 2.2Field soil (20-50cm) 5 6.4 6.4 0.7Field soil (50+cm) 5 6.4 6.5 0.7Ditch sediment (0-20cm) 2 5.7 5.6 3.4Field soil (0-20cm) 5 6.4 6 7.9Field soil (20-50cm) 5 6.2 5.8 6.9Field soil (50+cm) 5 6.3 6 2.7Ditch sediment (0-20cm) 2 6.7 6.4 10.2Field soil (0-20cm) 5 6 5.4 7.5Field soil (20-50cm) 5 6 5.8 3.4Field soil (50+cm) 5 6.4 6 2.5DairyTurfMixedUncultivated depth. At a soil depth of 50+ cm, the mixed and the uncultivated site retained relatively high SOM, in comparison to the dairy and turf sites which have <1.5% SOM. The mean clay content of the topsoil at all study sites was generally low (≤ 20%) characteristic of this region (Figure 2.7a). Silt content widely ranged between 18-75% across sites. Due to their geographical proximity, the dairy and uncultivated sites share the loam textural class; while the mixed site, enriched in silt, has a silt loam classification. The soils are coarser than the general classification of Buckerfield Bates soils—the turf farm having surface textures of loamy sand, representative of soil at 1m depth. There is limited and slight change from the surface to 20-50 cm, with textures either remaining the same, or becoming slightly coarser or finer. At the 50+ cm depth, silt and clay content are greatly reduced in the dairy and turf farms, having texture classifications of sandy loam and sand, respectively (Figure 2.7b). The mixed and uncultivated sites are only slightly affected, with a mean reduction of 4% and 1% for clay, and an increase in silt content. Ditch sediments have comparable or higher silt and clay content than their field soil counterparts (Figure 2.7c). Significant differences in clay content were found only between the turf and mixed sites; while for silt content they were found between the turf farm and the other study sites, and between the dairy and mixed sites.           Table 2.2 Mean values of active and exchangeable pH in water and 0.01 M CaCl2 respectively, and SOM content at the study sites. 26             Figure 2.7 The proportions of sand, silt and clay, and corresponding texture class of: a) topsoil (0-20 cm) and b) soil at 50+ cm depth at the study sites; and c) ditch sediments (0-20 cm) at the treatment sites.  a) b) c) Figure 2.6 Change in mean SOM content with depth (0-50+ cm) in the soil profile at the study sites. 27  2.4.2 Accumulation of P and associated elements in field soil and ditch sediments Consistent with their utility for elemental extraction, AR (Table 2.3) extracted the greatest proportion of all targeted elements (P, Ca, Mg, Fe, and Al) followed by AAO (Table 2.4), HCl (Table 2.5), Bray and Kurtz P-1(Table 2.6), and Mehlich-III (Table 2.7). Organic P concentrations (Table 2.8) were intermediate the P concentrations for AAO and HCl. Pearson’s r was significant (p<0.01) for AR with AAO (0.90), Bray (0.78) and Mehlich-III (0.95).  The relative proportions of elements extracted was similar for AR and HCl with P<Ca~Mg<Fe~Al. For AAO, Al concentrations were lower than Fe, while Ca could not be assessed as calcium oxalate is largely insoluble. Mg concentrations for the Bray extraction were greater than Ca, and Al greater than Fe, with Fe concentrations within a similar range to Mg. Trends for Mehlich-III were Ca>Mg and Al>Fe, with Ca and Al, and Fe and Mg within a similar range of concentrations.  SOM and clay content were found to be significantly associated (p< 0.01) with total P (r = 0.91, r = 0.81), stable P (r = 0.87, r = 0.65), available P as measured by Bray (r = 0.86, r = 0.64), available P as measured by Mehlich (r = 0.86, r = 0.71), and organic P (r = 0.89, r = 0.76). Silt content also demonstrated moderate association with total P (r = 0.61), stable P (r = 0.48), Bray P (r = 0.47), Mehlich P (r = 0.49), and organic P (r = 0.55).     28  Site Sample (depth)Sample Size (n)P Ca Mg Fe Al Field soil (0-20 cm) 5 1645 ± 53 7727 ± 399 10 636 ± 135 21 885 ± 1301 23 270 ± 985Field soil (20-50 cm) 5 1275 ± 131 9162 ±  427 11 919 ± 429 25 487 ± 2337 26 012 ± 1745Field soil (50+ cm) 5 743 ± 11 6288 ± 517 11 911 ± 570 24 243 ± 1386 18 344 ± 852Ditch sediment (0-20 cm) 2 1237 ± 100 6591 ± 180 10 528 ± 262 28 449 ± 66 24 038 ± 527Field soil (0-20 cm) 5 981 ± 51 7093 ± 710 10 617 ± 152 19 985 ± 793 17 227 ± 964Field soil (20-50 cm) 5 624 ± 31 8312 ± 304 10 413 ± 114 20 794 ± 1834 19 264 ± 1193 Field soil (50+ cm) 5 572 ± 17 8725 ± 810 10 812 ± 817 22 413 ± 2229 18 261 ± 1609Ditch sediment (0-20 cm) 2 1119 ± 64 8582 ± 233 9472 ± 93 18 677 ± 2041 21 963 ± 2222Field soil (0-20 cm) 5 2144 ± 81 8309 ± 951 10 425 ± 763 29 176 ± 2956 29 069 ± 2218 Field soil (20-50 cm) 5 1902 ± 140 6854 ± 218 9656 ± 147 25 603 ± 512 26 108 ± 297Field soil (50+ cm) 5 1004 ± 58 6332 ± 617 10 434 ± 677 34 581 ± 2846 26 738 ± 1445Ditch sediment (0-20 cm) 2 2647 ± 127 8367 ± 404 10 068 ± 10 30 092 ± 1169 33 159 ± 754Field soil (0-20 cm) 5 1623 ± 131 9555 ± 320 12 307 ± 363 21 493 ± 509 28 352 ± 943 Field soil (20-50 cm) 5 1153 ± 109 8877 ± 81 12 372 ± 220 26 517 ± 2030 27 358 ± 960Field soil (50+ cm) 5 933 ± 51 9189 ± 397 12 832 ± 653 35 860 ± 1734 27 650 ± 1160Concentration (mg/kg) ± SEMDairyTurfMixedUncultivated Site Sample (depth)Sample Size (n)P Mg Fe Al Field soil (0-2  cm) 5 1302 ± 62 1614 ± 42 8459 ± 61 3710 ± 69Field soil (20-50 cm) 5 99  ± 112 1461 ± 65 8500 ± 327 3459 ± 246Field soil (50+ cm) 5 451 ± 18 1247 ± 50 7983 ± 125 2344 ± 149Ditch sediment (0-20 cm) 2 876 ± 158 1710 ± 207 11 314 ± 1157 2490 ± 759 Field soil (0-20 cm) 5 657 ± 57 1063 ± 89 8560 ± 557 2974 ± 120Field soil (20-50 cm) 5 352 ± 19 1035 ± 51 7467 ± 137 1365 ± 67 il (50+ cm) 5 287 ± 14 1066 ± 36 7080 ± 329 1268 ± 43Ditch sediment (0-20 cm) 2 740 ± 10 1100 ± 68 9164 ± 1241 16 1 ± 312Field soil (0-20 cm) 5 1220 ± 129 498 ± 67 3874 ± 464 3305 ± 469Field soil (20-50 cm) 5 1209 ± 140 520 ± 29 4359 ± 209 4018 ± 380Field soil (50+ cm) 5 446 ± 31 400 ± 6 4370 ± 203 2218 ± 82Ditch sediment (0-20 cm) 2 1602 ± 261 699 ± 198 6016 ± 1796 2579 ± 675 Field soil (0-20 cm) 5 1364 ± 131 1880 ± 22 8744 ± 176 4483 ± 83 Field soil (20-50 cm) 5 719 ± 95 932 ± 240 5218 ± 1184 2814 ± 466Field soil (50+ cm) 5 442 ± 33 538 ± 30 3156 ± 142 1809 ± 201Concentration (mg/kg) ± SEMDairyTurfMixedUncultivated  Table 2.4 Mean values of the concentration and standard error of the mean for stable phosphorus and associated elements Mg, Fe, and Al from the AAO extraction.  Table 2.3 Mean values of the concentration and standard error of the mean (SEM) for total phosphorus and associated elements Ca, Mg, Fe, and Al from the AR extraction. 29  Site Sample (depth)Sample Size (n)P Ca Mg Fe Al Field soil (0-20 cm) 5 287 ± 13 1889 ± 119 1250 ± 99 1502 ± 76 2439 ± 53Field soil (20-50 cm) 5 221 ± 17 1590 ± 102 1145 ± 26 1547 ± 96 2247 ± 93Field soil (50+ cm) 5 225 ± 3 1120 ± 37 1271 ± 14 1910 ± 55 1657 ± 54Ditch sediment (0-20 cm) 2 220 ± 47 1550 ± 36 1291 ± 98 2945 ± 517 2023 ± 78Field soil (0-20 cm) 5 332 ± 26 1462 ± 462 1487 ± 187 2935 ± 328 1558 ± 152 Field soil (20-50 cm) 5 263 ± 5 1007 ± 57 1264 ± 45 2335 ± 47 1258 ± 27Field soil (50+ cm) 5 266 ± 23 1258 ± 279 1349 ± 225 2711 ± 667 1217 ± 153 Ditch sediment (0-20 cm) 2 406 ± 147 1461 ± 290 1544 ± 288 3381 ± 788 1552 ± 109Field soil (0-20 cm) 5 318 ± 16 2054 ± 96 823 ± 45 1025 ± 89 2822 ± 183Field soil (20-50 cm) 5 277 ± 28 2001 ± 52 802 ± 33 1085 ± 95 2909 ± 142Field soil (50+ cm) 5 146 ± 26 1689 ± 283 594 ± 174 1219 ± 432 2035 ± 219 Ditch sediment (0-20 cm) 2 318 ± 10 2373 ± 34 570 ± 16 495 ± 16 2186 ± 32Field soil (0-20 cm) 5 310 ± 26 1636 ± 46 1459 ± 24 1733 ± 26 2330 ± 18Field soil (20-50 cm) 5 262 ± 19 1322 ± 36 1287 ± 33 1644 ± 44 2360 ± 54Field soil (50+ cm) 5 243 ± 10 1364 ± 146 1204 ± 119 1684 ± 93 2232 ± 83UncultivatedMixedTurfDairy Concentration (mg/kg) ± SEMSite Sample (depth)Sample Size (n)P Ca Mg Fe Al Field soil (0-2  cm) 5 78 ± 2 91 ± 14 198 11 94  6 97 43Field soil (20-50 cm) 5 50 ± 5 87 ± 16 161 ± 10 110 ± 11 512 ± 45 Field soil (50+ c ) 5 28 ± 1 28 ± 8 161 ± 7 133 ± 12 533 ± 34Ditch sediment (0-20 cm) 2 32 ± 2 114 ± 21 307 ± 9 201 ± 44 552 ± 35 Field soil (0-2  cm) 5 24  2 73 ± 10 118  28 161  0 4 7 ± 41 Field soil (20-50 cm) 5 19 ± 2 65 ± 8 82 ± 10 207 ± 27 351 ± 57Field soil (50+ c ) 5 16 ± 1 40 ± 6 76 ± 33 215 ± 46 272 ± 33Ditch sediment (0-20 cm) 2 28 ± 1 148 ± 67 148 ± 67 87 ± 12 409 ± 9 Field soil (0-20 cm) 5 60 ± 3 86 ± 1 91 ± 4 54 ± 4 501 ± 29Field soil (20-50 cm) 5 58 ± 8 108 ± 5 86 ± 3 62 ± 4 574 ± 26Field soil (50+ cm) 5 24 ± 2 82 ± 8 92 ± 15 175 ± 47 565 ± 50Ditch sediment (0-20 cm) 2 73 ± 1 89 ± 4 128 ± 2 50 ± 0.3 408 ± 5 Field soil (0-20 cm) 5 60 ± 6 119 ± 2 218 ± 4 131 ± 19 666 ± 34Field soil (20-50 cm) 5 56 ± 17 86 ± 7 141 ± 19 127 ± 25  662 ± 33Field soil (50+ cm) 5 24 ± 3 58 ± 3 128 ± 13 99 ± 9 660 ± 30 UncultivatedConcentration (mg/kg) ± SEMDairyTurfMixed  Table 2.5 Mean values of the concentration and standard error of the mean for labile phosphorus and associated elements Ca, Mg, Fe, and Al from the 0.1M HCl extraction.  Table 2.6 Mean values of the concentration and standard error of the mean for available phosphorus and associated elements Ca, Mg, Fe, and Al from the Bray and Kurtz P-1 extraction.   30  Site Sample (depth)Sample Size (n)P Ca Mg Fe Al Field soil (0-20 cm) 5 111 ± 11 1296 ± 117 394 ± 43 259 ± 20 871 ± 90Field soil (20-50 cm) 5 65 ± 9 1378 ± 162 301 ± 17 293 ± 12 1116 ± 34Field soil (50+ cm) 5 19 ± 1 716 ± 42 242 ± 17 286 ± 16 1082 ± 65 Ditch sediment (0-20 cm) 2 28 ± 2 1049 ± 166 660 ± 97 431 ± 74 817 ± 259Field soil (0-20 cm) 5 33 ± 3 684 ± 179 254 ± 19 377 ± 16 686 ± 53Field soil (20-50 cm) 5 17 ± 1 592 ± 119 198 ± 11 345 ± 18 626 ± 29 Field soil (50+ cm) 5 23 ± 0.3 498 ± 86 167 ± 14 352 ± 26 610 ± 48Ditch sediment (0-20 cm) 2 38 ± 10 640 ± 18 262 ± 66 306 ± 17 650 ± 175Field soil (0-20 cm) 5 133 ± 12 1860 ± 58 371 ± 21 311 ± 16 1104 ± 139 Field soil (20-50 cm) 5 109 ± 19 1676 ± 79 310 ± 35 273 ± 21 1073 ± 146 Field soil (50+ cm) 5 27 ± 2 1265 ± 94 245 ± 4 328 ± 13 1278 ± 44 Ditch sediment (0-20 cm) 2 160 ± 0.1 2397 ± 30 591 ± 35 369 ± 18 957 ± 26Field soil (0-20 cm) 5 89 ± 12 1311 ± 71 466 ± 23 356 ± 11 1213 ± 15Field soil (20-50 cm) 5 56 ± 11 836 ± 84 350 ± 32 294 ± 25 1253 ± 145Field soil (50+ cm) 5 34 ± 5 658 ± 32 398 ± 43 296 ± 15 1297 ± 152 UncultivatedConcentration (mg/kg) ± SEMMixedTurfDairy Concentration (mg/kg) ± SEMSite Sample (depth)Sample Size (n)PField soil (0-20 cm) 5 811 ± 43Field soil (20-50 cm) 5 517 ± 70 Field soil (50+ cm) 5 136 ± 10Ditch sediment (0-20 cm) 2 347 ± 37Field soil (0-20 cm) 5 309 ± 53Field soil (20-50 m) 5 105 ± 21Field soil (50+ cm) 5 59 ± 16Ditch sediment (0-20 cm) 2 425 ± 98Field soil (0-20 cm) 5 1017 ± 46Field soil (20-50 cm) 5 829 ± 82 Field soil (50+ cm) 5 271 ± 30Ditch sediment (0-20 cm) 2 1234 ± 6Field soil (0-20 cm) 5 730 ± 95Field soil (20-50 cm) 5 467 ± 82Field soil (50+ cm) 5 275 ± 58MixedUncultivatedTurfDairy Table 2.7 Mean values of the concentration and standard error of the mean for available phosphorus and associated elements Ca, Mg, Fe, and Al from the Mehlich-III extraction.     Table 2.8 Mean values of the concentration and standard error of the mean for organic phosphorus from Bowman and Moir’s (1993) NaOH and EDTA extraction.  31  Table 2.9 Summary of minimum and maximum concentrations for total, stable, labile, plant-available and organic P across the study sites.  Phosphorus  Fractions Extraction Minimum Concentration (mg P/ kg soil): Site and sample type   Maximum Concentration (mg P/ kg soil): Site and sample type  Total  (Figure 2.8a) Aqua regia 555: turf farm; field soil at 50+ cm depth 2774: mixed site; surface (0-20 cm) ditch sediment  Stable (Figure 2.8b) Acid ammonium oxalate 273: turf farm; field soil at 50+ cm depth 1863: mixed site; surface (0-20 cm) ditch sediment  Labile (Figure 2.8c) 0.1 M HCl 120: mixed site; field soil at 50+ cm depth  553: turf farm; surface  (0-20 cm) ditch sediment Plant-available (Figure 2.8d) Bray-Kurtz P1 16: turf farm; field soil at 50+ cm depth 80: dairy farm; topsoil  (0-20 cm)  Plant-available (Figure 2.8e)  Mehlich-III 16: turf farm; field soil at 20-50 cm depth  160: mixed site; surface  (0-20 cm) ditch sediment Organic (Figure 2.8f) 0.25 M NaOH +  0.05 M EDTA 43: turf farm; field soil at 50+ cm depth 1240: mixed site; surface (0-20 cm) ditch sediment  The range of mean concentrations observed for P extractions are summarized in Table 2.9. For total, Mehlich, and organic P the mixed site had the highest concentration of P, overall and the turf farm the lowest (Figure 2.8a, e, and f). The dairy and uncultivated sites had similar concentrations in field soils. The decline in P concentrations is greatest between the surface and the 20-50 cm depth for the turf and uncultivated sites. Conversely, for the dairy and mixed sites, the decline is greatest between the 20-50 cm and 50+ cm depth. For ditch sediments, P concentrations at the turf and mixed sites tended to be elevated compared to corresponding topsoil samples; while at the dairy farm, P concentrations were consistently lower than in topsoil. Total P correlated best with Al (r= 0.64, p<0.01), and Mehlich P with Ca (r = 0.80, p<0.01) and Mg (r = 0.54, p<0.01).  For stable and Bray P, concentrations were notably elevated in fields soils at the uncultivated site, being above or comparable to the dairy or mixed sites (Figure 2.8b and d). Stable P was strongly associated with Al (r = 0.83, p<0.01). Additionally, Bray P correlated moderately with Ca (r = 0.50, p<0.01).Trends for labile P were different than the other extractions. Notably, the turf farm had the greatest P concentration in both field soil and ditch sediments (Figure 2.8c). For this extraction, P had weak relationships with Mg (r = 0.32, p<0.01) and Fe (r = 0.32, p < 0.05).  Kruskal-Wallis analysis revealed total, stable, and organic P to be significantly (p<0.01) different in the turf farm compared to study sites of high organic matter. For Mehlich P, the turf farm was distinct only from the mixed and uncultivated sites.  32  Summaries of P concentrations within site were done by normalizing P extraction values to total P values (Figure 2.9a-d). Near the soil surface, stable P accounts for 55-80% of total P, while organic P accounts for 10-50%. For all sites, stable and organic P proportions decline with depth in field soil, with proportions in ditch sediments being similar to, or above, those of topsoil. The three treatment sites show a general decline of ~16% stable P from the soil surface to 50+ cm depth; in the uncultivated site, this decline is more pronounced at ~37%. The decline of organic P from surface to 50+cm depth, is similar between the dairy and turf farms at ~25%, and between the mixed and uncultivated sites at ~16%.  Proportions of labile P tended to increase with depth in the field soil, with the exception of the mixed site which was stable at ~15%. Proportions of labile P were greatest at the turf farm, followed by the dairy and uncultivated sites, and least of all at the mixed site. Across all sites, available P was low with slight differences between Mehlich and Bray only at the dairy and mixed sites, the two sites of greatest organic matter content. Proportions range between 2-7%, the greatest decline occurring between the 20-50 cm and 50+ cm depth. Ratios of P and P-associated elements were indicative of natural and anthropogenic influences on the sites. The most distinct relationships include:   Mg/Ca: Due to the serpentinitic contributions, the study sites were enriched in Mg and expressed a ratio of 1.1-1.9.  From 0 to 20-50 cm depth in field soil, the ratio is elevated at the turf farm and similar across the other three sites. The trend for Mg/Ca was dairy > mixed > uncultivated > turf for 50+ cm depth of field soil and for ditch sediments. Natural enrichment is likely determined by proximity to the Sumas River; however, at the turf farm enrichment may be due to the application of Mg-based fertilizer or dolomite.   Fe/Al: Field soil at the surface and 50+cm depth of the three treatment sites tended to be enriched in Fe, while the uncultivated site and ditch sediments of the mixed site and turf farm had Fe/Al < 1. At 20-50 cm depth, Fe/Al is similar across all sites.   P/Ca+Mg and P/Fe+Al: The mixed site had the greatest P enrichment relative to elements responsible for P sorption; this difference is most evident from the surface to 20-50 cm depth in soil and in the ditch sediments.   P/Ca+Mg+Fe+Al: The mixed site still had the greatest P enrichment when all sources of sorption were accounted for; however this difference was only distinct in field soil at 20-50 cm depth, and in ditch sediments.  33  Figure 2.8 Mean concentrations of a) total P, b) stable P, c) labile P, d) available P (Bray), e) available P (Mehlich), and f) organic P at the study sites in field soil at depths of 0-50+ cm, and in surface (0-20 cm) ditch sediments. Error bars are standard error of the mean.   FS- Field Soil DS- Ditch Sediment   a) b) d) e) f) c) 34  Figure 2.9 Proportions of inorganic (stable, labile, and available) and organic P in field soil and ditch sediments compared to total P at the a) dairy farm, b) turf farm, c) mixed site, and d) uncultivated site.  a) b)c) d)a)35  2.4.3 Environmental P saturation as measured by the AAO and Mehlich-III extractions  The commonly used critical threshold value for DPSox is 25%, as defined on the basis of P solubility of 0.1mg/L of orthophosphate in a laboratory (Breeuwsma & Reijerink, 1992).Significant P solubility has been found to occur at lower P saturation degrees of 23% (Yuan & Lavkulich, 1995) and even 15%  (De Smet et al.,1996). For the study sites DPSox was found to range from 2 to 13% (Table 2.10). All samples from the dairy, turf and uncultivated sites were below any of the aforementioned critical thresholds. However, the mixed site demonstrates potential for losses of P with values of 12% in topsoil and 13% in ditch sediments, nearing the critical threshold proposed by De Smet et al. (1996).  According to Maguire & Sims (2002), the 25% DPSox value is equivalent to 6% for M3-PSR I and 7% for M3-PSR II. Sims et al. (2002) estimated the threshold M3-PSR I ratio to be 9% while Khiari et al. (2000) estimated M3-PSR II closer to 10%. For Quebec soils, Giroux & Tran (1996) proposed a critical value of 20% for both M3-PSR I and II. Overall, M3-PSR I values range from 1-11%, while M3-PSR II values range from 2-19% (Table 2.10). Values for M3-PSR I in topsoil at the dairy and mixed sites were above the lowest critical threshold. Values for M3-PSR II in field soil (0 to 20-50 cm) at the dairy and mixed sites, and the ditch sediments at the mixed sites were above the middle to high range of the critical threshold. The ditch sediment and topsoil at the turf farm along with the topsoil at the uncultivated site have M3-PSR II values near the low range of the critical threshold. Correlation tests between saturation indices and soil factors indicated no significant relationship with pH. Significant linear relationships (p < 0.01) were found between SOM and clay content with DPSox and M3-PSR I (Figure 2.10). For M3-PSR II, similar r-values (p < 0.01) were calculated: 0.80 for SOM and 0.64 for clay content. Silt content was moderately to weakly associated with DPSox (r = 0.62), M3-PSR I (0.44), and M3-PSR II (0.39).   36  AAO DPSoxM3-PSR I M3-PSR II M3-PSR IIISite Sample (depth)Sample Size (n)Pox /                (Alox + Feox)PM-3/ AlM-3 + FeM-3PM-3/AlM-3PM-III/CaM-III Field soil (0-20cm) 5 7 9 15 7Field soil (20-50cm) 5 5 4 7 4Field soil (50+cm) 5 3 1 2 2Ditch sediment (0-20cm) 2 4 2 4 2Field soil (0-20cm) 5 4 3 6 5Field soil (20-50cm) 5 2 1 3 2Field soil (50+cm) 5 2 2 5 4Ditch sediment (0-20cm) 2 4 3 7 5Field soil (0-20cm) 5 12 9 14 6Field soil (20-50cm) 5 11 8 12 5Field soil (50+cm) 5 5 2 2 2Ditch sediment (0-20cm) 2 13 11 19 5Field soil (0-20cm) 5 7 5 8 5Field soil (20-50cm) 5 7 3 5 5Field soil (50+cm) 5 6 2 3 4TurfMixedUncultivatedDairy Degree of P saturation (DPS) % Mehlich III  Table 2.10 Indices for environmental P saturation measured using the AAO (DPSox) and Mehlich-III (M3-PSR I and M3-PSR II) extractions.  37  a) b) c) d)   Figure 2.10 Linear relationships between selected soil factors and AAO and Mehlich-III P saturation indices: soil organic matter content vs. a) DPSox and b) M3-PSR I; clay content vs. c) DPSox and d) M3-PSR I. 38  C N PField soil (0-20cm) 5 3.2 ± 0.2 0.29 ± 0.02 0.081 ± 0.004 11 40 3.7 40:4:1Field soil (20-50cm) 5 2.0 ± 0.3 0.18 ±  0.03 0.052 ± 0.007 11 38 3.4 38:3:1Field soil (50+cm) 5 0.48 ±  0.03 0.05 ±  0.002 0.014 ± 0.001 9.6 35 3.7 35:4:1Ditch sediment (0-20cm) 2 2.5 ± 0.06 0.23 ± 0.004 0.035 ± 0.004 11 73 6.6 73:7:1Field soil (0-20cm) 4 0.72 ± 0.5 0.072 ± 0.01 0.031 ± 0.005 10 20 3.5 20:3:1Field soil (20-50cm) 5 0.31 ± 0.1 0.031 ± 0.007 0.010 ± 0.002 8.8 26 4 26:4:1Field soil (50+cm) 3 0.15 ± 0.8 0.022 ± 0.05 0.0059 ± 0.001 6.7 38 3.6 38:4:1Ditch sediment (0-20cm) 2 1.7 ± 0.02 0.16 ±  0.0007 0.042 ± 0.01 10 41 4 41:4:1Field soil (0-20cm) 5 3.9 ± 0.2 0.36 ± 0.02 0.1 ± 0.004 11 38 2 38:2:1Field soil (20-50cm) 5 3.3 ± 0.3 0.33 ± 0. 04 0.083 ± 0.008 9.9 39 3 39:3:1Field soil (50+cm) 5 0.98 ± 0.1 0.097 ± 0.007 0.027 ± 0.003 9.9 36 5.6 36:6:1Ditch sediment (0-20cm) 2 4.9 ± 0.07 0.50 ± 0.012 0.12 ± 0.0006 9.8 40 3.9 40:4:1Field soil (0-20cm) 5 5.4 ± 0.7 0.51 ± 0.06 0.073 ± 0.01 10 87 8.1 87:8:1Field soil (20-50cm) 5 1.8 ± 0.3 0.17 ± 0.02 0.047 ± 0.008 10 47 4.5 47:4:1Field soil (50+cm) 5 0.99 ±  0.1 0.094 ±  0.01 0.027 ± 0.004 10 39 3.7 39:4:1C:N:PN:P C:P C:N Concentration (%)  ± SEMSample (depth)Sample Size (n) SiteMixedDairyTurfUncultivated2.4.4 Carbon, nitrogen, and organic phosphorus concentrations and nutrient ratios  Across all sites, concentrations or organic carbon, total nitrogen, and organic P parallel the change in SOM with depth (Table 2.11). C:N ratio  (~10) narrowly varied with depth and across sites, with the exception of soil at 50+ cm depth at the turf farm which had a ratio of 6.7.  C:P ratio was ~40 in field soil at mixed and dairy sites, ~30 at the turf farm, and ranged from 87-37 with depth at the uncultivated site.  The C:P ratio narrowed with depth at all sites except at the turf farm where it increased. The N:P ratio in field soil at the dairy and turf farms were comparable, ranging from 3.5-4 with depth. However, the change in N:P ratio with depth was greater at the mixed site, increasing from 2-5.6.  The uncultivated site exhibited the greatest variability in SOM characteristics: a doubling of the C:P and N:P ratio of the treatment sites. Notably, only ditch sediments at the dairy farm had distinct enrichment of carbon and nitrogen.  C:N:P ratios were ~ 40:4:1 for field soil (0-50+ cm) and ditch sediments at all sites of high SOM content (dairy, mixed, and uncultivated), with the exception of variability of topsoil at the uncultivated site, and enrichment of N at the 50+ cm depth at the mixed site. This ratio is closer to an average of ~ 30:4:1 across depth at the turf farm. Ditch sediments at the dairy farm and topsoil at the uncultivated site were distinct having C:N:P ratios of 73:7:1 and 87:8:1, respectively.     Table 2.11 Concentrations and ratios of organic carbon, total nitrogen, and organic phosphorus at the study sites.  39  2.5 Discussion 2.5.1 Status of soil phosphorus in dairy, turf, mixed, and uncultivated sites Differences in P concentrations among sites were consistent for total, Mehlich and organic P, with the mixed site having largest P concentrations, the turf farm the least, and the dairy and uncultivated sites being intermediate. This may be attributed to the application of both manure and fertilizer at the mixed site, which can have an additive effect as compared to sites with singular nutrient additions (Eghball et al.1996). Further, nutrient management at the turf farm may be done with more precision as uniformity and efficiency of turf production is a priority to allow for multiple harvests in a year, thus limiting the likelihood of P accumulation in the soil. The narrow difference in P concentrations between the dairy and uncultivated sites may result from geographic proximity as P loading from the dairy field and the Sumas River can occur during flooding periods.  Furthermore, it is likely that there has been prior nutrient loading at the uncultivated site, which due to limited historical records has not been accounted for.   Concentrations and proportions of P and associated elements from all extractions, except the organic P fraction, were within the range of values reported by Bauge et al. (2013a and b). Mehlich P values at all the study sites were lower than values reported by Kowalenko et al. (2007) for Buckerfield soils which measured approximately 174 mg/kg.  Measurements of total, stable, Bray, Mehlich, and organic P affirm findings from previous studies that P is vertically mobile within the soil matrix, particularly down to the 20-50 cm depth Eghball et al., 1996; Pant et al., 2002). At a field soil depth of 50+ cm, the effect of accumulation at the treatment sites was minimal. Despite limited sampling, inference of lateral mobility can be made from accumulation of P in ditch sediments at the turf and mixed sites, which exhibited concentrations comparable to or above topsoil, and greater P saturation as measured by the DPS indices (Dunne et al., 2007; Liu et al., 2013; Sharpley et al., 2007).  P concentrations in ditch sediments at the dairy farm were consistently below values in topsoil, which is unexpected since the ditch had not been cleaned in five years. As such, this may suggest that vegetation growing in the ditch has removed some of the accumulated P or has contributed to stabilization of P in humic forms less susceptible to degradation and mobility.  The low variability of labile P with depth and across sites may be indicative of similar precipitation of calcium phosphates despite nutrient management (Devau et al., 2011; Ulén & Snäll, 2007). However, it is notable that proportions of labile P increased with depth except at the mixed site, and was greatest at the turf farm. Results suggest that at the study sites total P can be broadly estimated by 40  stable, available, and organic P fractions. Stable P was differentially extracted in field soil at the uncultivated site, which suggests that a larger proportion of P is bound to Fe and Al oxides when no nutrient management is applied. Limited conclusions may be drawn from Bray P which was consistent with stable P, despite extracting P bound to Ca rather than Al. Overall decline in organic P with depth (0-50+ cm) was greater for the dairy and turf farms compared to the mixed and uncultivated sites, which suggests greater retention overall at the latter two sites. Proportions of available P were low overall (<10%) and did not change considerably with depth. Contrary to the enrichment of Fe in serpentinitic soils (Hashimoto & Watanabe, 2014), total and stable P were associated solely with Al, while labile and available P were predominantly associated with Ca and moderately with Mg. In consideration of the circumneutral pH range found at these sites and the abundance of aluminosilicate minerals (Luttmerding, 1981), it is consistent with conventional P chemistry that Al dictate long-term P solubility, particularly through P sorption to Al oxides and hydroxides (Sharpley, 1995; Smeck, 1985). Ulen and Snall (2007) postulated that exchangeable Ca2+ and Mg2+ ions contribute to increased sorption by linking P to negatively charged clay minerals. Furthermore, Mg2+ is known to kinetically hinder nucleation and subsequent growth of HAP by competing for structural sites with Ca2+ (Cao & Harris, 2008). Hence, accumulation of Mg at the study sites may contribute to increased P availability through prevention of the formation of more stable Ca-phosphates. Elemental ratios were indicative of relative enrichment of Ca, Mg, Fe and Al which can influence P association to immediate or long-term reserves. Compared to the dairy and mixed sites, Mg was less enriched at the turf farm, while Fe and Mg were less enriched at the uncultivated site; this may suggest inherently lower contributions of serpentinitic material to these sites.  It is also possible that at the turf farm, Mg-amended soil was preferentially removed during turf harvesting. The enrichment of P to sorption sites in the soil was most distinct in field soil (0 to 20-50 cm) and ditch sediments at the mixed site. This is consistent with DPSox measurements which suggest that losses from this site are likely. This builds upon previous work that found P to be mainly mobile at the soil surface (Kowalenko et al., 2007).  Nonetheless there is potential for underestimation of DPSox values as Al predominantly dictates P solubility in these soils, yet the influence of Fe is given equal weight in the equation. The Mehlich sorption indices, M3-PSR I and M3-PSR II, suggest variable likelihood of P loss from the topsoil of all study sites and from ditch sediments at the fertilizer and mixed sites. By comparison the Mehlich sorption indices may be more sensitive than the DPSox, and may have greater utility given its established correlation with the Kelowna extraction typically performed in the Sumas.   41  2.5.2 Influence of selected soil factors on P accumulation and saturation Of the soil factors measured, pH was the least variable across sites and did not seem to distinguish P dynamics at the uncultivated site from the treatment sites. Despite the relatively low clay content in these soils, total, stable, available, and organic P along with all DPS indices were positively associated with clay content. This indicates that accumulation and desorption of P was more likely with increased clay content due to incorporation in clay minerals and increased proportion of sorption sites (Lair et al., 2009; Tiessen et al., 1984). Silt content, although high at all sites except the turf farm, only played a secondary role compared to clay content, with regards to P sorption.  It is definitive that soil organic matter content had the largest influence on P dynamics at these sites. This is evidenced by the fact that significant differences in total, stable, Mehlich, and organic P were only distinctly different at the turf farm compared to study sites of high organic matter content. Additionally, SOM increased potential for P desorption indicating that it is a significant contributor of both P and sorption sites (Smith et al., 1998; Yu et al., 2013).  There was no difference between the C:N ratios of the uncultivated and treatment sites which indicates overall enriched nitrogen content likely to facilitate mineralization (Parton et al., 1988). The C:N ratio remained constant with depth at the dairy, mixed and uncultivated sites despite changes in SOM content, indicating coupled cycling of carbon and nitrogen, which may be a product of similar inputs of organic matter to soils by primary producers (Cleveland & Liptzin, 2007; Groppo et al., 2015). The decrease in the C:N ratio at the turf farm is reflective of the overall low SOM content at depth as no manure is being added. The C:P ratio ranged from 30-40 at the treatment sites and is sustained with depth in the soil, indicative of accumulation of the active and highly available P fraction in SOM (Parton et al., 1988). Alternatively, topsoil at the uncultivated site and ditch sediments at the dairy farm had C:P ratios closer to  90, suggesting some accumulation of intermediate forms of P which are mineralized at a slower rate. At the study sites N:P ratios ranged from 2-8, which was within expected values for manure, but below typical crop requirements of 7-11,  suggesting enrichment of P particularly at the mixed site (Smith et al., 1998). It seems that there is minimal difference in the quality of organic matter between the dairy and mixed sites, and the uncultivated site.  In a comprehensive review of nutrient ratios from uncultivated sites, Cleveland and Litzpin (2007) found that despite variability in soil medium, element ratios, and microbial community structure, C:N:P ratios were highly constrained to 60:7:1 in soil microbial biomass, which is comparable to the C:N:P ratio found in ditch sediments at the dairy site, and topsoil at the uncultivated site. 42  Alternatively, Heuck et al. (2015) found molar microbial C:N:P to be 39:4:1 in laboratory studies, which is similar to the C:N:P ratio found at study sites of high SOM content. This indicates that at these C:N:P ratios, the nutritional needs of microbial populations are sufficiently met.  Hence, it can be inferred that P is effectively being released from cycling of organic pools at the dairy, mixed and uncultivated sites, perhaps more so than weathering of inorganic P from minerals.    43  3. Characterizing inorganic and organic phosphorus sources, and evaluating surface properties relevant for phosphorus retention   3.1 Synopsis The use of non-destructive analytical techniques, such as x-ray diffraction (XRD) and nuclear magnetic resonance (NMR) spectroscopy, may be used to capture the diversity of inorganic and organic compounds that contribute to P accumulation. The complexity, molecular structure, and intermolecular force binding these compounds ultimately determines their capacity to govern P supply and adsorption in the soil (Berg & Joern, 2006; Lookman et al., 1994). The solubility of P minerals and the formation of inorganic surface complexes on various aluminosilicates have been investigated to a much greater degree than the kinetics and adsorption of organic P by soils (Shang et al., 1990). Until recently characterization of organic P has been limited by lack of precision in methods of analysis. Optimization of solid-state and solution NMR spectroscopy in the application of soil analysis has enhanced the resolution of detected inorganic and organic P compounds (Cade-Menun, 2005; C. A. Shand et al., 1999). Organic P is of particular importance to intensively managed sites, such as the Sumas Prairie, because a large part of P loss during cultivation originates from the organic P fraction (Smith et al., 1998). Since adsorption-desorption reactions are believed to control soluble P concentrations, soil properties including surface charge, adsorption capacity, and bonding energy can indicate the likelihood for P retention and release (Smil, 2000; Syers et al., 2008). Currently, there has not been an assessment of P sources in British Columbia soils. As well, Kowalenko et al. (2007) found that many coastal BC soils do not adsorb P according to traditional theory and thus require alternate analysis methods to quantify binding mechanisms. The purpose of this work was first, to characterize P compounds using XRD and NMR, and secondly, to evaluate surface properties using zero point of charge and linear, Langmuir, and Freundlich equilibrium models. The complement of qualitative and quantitative analysis aided the understanding of how nutrient management has impacted the dynamics of P sorption in the Sumas Prairie.   3.2 Introduction The application of solid-state and solution NMR spectroscopy to environmental samples is a growing area of research. Solid-state NMR spectroscopy can determine short-range structural information, allowing for investigation of phosphate sorption mechanisms at the mineral-water interface (Li et al., 44  2013). Several studies have applied this technique to investigate inorganic phosphate and organophosphate sorption on aluminum oxyhydroxides and clays (Feuillie et al., 2013; Lookman et al., 1994). Specifically, solid-state NMR can be used to distinguish surface adsorption and surface precipitation of aluminum phosphate based on chemical shift. For solution NMR spectroscopy, biological P compounds visible in the spectrum include phosphonates, phosphate, phosphate monoesters, phosphate diesters, pyrophosphate, and polyphosphate (Cade-Menun, 2005). This technique is typically conducted using a soil extractant to achieve a high quality spectra, which can result in chemical alteration of some soil P compounds (Cade-Menun & Preston, 1996).  Additional challenges of NMR analysis include the heterogeneous physical and chemical properties of soils, relatively low soil P concentrations, and the association of P with paramagnetic ions (Cade-Menun, 2005). Cade-Menun et al. (2005) found with sufficient delay times of 1-2s, the effect of Fe and Mn on differential loss of peaks was sufficiently minimized.  At the study sites, Al was found to be the dominant element controlling long-term P release into soil solution, most likely in the form of amorphous Al oxides and hydroxides. Bauge et al. (2014) found the dominant mineralogy at the dairy and turf study sites to be characterized by quartz, chrysotile, mica, apatite, and struvite which suggests weak P sorption capacity by the crystalline fractions in the soil. Non-crystalline aluminosilicates (allophones), oxides, and hydroxides of Fe and Al, and to a lesser extent the edges of layer silicate clays, provide surface sites for chemisorption  (McBride, 1994). Adsorptive sites available to the soil solution are comprised of valence-unsatisfied OH-or H2O ligand bound to a metal ion, where surface charges are susceptible to changes in pH. Li et al. (2013) found that within a pH range of 4-10, bidentate binuclear inner-sphere surface complexes were the dominant mechanisms for P sorption on aluminum hydroxides. P rich-materials have been found to exist as discrete particles as well as coatings on aluminosilicates (Pierzynski et al., 1990). During the period between fertilizer application and dilution of soil water, a variety of precipitates can form under both high and low soluble P concentrations (Pierzynski et al., 1990). These precipitates include CaHPO4• 2H2O or MgHPO4 • 3H20; however, their retention in the soil is dependent on reactivity with Al hydroxides, the soil moisture content, and accumulation from multiple precipitation events (Pierzynski et al., 1990; The Clay Minerals Society, 1987).  Soil organic matter was found to be the most important soil factor determining the accumulation and sorption of P at the study sites. Organic P compounds can exist in stable forms as inositol phosphates and phosphonates, and in active forms as orthophosphate diesters, labile orthophosphate monoesters, and organic polyphosphates (Shen et al., 2011). In podzolic soils of Vancouver Island, British Columbia Po was present predominantly as phosphate monoesters adsorped on soil colloids, while 45  phosphate diesters were found throughout the soil profile (Cade-Menun et al., 2000). Further, Hashimoto and Watanabe (2014) found that up to 73% of P in serpentinitic forest soils occurred predominantly as recalcitrant orthophosphate monoesters and to a lesser extent as orthophosphate diesters. Pyrophosphate contributions were minor, comprising only 3-13% of soil P. Borggaard et al. (2005) summarized that organic anions can affect phosphate sorption in the following ways: competition for adsorption sites; dissolution of adsorbents; change of the surface charge of adsorbents; creation of new adsorption sites through adsorption of metal ions such as Al3+ and Fe3+; and retardation of crystal growth of poorly ordered Al and Fe oxides. Berg & Joern (2006) found that organic sources decreased the soil’s capacity to sorb Pi by different amounts in the following order myo-inositol hexaphosphate (IP6) < ATP < glucose-6-phosphate (G6P), with G6P having greater affinity than Pi for sorption sites. Sorption competition also occurs between organic P sources resulting in decreased sorption of G6P and ATP, but having minimal impact on IP6. Alternatively, adsorption on Al precipitate at an acidic pH was found to be greater for Pi and IP6 than for inositol monophosphate and G6P which may be related to the kinetics of the reaction (Shang et al., 1990). Adsorption of orthophosphate occurred more rapidly than for the organic phosphates, where reactions were more temperature dependent, thus requiring higher activation energy. In Australian soils, Mcbeath et al. (2004) found that pyrophosphate was preferentially sorbed compared to orthophosphate.  Mineral and organic soil components can 1) strongly interact with P, indicative of chemisorption, in which a co-valent or short-range electrostatic bonds form between the molecule and the surface, or 2) weakly interact with P, indicative of physical adsorption, in which the bonding interaction is not very energetic (McBride, 1994).  The most commonly used mathematical models to fit adsorption data are Langmuir and Freundlich equations (McBride, 1994). There is no mechanistic significance to the parameters measured by these equations, but they have practical utility in comparing P retention of different soils. The models are most useful when there is evidence that adsorption and not precipitation is controlling P fixation. It is proposed that a sizable fraction of P adsorbed by soil is rapidly converted to non-labile forms, and that phosphate desorption is slow (McBride, 1994). Also, it is likely that chemisorption and precipitation are occurring simultaneously, with the latter reaction accounting for much of the non-labile fraction. As P concentrations are lowered, sparingly soluble phosphates will dissolve until equilibrium is reached between the adsorption complex and the solubility of the least stable P compound present (Lindsay et al., 1989).   ZPC (often interchangeably used with isoelectric point) describes the pH at which there is zero net charge due to the equal amount of H+ and OH- adsorped on variable charge components (Sakurai et 46  al., 1989; Sollins et al., 1988). ZPC is highly correlated with pH, Fe and Al oxides extracted using oxalate, and specific surface area; as well, the magnitude of zeta potential is correlated with the amount of exchangeable Ca and Al (Sakurai et al., 1989). Since orthophosphate is an anion and agricultural soils tend to be negatively charged and have variable charges due to high organic matter content, steady state is rarely achieved posing a challenge for modelling of solute transport through equilibrium models. Thus, an advantage of concurrently measuring ZPC is to understand the sorption properties of the soil surface across a range of pH values. In this way ZPC can act as an indicator of the capacity for P to become mobile.  3.3 Methods 3.3.1 Sample collection The samples used for analysis were the same as collected in Chapter 2 (Section 2.3.1). Composite samples of the replicates in field soil and ditch sediments (stratified by depth) were used, and individual samples were selected, when appropriate, to express the level of variability in the measurements.  3.3.2 Soil mineralogy: X-ray diffraction (XRD) The use of XRD for identification of unknown crystalline material is relatively common in soil analysis. The principle of XRD is based on constructive interference of monochromatic X-rays and a crystalline sample (McBride, 1994). Constructive interference is achieved when conditions satisfy Bragg’s Law, which relates the wavelength of electromagnetic radiation to the diffraction angle and the lattice spacing in a crystalline sample. Diffraction peaks are converted to d-spacings to allow for identification of the mineral because each mineral has a set of unique d-spacings. The challenge of using XRD on soil samples is that mixed materials can result in ambiguity of mineral identification (McBride, 1994), particularly since P is associated with amorphous materials. The reactive fraction of the soil, the <63µm fraction, was selectively collected by grinding composite field soil and ditch soil samples using a mortar and pestle, and sieving accordingly (Baugé et al., 2014). A dry-sieve method, was preferred to wet-sieving due to interest in maintaining the integrity of water-soluble minerals. A second set of samples with pre-treatment of the AAO solution was prepared for XRD analysis to enhance peak formation on the XRD spectra by reducing the proportion of amorphous materials. This procedure was similar to the AAO extraction in Chapter 2 (Section 2.3.4) with additional washing of the soil or sediment three times with distilled water following the removal of the supernatant. This was paired with centrifuging at 2000 rpm for five minutes to 47  separate the solid and liquid fractions. The samples were dried at 70°C for two hours. To follow, the samples were dry-sieved to isolate the 63 µm fraction, with no grinding performed.  The samples were smear-mounted on glass plates using anhydrous ethanol. XRD was measured using a Bruker D8 Focus Bragg-Brentano diffractometer with CoKα radiation and a step size of 0.04˚ over a range of 3-80˚ 2θ at 0.9 s/step (Holmes & Lavkulich, 2014). The analysis used an Fe monochromator foil, 0.6 mm divergence slit, incident and diffracted bean soller slits, and a Lynx Eye detector. A long, fine focus Co X-ray tube operating at 5kV and 40 mA was used with a take-off angle of 6˚. The rotation speed was 50 rpm. Phase identification was done using a search-match software by Bruker (DIFFRACplus EVA 16); and crystal structure data were obtained from the International Centre for Diffraction Database PDF-4+2010. This analysis took place at the Earth, Ocean, and Atmospheric Sciences Department at the University of British Columbia.    3.3.3 Inorganic and organic phosphorus compounds: Solid and liquid nuclear magnetic resonance (NMR) NMR spectroscopy is ideal for phosphorus analysis as 31P has a large gyromagnetic ratio and 100% abundance (Cade-Menun et al., 2002). During an experiment, electromagnetic radiation energy pumps alpha (low-energy) oriented nuclei into the beta state (high-energy). Once the energy is removed, the energized nuclei relax back to the alpha state, and the emission of energy associated with the relaxation process is called resonance which can be detected and converted to the peaks seen in a NMR spectrum. A resonating nucleus only experiences a portion of the magnetic field due to shielding or de-shielding from the molecule in which the nucleus is located (Cade-Menun, 2005). Different nuclei absorb and emit energy in different ways, resulting in peaks at different places in the spectrum, depending on their chemical bonds. Peak intensity is proportional to the number of each type of nuclei emitting energy. Solid NMR was conducted to assess the presence of struvite (NH4MgPO4 • 6H2O) and calcium phosphate (CaHPO4•2H2O). The analysis was performed on the untreated reactive fraction (<63 µm) (Section 3.3.2) of composite of surface soil (0-20 cm) samples from the treatment sites, given that these show the greatest diversity in organic matter and clay content. A composite sample of the ditch sediments from the mixed site was also analyzed because it expressed the greatest enrichment of P across all samples. Synthetic struvite and laboratory grade calcium phosphate were used as reference samples 48  Samples were run by Dr. Andrew Lewis at Simon Fraser University on a Bruker AVANCE III NMR spectrometer with a 9.4 Tesla (400 MHz) 89 mm (widebore) ultrashield plus superconducting magnet and a double-tuned 4 mm Bruker magic angle spinning (MAS) probe (MAS 4 BL CP BB DVT) tuned to 31P (161.96 MHz). 100-150 mg of the dry, finely powdered samples was packed into 4 mm outer diameter zirconia rotors and capped with PTFE caps.  Samples were spun at 15,000 Hz using dry compressed air and a Bruker MAS II controller. As a result of frictional-induced heating caused by the spinning, the sample temperature was approximately 305 K.   The magnet field was shimmed using a sample of water in a non-spinning MAS rotor and chemical shifts were referenced using an external sample of 85% H3PO4 (set to 0 ppm).  A one-pulse (pulse program “zg”) experiment was used (i.e. no 1H decoupling was applied) with a 2.1 µs 90-degree r.f. pulse applied at 0 ppm, a spectral width of 503 ppm (81,522 Hz), 2048 complex data points were acquired with a dwell time of 6.13 µs for (acquisition time 12.6 ms).  The delay between successive acquisitions (D1) was set to 10 s, and 8 FIDs were co-added for the struvite and CaHPO4·2H2O mineral reference samples (experiment time 1.5 min), or 1,000-5,000 scans for the soil sample (3-27 h acquisition time).  Data were zero-filled to 16,384 points, apodized by applying a 100 Hz exponential function, then Fourier transformed and baseline corrected using a 5th order polynomial.   P forms associated with organic compounds were characterized using Bowman and Moir’s (1993) extraction, followed by analysis through solution 31P-NMR. The analysis was performed on 20 samples across all sites: two surface (0-20 cm) samples from each study site; two surface (0-20 cm) ditch sediment samples from each treatment site; and two subsoil (20-50 cm and 50+ cm) samples from each treatment site. The extraction procedure was performed as outlined in Chapter 1 (Section 2.3.4). Following the extraction, 20 mL of the soil extract was filtered through Whatman 44 (3µm) filter paper; then frozen using liquid nitrogen and lyophilized over two days. The freeze-dried samples were re-suspended in 2 mL of 90 % H2O and 10 % D2O solution. Because of the high suspended sediment content, the samples were filtered again through a 0.2 µm syringe. The samples were kept at 4˚C prior to analysis.  All samples were analyzed at the Simon Fraser University NMR facility. Prior to running the samples, the procedure was optimized to account for pH differences, sensitivity to low concentrations, the effect of lyophilization, and the effect of filtration on the magnitude and references (chemical shifts) in the 31P spectra. A difference in the chemical shift value (~5.1-5.2 ppm) of phosphate compared to reports in the literature (~5.7-6.1) was found. This prompted a spiking test using CaHPO4•2H2O which confirmed the identity of the phosphate peak.  49  For preparation, 0.6 mL of each sample was pipetted into 5 mm 600 MHz-grade NMR tubes. Sample analysis was performed on the Bruker AVANCE II 600 MHz NMR spectrometer running TopSpin version 2.1 and equipped with a 5 mm QNP cryoprobe for 31P (resonance frequency of 242.9 MHz). The process was automated for each sample: samples were locked to deuterium oxide and shimmed to achieve symmetrical peak shapes, a 31P NMR spectrum was acquired using inverse-gated proton (1H) GARP composite pulse decoupling, and the chemical shift scale of each spectrum was individually referenced to an external sample of 85% H3PO4 (peak set to 0 ppm). The following acquisition and processing parameters were set to obtain quantitative spectra:  Spectral width (SW): 49 019 Hz  31P transmitter offset (O1p): 0 ppm  1H transmitter offset (O2p): 4 ppm   Number of complex data points (TD): 131 072  Acquisition time (AQ): 1.34 s  31P 90° r.f. pulse width (P1): 10.2 µs  1H 180° r.f. pulse width for decoupling: 70 µs  Relaxation delay (D1): 8.7 s  Receiver gain: 1620  Number of dummy scans: 2  Number of accumulated scans: 350  Sample temperature: 298 K  Baseline correction: 5th order polynomial spline   Apodization (LB): 2 Hz Exponential   Data acquisition time: 59 min per spectrum PULCON (Dreier & Wider, 2006) is a method which correlates the absolute peak intensities of two different spectra and allows the measurement of concentrations of compounds in samples based on a reference compound whose concentration is known. The calibration of PULCON requires a 1D spectrum measured on a sample of known concentration, which was acquired under “quantitative” conditions: a tuned and matched probe, a calibrated 90° r.f. pulse, a relaxation delay equal to at least 5×T1 (the longitudinal relaxation time), an acquisition time longer than T2 (the transverse relaxation time), and a sufficient signal to noise ratio. Concentration measurements with PULCON use the principle of reciprocity which states that the lengths of a 90° or 360° radio frequency (r.f.) pulse are inversely proportional to the NMR signal intensity (Hoult & Richards, 1976; Hoult, 2000). Therefore, provided that the concentration of one of the samples is known precisely, and that the 90° pulses have been well calibrated for each sample, the unknown concentrations can be obtained using Equation 1 (Dreier & Wider, 2006). 𝐶𝑢𝑛𝑘  = 𝑘𝐶𝑟𝑒𝑓𝐴𝑢𝑛𝑘 𝑇𝑢𝑛𝑘 𝜃90𝑢𝑛𝑘 𝑁𝑢𝑛𝑘 𝑛𝑟𝑒𝑓 𝐴𝑟𝑒𝑓 𝑇𝑟𝑒𝑓 𝜃90𝑟𝑒𝑓𝑁𝑟𝑒𝑓 𝑛𝑢𝑛𝑘 (4) 50  Here the unk and ref indices stand for unknown and reference, respectively, C is the concentration, A is the peak area of the NMR resonance, T is the temperature of the sample (in Kelvin), θ90 is the 90° r.f. pulse length, N is the number of (magnetically) equivalent nuclei in the compound contributing to the peak being integrated, n is the number of transients used for the experiments, and k is a correction factor taking into account the use of different receiver gains for measurement of the reference and of the unknown samples, or incomplete relaxation. This equation is valid when the experiments are recorded with the same NMR probe, tuned and matched properly on each sample. The reference compound can be added as an internal standard or run as a separate external sample, and more importantly, the reference material can be any compound containing the nucleus of interest (31P in this case), i.e. it does not have to be the same chemical whose concentration is being quantified in the "unknown" sample. Integration of peak area(s) of the samples is outlined in Figure 3.1 a and b. A good linear fit between 31P concentrations and NMR peak area confirms the application of a single point calibration for calculating concentrations of P compounds (i.e. phosphate, phosphate monoesters and other compounds, and pyrophosphate) in the spectra (Figure 3.2).    51       b) a) Figure 3.1 Integration of 31P peaks in samples using either of the two methods a) integration of distinct peak(s) of phosphate, phosphate monoesters and other compounds, and pyrophosphate; and b) deconvolution of peak overlap using Lorentzian fit followed by integration of the two distinct sections of the spectrum. Integral proportions were normalized to sum to a value of 1. 52               3.3.4 Zeta potential: zero point of charge Zero point of charge was measured with a ZetaProbe machine which used an electroacoustic technique that applied an electric field to the samples, causing charged particles to oscillate; this sound wave was detected and analyzed to transform the electrophoretic motion of the particles to zeta potential. Analysis was conducted on composite topsoil samples at each site. Ten grams of soil were mixed with 250 mL of 0.01 M NaCl solution to maintain consistent ionic strength. The mixture was stirred and left to equilibriate overnight. Initially, the natural pH of the soil was lowered to 4. The ZetaProbe machine was then used to measure zeta potential while automated titration commenced from pH 4 to 8 using 0.1 M HCl and 0.1 M NaOH to adjust pH at half units. The zeta potential describes the voltage difference between the Stern layer and the bulk electrolyte; the ZPC can be determined when zeta potential reaches zero.  Parameters were set at 1.2 g/cm3 for particle density, and 50 for the particle dielectric constant. This analysis took place in the Mining Department at the University of British Columbia.   Figure 3.2 The linear relationship between solutions of known concentrations of NH4H2PO4 and their respective 31P NMR peak area. 53  3.3.5 Phosphorus adsorption capacity: Linear, Langmuir and Freundlich models An adsorption isotherm describes a plot of the quantity of adsorbate (phosphate) retained by a solid (soil surface) as a function of the concentration of that adsorbate in the solution phase (soil solution) at equilibrium with the solid (McBride, 1994). The linear adsorption isotherm assumes a direct linear relationship between adsorped P and P in the equilibrium solution, with the k slope describing the relative change of adsorped P to equilibrium P (Hoseini & Talesnhmikaiel, 2013). The Langmuir model was developed in 1918 to describe the kinetics of vapour adsorption on a homogeneous surface to form a monolayer, where the surface is assumed to possess a certain number of sites (McBride, 1994; Zelazny & White, 1989). All molecules are adsorbed with the same bond strength on identical sites; thus, each molecule has an equal chance of desorbing from the surface at any particular time.  It is assumed that equilibrium state is reached. The linearized Langmuir equation is: 𝐶(𝑥𝑚)=1𝑘𝑏+𝑐𝑏 (5) where C is the P concentration in the equilibrium solution (mg/L), x/m is P adsorped by the solid phase (mg/kg), and b and k are Langmuir constants. The Langmuir b describes the P adsorption maximum (mg/kg), and k is the affinity constant related to the bonding energy of the soil. Some shortcomings of the Langmuir equation include the inability to be applied to heterogeneous surfaces, and to take coupled adsorption-desorption reactions into account. While the Langmuir equation describes chemisorption well enough, it is not appropriate to describe physical sorption which can occur in multilayers.  The Freundlich equation was empirically derived in 1909, to allow for a logarithmic increase in adsorption energy with increasing coverage (McBride, 1994; Zelazny & White, 1989). This isotherm model is considered appropriate for describing multilayer sorption and sorption on heterogeneous surfaces. The modified Freundlich equation is:  𝑥𝑚= 𝑘𝐶1𝑛 (6) where C is the P concentration in the equilibrium solution (mg/L), x/m is P adsorped by the solid phase (mg/kg), and k and 1/n are Freundlich constants. The Freundlich k refers to the ratio of P adsorped to the P in the soil solution, while 1/n describes the linearity of the adsorption curve. 54  From these isotherms, buffering capacity can be calculated to evaluate the ease of release of labile P into the soil solution (Bridgham et al., 2001; Shirvani et al., 2005). Buffering capacity indices were determined as follows (Shirvani et al., 2005). 1. Maximum buffering capacity (MBC) was the limiting slope of the Langmuir equation. 2. Buffering index (BI) was calculated from the slope of the Freundlich equation at 1µg P/mL of equilibrium soil solution. 3. Phosphorus buffering capacity (PBC) was measured as the slope of the linear P sorption equation.  4. A single-point index of buffering capacity, the phosphorus sorption index (PSI), was determined as the amount of P sorbed (mg/kg soil) from a solution containing 25 mg/L P divided by the P concentration remaining in the solution after 24 hours of shaking.  Two replicates from each depth of field soil (0-20 cm, 20-50 cm, 50+ cm) and two replicate ditch sediment samples from each applicable site were used for analysis. Three grams of soil were mixed with 30 mL of NH4H2PO4 in 0.01M LiCl solution at 10, 25, 50, 100, 200, 300 and 500 mg/L of P (Anghinoni et al., 1996). NH4H2PO is a soluble salt which dissociates to form H2PO42- the most likely phosphate anion to occur in soils within a neutral to slightly acidic pH. Surface samples (0-20 cm) were reacted to a range of 10-500 mg/L while subsurface samples (20-50 +cm) were reacted to a range of 10-200 mg/mL. Following mixing of the samples, soils were shaken for 30 minutes and left to equilibriate for 24 hours. After this time, samples were shaken for 30 minutes prior to centrifuging at 2000 rpm for five minutes. The supernatant was decanted and filtering using Whatman 44 was completed when the supernatant retained high suspended sediment content. Samples were then analyzed using the ICP-AES from the UBC Earth, Ocean, and Atmospheric Sciences Department. Fitting of the linear model, the linearized Langmuir model, and the Freundlich power model were done using Microsoft Excel 2010 (32 bit). When samples did not fit the model(s) at the lowest P concentration (10mg/L) either due to experimental error or to calibration of the ICP-AES, the point was removed for better model(s) fitting.    55  3.4 Results 3.4.1 Soil mineralogy and 31P-NMR inorganic compounds  Mineralogy of field soils and ditch sediments were consistent across the study sites. Five minerals characterized the XRD spectra of the samples: quartz, albite, clinochlore, muscovite, and apatite (Figure 3.3a). Only a slight decline in the spectral peaks of quartz and albite were found in the XRD spectra of samples with chemical pre-treatment of AAO (Figure 3.3b). This affirms that the AAO extraction removes only non-crystalline compounds.  High Mg content of Sumas soils has recently spurred investigation into the potential for the formation of struvite, NH4MgPO4, predominantly in sites of high SOM content (Baugé et al., 2014). Notably, struvite XRD spectral peaks were detected in all samples, independent of the degree of organic matter enrichment (Figure 3.4). Nonetheless, solid 31P NMR found that the spectral peak of synthetic struvite (δ = 5.5ppm) did not align with the mineral peaks (δ =2 ppm) of selected samples from the treatment sites (Figure 3.5), but may have contributed to the well-defined shoulder along the high field region of the spectra, most evidently in topsoil from the dairy and mixed sites (Figure 3.5 inset). Instead mineral peaks were found to coincide with a calcium phosphate precipitate, CaHPO4 • 2H2O. Peak broadness along with asymmetry, particularly along the low field region of the spectra, were pronounced in samples from the dairy and mixed sites, by comparison to the turf peak which was broad but exhibited symmetry.    56  a) b) Figure 3.3 XRD spectra for a) topsoil (0-20 cm) from the dairy farm with no chemical pre-treatment performed; and b) topsoil (0-20 cm) from the dairy farm with an AAO pre-treatment. 57      Figure 3.4 XRD spectra demonstrating the presence of struvite in topsoil (0-20 cm) at the dairy farm. 58     Figure 3.5 The solid-state 31P- NMR spectra of topsoil (0-20 cm) from the treatment sites, and ditch sediments (0-20 cm) from the mixed site compared with mineral peaks of calcium phosphate and synthetic struvite. Inset (left): The struvite peak relative to the broad, asymmetric inorganic P peak in dairy field soil.   59  3.4.2 31P-NMR compounds from organic soil extracts Within the major chemical shift region (10 to -10 ppm) pertinent to these soils, the mixed site contained the largest diversity and concentrations of 31P forms in both field soil and ditch sediments followed by the uncultivated, dairy, and turf sites (Figure 3.6a-d and Figure 3.7a-c). The diversity of P compounds was maintained until the largest depth (50+ cm) of field soil at the mixed site, and was diminished below the mid-level depth (20-50 cm) at the dairy farm and topsoil at the turf farm (Figure 3.8). Common to all samples was the phosphate peak at δ = 5.1-5.2 ppm and the pyrophosphate peak at δ = 5 ppm, characterizing the inorganic components found in the soil extracts (Cade-Menun, 2005). Phosphate monoesters (δ = 3-6 ppm) and mononucleotides (δ = 4.78-4.32) were found in all samples with the exception of field soil samples at depth (20-50+ cm) from the turf farm. Polynucleotides (δ = 3.5 ppm) were found in some samples, with no clear trend in occurrence. Choline phosphate (δ = 4.05 ppm) tended to be found in ditch sediments, with some exception for topsoil from the turf and uncultivated sites. Ethanolamine phosphate (δ = 4.71) was found predominantly at the mixed site, while phytic acid (δ = 5.85, 4.92, 4.55, 4.43 ppm) was found in all samples above the 50+ cm depth of field soil.  Maximum NMR P concentration was found in ditch sediments at the mixed site (1243 mg/kg) and minimal concentration at the 50+ cm depth in field soil at the turf farm (51 mg/kg), which was consistent with P extraction determinations (Table 3.1).  The range of  integrated concentrations was comparable (within 15%) to organic P extraction measurements analyzed using ICP-AES, with exception to soil samples from  the dairy and turf farms which exhibited differences between 30-70%.  Across all sites, phosphate concentrations decreased with depth but consistently comprised 70-85% of total NMR P (Figure 3.9 a, b, and c). Po, which included phosphate monoesters and other aforementioned compounds, decreased with depth at all sites and ranged from 10% at the turf farm to over 30% at the 50+ cm depth at the mixed site. Proportions of Po, averaged 25% in topsoil at all sites of high SOM content. At the dairy farm, proportions of Po in ditch sediments were more similar to field soil at the 50+ cm depth, while at the turf and mixed sites proportions were more similar to topsoil. Pyrophosphate concentrations decreased with depth at the turf and mixed sites, but increased with depth at the dairy farm. Proportions of pyrophosphate were similar across sites of high SOM content, comprising ~2-4% of total NMR 31P, with the exception to field soil at the 50+ cm depth in the dairy farm. The largest proportions of pyrophosphate were found at the turf farm, and increased markedly from 5 to 22 % with depth in field soil. Concentrations of Pi and Po were consistently greater in ditch sediments than in topsoil at the turf and mixed sites.   60  a) b) c) d)    Figure 3.6 Selected 31P NMR spectra of topsoil (0-20 cm) from the a) dairy farm, b) turf farm, c) mixed site, and d) uncultivated site.  61   Figure 3.7 Selected 31P NMR spectra of ditch sediments (0-20 cm) from the treatment sites: a) dairy farm, b) turf farm, and c) mixed site.  a) b) c) 62     Figure 3.8 31P scaled NMR spectra of field soils from the treatment sites (dairy, turf, and mixed) at depths of) 20-50 cm and b) 50+ cm.  a) b) 63  Concentration (mg/kg) ± SEM (Proportion of total NMR phosphorus)Site Sample (depth)Sample Size (n)PhosphatePhosphate monoesters and other compoundsPyrophosphateTotal NMR phosphorusField soil (0-20 cm) 2 381 ± 78 (0.75) 112 ± 14 (0.23) 13 ± 4 (0.02) 506 ± 96Field soil (20-50 cm) 1 377 (0.73) 124 (0.24) 14 (0.03) 515Field soil (50+ cm) 1 126 (0.70) 34 (0.18) 23 (0.12) 184Ditch sediment (0-20 cm) 2 255 ± 71 (0.79) 50 ± 33 (0.17) 16 ± 7 (0.04) 322 ± 45Field soil (0-20 cm) 2 229 ± 111 (0.85) 40 (0.10) 19 ± 2 (0.05) 268 ± 134Field soil (20-50 cm) 1 112 (0.86) - 18 (0.14) 131Field soil (50+ cm) 1 39 (0.78) - 11 (0.22) 51Ditch sediment (0-20 cm) 2 348 ± 71 (0.76) 59 ± 3 (0.13) 47 ± 11 (0.11) 454 ± 127Field soil (0-20 cm) 2 767 ± 121 (0.77) 205 ± 5 (0.21) 13 ± 11 (0.02) 985 ± 127Field soil (20-50 cm) 1 490 (0.71) 186 (0.27) 15 (0.02) 691Field soil (50+ cm) 1 153 (0.65) 74 (0.31) 9 (0.04) 235Ditch sediment (0-20 cm) 2 940 ± 20 (0.76) 272 ± 26 (0.22) 24 ± 0.4 (0.02) 1237 ± 6Uncultivated Field soil (0-20 cm) 2 462 ± 59 (0.72) 167 ± 21 (0.26) 10 ± 0.5 (0.02) 639 ± 81Dairy TurfMixed    Table 3.1 Mean concentrations and standard error of the mean (when applicable), along with proportions of phosphate, phosphate monoesters and other compounds, and pyrophosphate in organic P extracts of field soil and ditch sediments at the study sites. 64    Figure 3.9 Concentrations (mean values when applicable) of types of P compounds (phosphate, phosphate monoesters and other compounds, and pyrophosphate) in organic extracts of field soil and ditch sediments from the treatment sites: a) dairy, b) turf, and c) mixed.  a bc) FS- Field Soil DS- Ditch Sediment 65  a) b) c) d) 3.4.3 Surface charge of soil Within the parameters investigated, surface soils at the study sites were found not to have a zero point of charge. Despite the potential for pH-dependent charges associated with high organic matter content, soils at all sites were found to be mildly negatively charged within the pH range of 4-8. Selected plots showed zeta potential to range between ~ -6 to -12 mV for the treatments sites, and narrow to ~ -6 to -8 mV for the uncultivated site (Figure 3.10). Lack of measurable ZPC may be due to the slow kinetics of reactions occurring in the soil in response to pH changes during titration of the soil solution. An assumption of this experiment was that an equilibrium state was reached at each pH which allows for measurement of net surface charge; this assumption may not have been fulfilled due to the complexity of interactions between inorganic and organic materials in these soils. However, in natural systems a hypothetical equilibrium state may be just as elusive.    Figure 3.10 The change in zeta potential of topsoil (0-20 cm) within a pH range of 4-8 at the a) dairy farm, b) turf farm, c) mixed site, and d) uncultivated site. 66  3.4.4  Phosphorus adsorption capacity: Linear, Langmuir and Freundlich models  In general, the adsorption isotherms of the samples demonstrated good fit of the linear, modified Langmuir, and Freundlich models, with exception to selected samples from the mixed and dairy sites (Table 3.2). Indices of buffering capacity (BC) were also found to closely agree with each other, having significant (p< 0.01) Pearson r values > 0.80 across all combinations (Table 3.3). Langmuir isotherms demonstrated a tendency for points to cluster below the 50 mg/L concentration (particularly for topsoil); however, linearity was achieved at higher concentrations (Figure 3.11). None of the isotherms approached an asymptote, indicating that maximum adsorption capacity at the study sites has not been reached. The linear shape of the isotherm suggests constant relative affinity of the adsorbate molecules to the adsorbent, generally observed at a low adsorption range (McBride, 1994). Topsoil isotherms showed greater consistency across all sites, while field soil samples at depth (20-50+ cm) and ditch sediments demonstrated distinction of the turf farm due to its low adsorption capacity. The affinity constant k for the Langmuir isotherm indicated greater bonding energy in the 20-50+ cm depth compared to the topsoil and ditch sediments for all sites, except ditch sediments at the dairy farm. In topsoil, the mixed site had the lowest Langmuir k value and the uncultivated site the greatest. However, at lower depths the mixed and uncultivated sites had the highest k values overall, while the turf and dairy farms were similarly low with exception to ditch sediments at the dairy farm.  Maximum P adsorption values (b) were greatest in the ditch sediments at 3670 mg/kg and lowest at the 50+ cm depth of field soil at the turf farm at 356 mg/kg.  Adsorption capacity decreased in the following order mixed > uncultivated > dairy > turf. The mixed and uncultivated sites had minimal decline in adsorption with depth (0-50+ cm), while the turf farm had maximal decline, and the dairy farm was intermediate. Maximum adsorption in ditch sediments was greater than topsoil at all sites.  The Freundlich isotherms demonstrated separation of sites similar to the Langmuir isotherms (Figure 3.12). The curved shape of isotherms of soil samples from the turf farm, and ditch sediments from all treatment sites reflects an approach towards saturation. The overall trend was decreasing linearity with depth, the greatest difference occurring between the 20-50 and 50+ cm depth. Freundlich k increased with depth at all sites of high organic matter content, indicating increased sorption. In topsoil, soluble P decreased in the following order uncultivated > mixed ~ turf > dairy.  Overall trends in Freundlich k were more similar in the mixed and uncultivated sites. Ditch sediments had similar soluble P to topsoil at the turf and mixed sites; sediments at the dairy farm had exceptionally high k values comparable to field soil at 50+ cm depth in the uncultivated site.  67  Overall, buffering capacity was greatest at the mixed and uncultivated sites, minimal at the turf farm, and intermediate at the dairy farm (Table 3.3). BC values in mL/g were in the following range: 5-48 for MBC; 6-46 for BI; 1-12 for PBC; and 5-64 for PSI with values consistently being lowest in topsoil at the turf farm and highest in soil at 50+ cm depth at the mixed site. At the mixed and uncultivated sites, BC increases with depth in field soil, reaching a maximum at the 50+ cm depth. Alternatively, at the turf farm BC declines with depth, while at dairy farm maximum BC is calculated at the intermediate 20-50 cm depth in field soil. Trends in BC for ditch sediments were similar to other parameters measured by the Langmuir and Freundlich equations.  Pearson’s correlation test indicated significant (p < 0.01) positive relationships between maximum P adsorption and stable P (r = 0.86), available Mehlich P (r = 0.83), organic P (r = 0.83), SOM (r = 0.83), and clay content (r = 0.69); weaker associations were measured for organic carbon content (r = 0.50) and total nitrogen (0.58). Conversely, Langmuir k was negatively correlated with stable P (r = -0.60), available Mehlich P (r = -0.53), and organic P (r = -0.53).       68  Site Sample (depth)Sample Size (n)MBC BI PBC PSI Field soil (0-20cm)* 2 10.3 5.7 2.8 8.3Field soil (20-50cm) 2 16.2 18.9 5.3 21.8Field soil (50+cm) 2 13.0 15.9 3.2 16.9Ditch sediment (0-20cm) 2 28.1 35.5 3.9 40.2Field soil (0-20cm) 2 6.7 9.3 1.7 8.3Field soil (20-50cm) 2 5.0 6.2 1.4 4.8Field soil (50+cm) 2 5.6 7.7 1.3 5.2Ditch sediment (0-20cm)* 2 5.9 7.9 2.1 6.6Field soil (0-20cm)* 2 8.0 11.1 3.1 5.2Field soil (20-50cm)* 2 16.8 19.7 7.7 17.7Field soil (50+cm) 2 47.6 46.1 11.6 64.2Ditch sediment (0-20cm)* 2 9.7 11.9 4.5 5.8Field soil (0-20cm) 2 15.0 20.3 3.4 18.4Field soil (20-50cm) 2 35.0 34.7 8.2 42.3Field soil (50+cm) 2 44.3 42.0 9.0 54.1* Isotherms developed using 25 to 200 or 500 mg NH4(H2PO4)/ L  for one or both samples.  MixedUncultivatedBuffering Capacity Indices (mL/g soil)Dairy TurfSite Sample (depth)Sample Size (n)r2k r2k b- Maximum P adsorption (mg/kg)r2k 1/nField soil (0-20cm)* 2 0.88 3.01 0.46 0.003 2224 0.88 11.09 0.849Field soil (20-50cm) 2 0.93 5.34 0.77 0.013 1249 0.90 26.12 0.738Field soil (50+cm) 2 0.90 3.21 0.90 0.017 775 0.88 24.18 0.666Ditch sediment (0-20cm) 2 0.86 3.89 0.97 0.016 1687 0.90 61.54 0.589Field soil (0-20cm) 2 0.90 1.73 0.90 0.006 1106 0.92 12.92 0.733Field soil (20-50cm) 2 0.90 1.42 0.94 0.012 432 0.94 8.68 0.721Field soil (50+cm) 2 0.93 1.26 0.98 0.016 356 0.96 12.85 0.610Ditch sediment (0-20cm)* 2 0.91 2.08 0.96 0.007 1203 0.93 17.82 0.692Field soil (0-20cm)* 2 0.90 3.09 0.65 0.004 2518 0.92 14.22 0.805Field soil (20-50cm)* 2 0.87 8.13 0.45 0.012 1818 0.77 32.99 0.781Field soil (50+cm) 2 0.87 11.59 0.83 0.029 1666 0.85 67.03 0.698Ditch sediment (0-20cm)* 2 0.91 4.55 0.47 0.003 3670 0.91 13.80 0.866Field soil (0-20cm) 2 0.89 3.44 0.89 0.009 1791 0.90 29.81 0.692Field soil (20-50cm) 2 0.88 8.23 0.90 0.026 1403 0.89 51.06 0.694Field soil (50+cm) 2 0.87 9.00 0.93 0.032 1378 0.87 64.36 0.658* Isotherms eveloped using 25 to 200 or 500 mg NH4(H2PO4)/ L  for one or both samples.  FREUNDLICHLANGMUIRLINEARDairyTurfUncultivatedMixed  Table 3.3 Buffering capacity indices of field soils and ditch sediments at the study sites, including: maximum buffering capacity (MBC), buffering index (BI), phosphorus buffering capacity (PBC), and phosphorus sorption index (PSI).  Table 3.2 Parameters of the linear, Langmuir, and Freundlich models for phosphorus adsorption in field soils and ditch sediments at the study sites. 69       Figure 3.11 Selected Langmuir isotherms for: a) topsoil (0-20 cm), b) soil at 20-50 cm depth, and c) soil at 50+ cm depth at the study sites; and d) ditch sediments (0-20 cm) at the treatment sites.   a) b) c) d) 70   Figure 3.12 Selected Freundlich isotherms for: a) topsoil (0-20 cm), b) soil at 20-50 cm depth, and c) soil at 50+ cm depth at the study sites; and d) ditch sediments (0-20 cm) at the treatment sites. a) b) c) d) 71  3.5 Discussion 3.5.1 Inorganic and organic phosphorus compounds The mineralogy of the study sites indicated a low proportion of apatite, was inconclusive for the presence of struvite, and strongly suggestive of the presence of CaHPO4 • 2H2O. Feldspars and low-reactivity aluminosilicates comprised the dominant mineral fractions, inferring limited potential for amphoteric charged surfaces. This is contrary to Hashimoto and Watanabe’s (2014) findings of Fe minerals being the major P species in serpentinitic forest soils, while associations of P with Ca or apatite like compounds were generally minor. Their XANES investigation did not find a distinct struvite peak, and further analysis did not determine struvite to be a major P component. Notably, laboratory reactions of MgHP04 •3H20 precipitate with amorphous Al oxides resulted in full dissolution within 3 days under high relative humidity, which suggests that while struvite may be formed in Sumas soils, it is unlikely to be retained for long periods of time (The Clay Minerals Society, 1987).  Despite the correlative association of Al with total and stable P fractions, solid-state 31P-NMR demonstrated that mineral peaks from the treatment sites most closely aligned with a calcium phosphate precipitate, CaHPO4 • 2H2O. The broad, asymmetric nature of these peaks suggests the presence of a minor component, in addition to a major component, within the range of -2 to -20 ppm.  As such, it is possible that P adsorped on amorphous aluminum oxides, which by nature would not produce distinct peaks, occupy this range of the spectra as  Li et al. (2013) found evidence of P adsorped on bayerite, gibbsite, boehmite, and alumina within this range. It cannot be determined from the current spectra whether calcium phosphate or amorphous aluminum phosphates dominate inorganic P sources. Application of superphosphate has been known to dissolve in the presence of Al oxides to produce non-crystalline aluminum phosphate as well as CaHPO4 • 2H2O, which can remain in soils having high phosphate retention capacity at low relative humidity (The Clay Minerals Society, 1987). Further, the formation of various calcium phosphates have been reported in manure-amended soils as animal feed is often enriched in Ca and P (Cao & Harris, 2008; Kronqvist, 2011)). Pierzynski et al. (1990) posit that P-rich particles reflect accumulation of products from multiple precipitation events which is likely given continuous application of fertilizer and manure. Given the evidence provided, it is inconclusive whether calcium phosphate precipitate is occurring as a discrete particle or as a coating on Al oxides.  In serpentinitic soils, Hashimoto and Watanabe (2014) proposed that P speciation was severely low and relied substantially on P from organic subcycles. Integrated 31P concentrations can be categorized as Pi, comprised of orthophosphate and pyrophosphate fractions, and Po comprised of predominantly 72  phosphate monoesters as well as other compounds (potentially mononucleotides, polynucleotides, choline phosphate, ethanolamine phosphate, and phytic acid). Overall, Pi accounted for 69-100% and Po for 0-31% of NMR P. Both species and relative proportions were consistent with Hashimoto and Watanabe (2014), with the exception of phosphate diesters which were absent in Sumas soils perhaps due to their high susceptibility to microbial degradation. Despite singular application of manure at the dairy farm, the mixed site still contained higher Po concentrations. Many studies have found that additions of mineral fertilizer can increase Po content in soils, mainly as orthophosphate monoesters, either through increased plant material growth or a reduction of mineralization due to increased P availability (Turner et al, 2002). Further, in NMR studies of dairy manure He et al. (2009) found that orthophosphate comprised 70-89% of P in extracts, 24% were associated with soluble Ca and Mg phosphate species, and stable species were mainly comprised of phytate bonded to Al(III). As such, the composition of P in organic extracts from the study sites was consistent with soils that have been nutritionally amended.  The differential affinity of Pi and Po for sorption sites will determine their potential for stabilization in the soil. However, at low P concentrations, P will be adsorbed as inner sphere surface complexes with a maximum of one phosphate per surface Al atom (Lookman et al., 1994). It is likely that inositol phosphates comprise the majority of Po, which includes phytic acid (IP6) as well as typically 60% of phosphate monoesters (Turner et al., 2002). Since IP6 has a high charge density, it undergoes strong interaction with the soil, typically being adsorped to clays or precipitated as insoluble salts and thus limiting the likelihood for desorption, which can explain the consistency of PO fractions maintained with depth at the dairy and mixed sites. Celi et al. (1999) found that IP6 had greater affinity than orthophosphate for goethite, illite and kaolinite sorption. Pyrophosphate originates from the microbial degradation of inorganic polyphosphate and is typically immobilized in soil and persist for long periods of time (Blanchar & Hossner, 1969). It has also been found to be preferentially sorbed compared to orthophosphate, which may explain its relative accumulation at the turf farm (Mcbeath et al., 2004). Adsorption of P nucleotides on clay minerals is substantially influence by pH and subsequent changes in surface charge: above pH 6.3, the phosphate group carries two negative charges and the base is neutral, while between pH 6.3and 2.4 nucleotides were left with one negative charge on the phosphate group (Feuillie et al., 2013). Since soil surfaces tend to be negatively charged, a lower negative charge on the phosphate group would facilitate greater sorption.   73  3.5.2 Dynamics of phosphorus adsorption It is evident that surface charge can substantially influence P retention as Pi and Po forms are predominantly anionic. As such, the negative charge of these soils across a pH range of 4-8 may be explained by the following factors: i) low clay content and low reactivity of dominant clay minerals resulting in generally negative but overall limited variable charges; ii) high SOM content with predominant hydroxyl, phenolic, and carboxyl groups which contribute substantial negative charges (McBride, 1994); and iii) slow kinetic response of aluminosilicates to reach an equilibrium state particularly since surface interactions can include adsorption and precipitation with organic fractions, where reactions may occur more slowly (Shang et al., 1990).  ZPC analysis conducted by Celi et al. (1999)  have measure surface charges on the scale of +40 mV to -55 mV for goethite; thus resulting in substantial changes in P sorption. Since changes in zeta potential ranged from only -6 mV to -12 mV in Sumas soils, this may imply generally weak sorption strength for P anions complexed through associations with cations in the soil solution.  Adsorption capacity was associated with SOM and clay content as they provide sorption sites that facilitate P accumulation. Associations were also found with carbon and nitrogen content which suggests some degree of coupled cycling of the nutrients. Sorption affinity was negatively correlated with stable, organic, and available P which affirms that these fractions capture P occupying the sorption sites. Since solid-state 31P NMR suggests that precipitation and chemisorption were simultaneously occurring, it is possible that the Langmuir and Freundlich models do not fully capture adsorption properties in these soils; however, these models do demonstrate consistency with P extractions findings. The Langmuir isotherms revealed low adsorption at the study sites; calculated maximum adsorption values were above stable, labile, available, and organic P concentrations, but were comparable to or exceeded total P concentrations particularly with depth in the soil. The Freundlich isotherms demonstrated that soil and ditch sediments from the turf farm approach an asymptotic value, which was contrary to DPS measurements at this site. Ditch sediments from all treatment sites were nearing saturation which was consistent with high P concentrations and DPS measurements. There was consistency between the Langmuir and Freundlich models which indicated increased bonding affinity and decreased soluble P with depth in sites of high SOM content. While the mixed site had lowest bonding affinity, the dairy farm had the largest soluble P in topsoil. Buffering capacity of the sites decreased in the following order mixed > uncultivated > dairy > turf. With increasing soil depth, adsorption properties of the mixed and uncultivated sites became more alike and distinct from the dairy and turf farms. Adsorption properties of ditch sediments were comparable to topsoil at the turf 74  and mixed sites. Ditch sediments at the dairy farm had consistently greater bonding affinity, lower soluble P, and comparable adsorption capacity to field soil which may be related to the distinct enrichment of carbon and nitrogen in this sample resulting in a slower mineralization rate.  Given the evidence provided, sorption dynamics as related to the Pi and Po forms present in Sumas soils may be broadly inferred. Since the soil surface is mildly negatively charged, there is greater preference for sorption of phosphate groups with lower negative charges. Soluble P concentrations are low, favouring the formation of inner sphere surface complexes on amorphous Al oxides which provide a more stable source of P. Long term addition of fertilizer and manure have amended the expected deficiency of Ca in these soils, facilitating the simultaneous formation of calcium phosphates which provide a more labile source of P. There is evidence that organic matter is a source of P, provides sorption sites for the retention P, and can chemically stabilize P in the soil. While Po is typically assumed to be more mobile in soils, several studies have found greater sorption affinity of phosphate monoesters and Po compounds, particularly inositol phosphates, compared to orthophosphate. Thus competition within sources of Po and between Po and Pi most likely influence P sorption and desorption in these soils. This is particularly evident in the dairy, mixed, and uncultivated sites where SOM content is high. At the turf farm, competition may occur between orthophosphate and pyrophosphate; however, low SOM and lack of competition with Po may limit P retention as well as release. With variable pH, changes in surface charge of Po compounds can affect their potential for sorption. The retention of Po is likely facilitated by cations in the soil, potentially Mg since it is abundant. It is likely that changes in surface properties such as increased bonding affinity and decreased soluble P with depth, as well as trends in buffering capacity are related to the following factors: P concentrations in labile pools; stabilization of Pi and Po through association with amorphous Al oxides, clay colloids or SOM; and diversity of P compounds which may enable a variety of bonding mechanisms.      75  4. Conclusions  The specific conclusions of this study of phosphorus dynamics in the Sumas Prairie can be summarized in the following: 1. Measurements of total, stable, available (Bray and Mehlich), and organic P affirm P is vertically mobile within the soil matrix of all study sites, particularly down to the 20-50 cm soil depth. Accumulation of P in ditch sediments was evident at the turf and mixed sites. Overall P concentrations were largest at the mixed site, lowest at the turf farm, and similar at the dairy and uncultivated sites; although statistically, P concentrations were only distinct at the turf farm.   2. P fractionation indicated the following relative composition of total P (AR): 55-80% stable P (AAO), 12-46% labile P (HCl), 2-5% available P (Bray), 2-7% available P (Mehlich), and 10-50% organic P (NaOH-EDTA).   3. Mineral P sources are attributed to apatite, P adsorped on amorphous aluminum oxide as a stable source of P, and calcium phosphate precipitate as a labile source of P. There was inconclusive evidence of the presence of struvite in these soils. The effect of Mg in P sorption is likely related to facilitating Po sorption and inhibiting the formation of calcium phosphates.  4. 31P NMR analysis of organic extracts indicated that Pi, comprised of orthophosphate and pyrophosphate fractions accounted for 69-100% of extracted P, and Po comprised of predominantly phosphate monoesters as well as other compounds accounted for 0-31%. .  5. There is evidence that organic matter is a source of P, provides sorption sites for the retention of P, and can chemically stabilize P in the soil. As such SOM was the most important factor differentiating the type and status of soil P, as well as adsorption and saturation characteristics among the study sites. Despite differences in P concentrations, carbon, nitrogen, and phosphorus nutrient ratios did not substantially differ among the treatment and uncultivated sites, and showed overall efficient mineralization dictated by the microbial populations at each site.  6. P content in soils and ditch sediments at all study sites were below maximum adsorption; however, samples from the turf farm as well as ditch sediments may be nearing saturation. According to the Langmuir and Freundlich models, P bonding affinity, sorption, and buffering capacity increased with depth at the mixed and uncultivated sites, and were less conclusive at the dairy and turf farms. There is evidence that the turf and uncultivated sites were less impacted by serpentinitic contributions compared to the dairy and mixed sites. However, adsorption properties were more similar for the mixed and uncultivated sites compared to the dairy and turf farms suggesting that the influence of SOM is more important for determining sorption properties.  76  7. Processes of sorption and desorption in these soils are expected to be influenced by competition within sources of Po and between Po and Pi forms, particularly in the dairy, mixed, and uncultivated sites where SOM content is high.  8. The narrow pH range of Sumas soils was found not to be associated with P content. Clay content and SOM were positively associated with all P pools, and with increased degrees of phosphorus saturation measured by DPSox, M3-PSR I, and M3-PSR II.  9. There is potential for P loss and subsequent environmental contamination from topsoil and ditch sediments at the mixed site according to DPSox, and from topsoil and ditch sediments at all sites according to M3-PSR I and M3-PSR II. 10. In evaluating the utility of selected extractions, the consistency of AR, Mehlich-III, and NaOH and EDTA is salient for future use in the region and in other sites of similar geochemical properties. Since AR extraction is not a standard soil test and is exacting to perform, AAO is an effective alternative to use in soils with circumneutral pH.  The results demonstrate that phosphorus availability and movement is influenced by both the organic and inorganic forms within the soil system and there is a dynamic exchange among the various forms. The NMR data contributes to understanding of the portions of soil phosphorus that are relatively labile and those that are more stable. This requires further examination but does aid in the rationalization of phosphorus mobility in soil systems and puts into question the generally held belief that P is a non-mobile element in soil systems. This belief is based, in part, by assuming that soil systems are in thermodynamic equilibrium. In fact, soils are in “dynamic steady state”, that is chemical species are transient meta-stable entities that are the result of environmental (management) pressures. Thus, although equilibrium concepts, such as solubility products,  help to explain the ultimate controls on species viability, meta-stable phases such as P forms that are biological available do exist in the system. This dynamic helps explain how relatively insoluble elements, such as P, can become mobile. These observations would benefit from examination of P species during the humification process, as occurs during composting, and in aiding in the management of animal manures prior to their land applications. This study would also benefit from the examination of several more fields with different management regimes and also with the study of soil solution chemistry through the use of soil lysimeters.  77  Bibliography Abdala, D. B., da Silva, I. R., Vergütz, L., & Sparks, D. L. (2015). Long-term manure application effects on phosphorus speciation, kinetics and distribution in highly weathered agricultural soils. Chemosphere, 119, 504–514. doi:10.1016/j.chemosphere.2014.07.029 Anghinoni, I., Baligar, V. C., & Wright, R. J. (1996). Phosphorus sorption isotherm characteristics and availability parameters of Appalachian acidic soils. Communications in Soil Science and Plant Analysis, 27, 2033–2048. Ashley, K., Cordell, D., & Mavinic, D. (2011). A brief history of phosphorus: From the philosopher’s stone to nutrient recovery and reuse. Chemosphere, 84(6), 737–746. doi:10.1016/j.chemosphere.2011.03.001 Atkinson, H. J., Giles, G. R., MacLean, A. J., & Wright, J. R. (1958). Chemical methods of soil analysis. Contribution no. 169 (Revised). Ottawa: Chemical Division of Science Services. CDA. Baken, S., Verbeeck, M., Verheyen, D., Diels, J., & Smolders, E. (2015). Phosphorus losses from agricultural land to natural waters are reduced by immobilization in iron-rich sediments of drainage ditches. Water Research, 71, 160–170. doi:10.1016/j.watres.2015.01.008 Baugé, S. M. Y., Lavkulich, L. M. (Les), Wilson, J. E., & Schreier, H. E. (2014). Comparison of surface properties of synthetic and soil struvite. Canadian Journal of Soil Science, 94, 169–176. doi:10.4141/cjss2013-099 Baugé, S. M. Y., Lavkulich, L. M., & Schreier, H. E. (2013a). Phosphorus and trace metals in serpentine-affected soils of the Sumas Basin, British Columbia. Canadian Journal of Soil Science, 93(3), 359–367. doi:10.4141/cjss2012-138 Baugé, S. M. Y., Lavkulich, L. M., & Schreier, H. E. (2013b). Serpentine affected soils and the formation of magnesium phosphates (struvite). Canadian Journal of Soil Science, 93(2), 161–172. doi:10.4141/cjss2012-117 Beauchemin, S., & Simard, R. R. (1999). Soil phosphorus saturation degree: Review of some indices and their suitability for P management in Québec, Canada. Canadian Journal of Soil Science, 79(4), 615–625. doi:10.4141/S98-087 Benbi, D. K., & Gilkes, R. J. (1987). The movement into soil of P from superphosphate grains and its availability to plants. Fertilizer Research, 12, 21–36. doi:10.1007/BF01049418 Berg, A. S., & Joern, B. C. (2006). Sorption dynamics of organic and inorganic phosphorus compounds in soil. Journal of Environment Quality, 35, 1855. doi:10.2134/jeq2005.0420 Berka, C., Schreier, H., & Hall, K. (2001). Linking water quality with agricultural intensification in a rural watershed. Water, Air, and Soil Pollution, 127, 389–401. Bjorkman, T., & Reiners, S. (2013). Application of bicarbonate to high-phosphorus soils to increase plant-available phosphorus. Soil Science Society of America Journal, 78, 319–324. 78  Blanchar, R. W., & Hossner, L. R. (1969). Hydrolysis and sorption reactions of orthophosphate, pyrophosphate, tripolyphosphate, and trimetaphosphate anions added to an Elliot soil. Soil Science Society of America, 33(1), 141–144. Bonifacio, E., Celi, L., & Barberis, E. (2001). Effect of phosphate on the dissolution of serpentine. Soil Science, 166(10), 708–716. doi:10.1097/00010694-200110000-00006 Borggaard, O. K., Raben-Lange, B., Gimsing, a. L., & Strobel, B. W. (2005). Influence of humic substances on phosphate adsorption by aluminium and iron oxides. Geoderma, 127, 270–279. doi:10.1016/j.geoderma.2004.12.011 Bowman, R. A., & Cole, C. V. (1978). Transformations of organic phosphorus substrates in soils as evaluated by NaHCO3 extraction. Soil Science. doi:10.1097/00010694-197801000-00008 Bowman, R. A., & Moir, J. O. (1993). Division s-2 notes, 1518(May 1992), 1516–1518. Boyle, C. A., Lavkulich, L.M., Schreier, H., & Kiss, E. (1997). Changes in land cover and subsequent effects on Lower Fraser Basin ecosystems from 1827 to 1990. Environmental Management, 21(2), 185–196. Bray, R. H., & Kurtz, L. T. (1945). Determination of total, organic, and available forms of phosphorus in soils. Soil Science, 59, 39–45. Breeuwsma, A., Reijerink, J. G. ., & Schoumans, O. F. (1995). Impact of manure on accumulation and leaching of phosphate in areas of intensive livestock farming. In K. Steele (Ed.), Animal waste and the land water interface (pp. 239–251). New York: Lewis Publishing. Breeuwsma, A., & Reijerink, J. G. A. (1992). Phosphate-saturated soils: A “new” environmental issue. In Chemical time bombs. Proceedings of the European Conference, (pp. 79– 85). Hoofddorp, the Netherlands: Ecodevelopment., Foundation for Veldhoven, the Netherlands. Bridgham, S. D., Johnston, C. A., Schubauer-Berigan, J. P., & Weishampel, P. (2001). Phosphorus sorption dynamics in soils and coupling with surface and pore water in riverine wetlands. Soil Sci Soc Am J, 65, 577–588. Cade-Menun, B. J. (2005). Chapter 2. Using phosphorus-31 nuclear magnetic resonance spectroscopy to characterize organic phosphorus in environmental samples. In B. L. Turner, E. Frossard, & D. S. Baldwin (Eds.), Organic phosphorus in the environment (pp. 21–44). Balboa, Ancon, Panama. Cade-Menun, B. J., Berch, S. M., Preston, C. M., & Lavkulich, L. M. (2000). Phosphorus forms and related soil chemistry of Podzolic soils on northern Vancouver Island. II. The effects of clear-cutting and burning. Canadian Journal of Forest Research, 30, 1726–1741. doi:10.1139/x00-099 Cade-Menun, B. J., Liu, C. W., Nunlist, R., & McColl, J. G. (2002). Soil and litter phosphorus-31 nuclear magnetic resonance spectroscopy: Extractants, metals, and phosphorus relaxation times. Journal of Environment Quality, 31, 457–465. 79  Cade-Menun, B. J., & Preston, C. M. (1996). A comparison of soil extraction procedures for 31P NMR Spectroscopy. Soil Science, 161, 770–785. doi:10.1097/00010694-199611000-00006 Canadian Council of Ministers of the Environment. (2004). Canadian water quality guidelines for the protection of aquatic life - phosphorus: Canadian guidance framework for the management of freshwater systems. Canadian Environmental Quality Guidelines, 6. Cao, X., & Harris, W. (2008). Carbonate and magnesium interactive effect on calcium phosphate precipitation. Environmental Science and Technology, 42, 436–442. Celi, L., Lamacchia, S., Marsan, F. A., & Barberis, E. (1999). Interaction of inositol hexaphosphate on clays: Adsorption and charging phenomena. Soil Science, 164, 574–585. doi:10.1097/00010694-199908000-00005 Cheng, M., & Ma, L. Q. (2001). Comparison of three aqua regia digestion methods for twenty Florida soils. Soil Science Society of America, 65, 491–499. City of Abbotsford. (2014). 2013 Abbotsford ditch maintenance program: Fact sheet (pp. 1–2). Cleveland, C. C., & Liptzin, D. (2007). C:N:P stoichiometry in soil: Is there a “Redfield ratio” for the microbial biomass? Biogeochemistry, 85(3), 235–252. doi:10.1007/s10533-007-9132-0 Cordell, D., Drangert, J. O., & White, S. (2009). The story of phosphorus: Global food security and food for thought. Global Environmental Change, 19, 292–305. doi:10.1016/j.gloenvcha.2008.10.009 Correll, D. L. (1998). The role of phosphorus in the eutrophication of receiving waters: A review. Journal of Environment Quality, 27, 261. doi:10.2134/jeq1998.00472425002700020004x Daniel, T. C., Sharpley, A. N., & Lemunyon, J. L. (1998). Agricultural phosphorus and eutrophication: A symposium overview. Journal of Environment Quality, 27, 251. doi:10.2134/jeq1998.00472425002700020002x Dao, T. H., & Schwartz, R. C. (2010). Mineralizable phosphorus, nitrogen, and carbon relationships in dairy manure at various carbon-to-phosphorus ratios. Bioresource Technology, 101(10), 3567–3574. doi:10.1016/j.biortech.2009.12.070 De Smet, J., Hofman, G., Vanderdeelen, J., Meirvenne, M., & Baert, L. (1996). Phosphate enrichment in the sandy loam soils of West-Flanders, Belgium. Agroecosystems, 43, 209–215. De Solla, S. R., Pettit, K. E., Bishop, C. a., Cheng, K. M., & Elliott, J. E. (2002). Effects of agricultural runoff on native amphibians in the Lower Fraser River Valley, British Columbia, Canada. Environmental Toxicology and Chemistry, 21(2), 353–360. doi:10.1897/1551-5028(2002)021<0353:EOARON>2.0.CO;2 Devau, N., Hinsinger, P., Le Cadre, E., Colomb, B., & Gérard, F. (2011). Fertilization and pH effects on processes and mechanisms controlling dissolved inorganic phosphorus in soils. Geochimica et Cosmochimica Acta, 75, 2980–2996. doi:10.1016/j.gca.2011.02.034 80  Dreier, L., & Wider, G. (2006). Concentration measurements by PULCON using X-filtered or 2D NMR spectra. Magnetic Resonance Chemistry, 44, 206–212. Dunne, E., McKee, K., Clark, M., Grunwald, S., & Reddy, K. (2007). Phosphorus in agricultural ditch soil and potential implications for water quality. Journal of Soil and Water Conservation, 62, 244–252. Eghball, B., Binford, G.D. & Baltensperger, D.D. (1996). Phosphorus movement and adsorption in a soil receiving long-term manure and fertilizer application. J. Environ. Qual. 25: 1339– 1343. Environment Canada. (2013). Phosphorus in Canada’s Aquatic Ecosystems. Retrieved March 15, 2015, from http://www.ec.gc.ca/eaudouce-freshwater/default.asp?lang=En&n=0A77A85E-1&printfullpage=true Feuillie, C., Daniel, I., Michot, L. J., & Pedreira-Segade, U. (2013). Adsorption of nucleotides onto Fe-Mg-Al rich swelling clays. Geochimica et Cosmochimica Acta, 120, 97–108. doi:10.1016/j.gca.2013.06.021 Frossard, E., Stewart, J. W. B., & St. Arnaud, R. J. (1989). Distribution and mobility of phosphorus in grassland and forest soils of Saskatchewan. Canadian Journal of Soil Science, 69, 401–416. doi:10.4141/cjss89-040 Gahoonia, T.S., Claassen, N., & Jungk, A. (1992). Mobilization of phosphate in different soils by ryegrass supplied with ammonium or nitrate. Plant and Soil, 140, 241–248. Giroux, M., & Tran, T. S. (1996). Critères agronomiques et environnementaux liés à la disponibilité, la solubilité et la saturation du phosphore des sols agricoles du Québec. Agrosol, 9(2), 51–57. Government of Canada. (2015). Canadian Climate Normals 1981-2010 Station Data. Retrieved January 10, 2015, from http://climate.weather.gc.ca/climate_normals/results_1981_2010_e.html?stnID=702&autofwd=1 Grant, R., Laubel, A., Kronvang, B., Andersen, H. E., Svendsen, L. M., & Fuglsang, A. (1996). Loss of dissolved and particulate phosphorus from arable catchments by subsurface drainage. Water Research, 30(11), 2633–2642. Groppo, J. D., Lins, S. R. M., Camargo, P. B., Assad, E. D., Pinto, H. S., Martins, S. C., Martinelli, L.A. (2015). Changes in soil carbon, nitrogen and phosphorus due to land-use changes in Brazil. Biogeosciences Discussions, 12, 2533–2571. doi:10.5194/bgd-12-2533-2015 Güngör, K., Jürgensen, A., & Karthikeyan, K. G. (2007). Determination of phosphorus speciation in dairy manure using XRD and XANES spectroscopy. Journal of Environmental Quality, 36, 1856–1863. doi:10.2134/jeq2006.0563 Gustafsson, J. P., Mwamila, L. B., & Kergoat, K. (2012). The pH dependence of phosphate sorption and desorption in Swedish agricultural soils. Geoderma, 189-190, 304–311. doi:10.1016/j.geoderma.2012.05.014 81  Harrison, A. F. (1982). Labile organic phosphorus mineralization in relationship to soil properties. Soil Biology and Biochemistry, 14, 343–351. doi:10.1016/0038-0717(82)90004-9 Hashimoto, Y., & Watanabe, Y. (2014). Combined applications of chemical fractionation, solution 31P-NMR and P K-edge XANES to determine phosphorus speciation in soils formed on serpentine landscapes. Geoderma, 230-231, 143–150. doi:10.1016/j.geoderma.2014.04.001 He, Z., Honeycutt, C. W., Griffin, T. S., Cade-Menun, B. J., Pellechia, P. J., & Dou, Z. (2009). Phosphorus forms in conventional and organic dairy manure identified by solution and solid state p-31 NMR spectroscopy. Journal of Environmental Quality, 38, 1909–1918. doi:10.2134/jeq2008.0445 Heathwaite, A. L., & Dils, R. M. (2000). Characterising phosphorus loss in surface and subsurface hydrological pathways. Science of the Total Environment, 251-252, 523–538. doi:10.1016/S0048-9697(00)00393-4 Hedley, M. J., Stewart, J. W. B., & Chauhan, B. S. (1982). Changes in inorganic and organic soil phosphorus fractions induced by cultivation practices and by laboratory incubations. Soil Sci Soc Am J, 46, 970–976. Hendershot, W. H., Lalande, H., & Duquette, M. (1993). Chapter 16: Soil reaction and exchangeable acidity. In M. R. Carter (Ed.), Soil Sampling and Methods of Analysis (pp. 141–143). Boca Raton: Lewis Publishers. Heuck, C., Weig, A., & Spohn, M. (2015). Soil microbial biomass C:N:P stoichiometry and microbial use of organic phosphorus. Soil Biology and Biochemistry, 85, 119–129. doi:10.1016/j.soilbio.2015.02.029 Holmes, E. P., & Lavkulich, L. M. (2014). The effects of naturally occurring acids on the surface properties of chrysotile asbestos. Journal of Environmental Science and Health. Part A, Toxic/hazardous Substances & Environmental Engineering, 49, 1445–52. doi:10.1080/10934529.2014.928558 Hoseini, Y., & Talesnhmikaiel, R. D. (2013). Comparison of phosphorus adsorption isotherms in soil and its relation to soil properties. International Journal of Agriculture: Research and Review, 3(1), 163–171. Hoult, D. I. (2000). The principle of reciprocity in signal strength calculations—A mathematical guide. Concepts in Magnetic Resonance, 12(4), 173–187. Hoult, D. I., & Richards, R. E. (1976). The signal-to-noise ratio of the nuclear magnetic resonance experiment. Journal of Magnetic Resonance, 24(1), 71–85. Huang, L., Thompson, A., & Zhang, G. (2014). Soil & Tillage Research Long-term paddy cultivation significantly alters topsoil phosphorus transformation and degrades phosphorus sorption capacity, 142, 32–41. IRC Integrated Resource Consultants Inc. (1994). Agricultural landuse survey in the Sumas River watershed summary report. 82  Jackman, R. H., & Black, C. A. (1952). Solubility of iron, aluminium, calcium, and magnesium inositol phosphates at different pH values. Soil Science, 72(1), 179-186.  Johnston, A. E., Poulton, P. R., Fixen, P. E., & Curtin, D. (2014). Phosphorus. Its Efficient Use in Agriculture. Advances in Agronomy (1st ed., Vol. 123, pp. 177–228). Elsevier Inc. doi:10.1016/B978-0-12-420225-2.00005-4 Kazakou, E., Dimitrakopoulos, P. G., Baker, a. J. M., Reeves, R. D., & Troumbis, A. Y. (2008). Hypotheses, mechanisms and trade-offs of tolerance and adaptation to serpentine soils: From species to ecosystem level. Biological Reviews, 83, 495–508. doi:10.1111/j.1469-185X.2008.00051.x Kettler, T. A., Doran, J. W., & Gilbert, T. L. (2001). Simplified method for soil particle-size determination to accompany soil-quality analyses. Soil Science Society of America Journal, 65(3), 849. doi:10.2136/sssaj2001.653849x Khiari, L., Parent, L. E., Pellerin, A., Alimi, A. R. A., Simard, R. R., Tremblay, C., & Fortin, J. (2000). An agroenvironmental phosphate saturation index for acid, coarse-textured, soils. Journal of Environment Quality, 29, 1561–1567. King, K. W., Williams, M. R., & Fausey, N. R. (2014). Contributions of systematic tile drainage to watershed-scale phosphorus transport. Journal of Environment Quality, 44(2), 486-494. doi:10.2134/jeq2014.04.0149 Kowalenko, C. G., Schmidt, O., & Hughes-Games, G. (2007). Fraser Valley Soil Nutrient Study 2005 (pp. 1–47). Kronqvist, C. (2011). Minerals to dairy cows with focus on calcium and magnesium balance. Swedish University of Agricultural Sciences. Lair, G. J., Zehetner, F., Khan, Z. H., & Gerzabek, M. H. (2009). Phosphorus sorption-desorption in alluvial soils of a young weathering sequence at the Danube River. Geoderma, 149(1-2), 39–44. doi:10.1016/j.geoderma.2008.11.011 Li, M., Osaki, M., Rao, I. M., & Tadano, T. (1997). Secretion of phytase from the roots of several plant species under phosphorus-deficient conditions. Plant and Soil, 195, 161–169. Li, W., Feng, X., Yan, Y., Sparks, D. L., & Phillips, B. L. (2013). Solid-state NMR spectroscopic study of phosphate sorption mechanisms on aluminum (hydr)oxides. Environmental Science and Technology, 47, 8308–8315. doi:10.1021/es400874s Liang, X., Liu, J., Chen, Y., Li, H., Ye, Y., Nie, Z., & Xu, Z. (2010). Effect of pH on the release of soil colloidal phosphorus. Journal of Soils and Sediments, 10, 1548–1556. doi:10.1007/s11368-010-0275-6 Lindsay, W. L., Vlek, P. L. G., & Chien, S. H. (1989). Chapter 22. Phosphate minerals. In J. B. Dixon & S. B. Weed (Eds.), Minerals in Soil Environments (Second Edi, pp. 1089–1130). Madison, Wisconsin: Soil Science of America. 83  Liu, F., Xiao, R., Wang, Y., Li, Y., Zhang, S., Luo, Q., & Wu, J. (2013). Effect of a novel constructed drainage ditch on the phosphorus sorption capacity of ditch soils in an agricultural headwater catchment in subtropical central China. Ecological Engineering, 58, 69–76. doi:10.1016/j.ecoleng.2013.06.008 Lookman, R., Grobet, P., Merckx, R., & Vlassak, K. (1994). Phosphate sorption by synthetic amorphous aluminum hydroxides; a 27Al and 31P solid-state MAS NMR spectroscopy study. European Journal of Soil Science, (March), 37–44. Luttmerding, H. A. (1981). Soils of the Langley-Vancouver Map Area. Report No. 15. Volume 3 Description of Soils (pp. 1–242). Kelowna. Luttmerding, H. A. (1984). Soils of the Langley-Vancouver Map Area. Report No. 15. Volume 5 (RAB Bullet, pp. 1–116). Kelowna: BC Ministry of Environment. MacDonald, G. K., Bennett, E. M., Potter, P. A., & Ramankutty, N. (2011). Agronomic phosphorus imbalances across the world’s croplands. Proceedings of the National Academy of Sciences of the United States of America, 108(7), 3086–3091. doi:10.1073/pnas.1010808108 Maguire, R. O., & Sims, J. T. (2002). Soil testing to predict phosphorus leaching. Journal of Environment Quality, 31, 1601–1609. Manzoni, S., Trofymow, J. A., Jackson, R. B., & Porporato, A. (2010). Stoichiometric controls on carbon, nitrogen, and phosphorus dynamics in decomposing litter. Ecological Monographs, 80(1), 89–106. Mcbeath, T. M., Lombi, E., & Mclaughlin, M. (2004). Sorption of orthophosphate and pyrophosphate in Australian soils. 3rd Australian New Zealand Soils Conference, 5-9 December 2004. University of Sydney, Australia.  McBride, M. B. (1994). Environmental chemistry of soils. Oxford: Oxford University Press. McDowell, R. W., & Sharpely, A. N. (2003). Phosphorus solubility and released kinetics as a function of soil test P concentration. Geoderma, 112, 143–154. McDowell, R. W., & Sharpley, A. N. (2001). Phosphorus losses in subsurface flow before and after manure application to intensively farmed land. Science of the Total Environment, 278, 113–125. doi:10.1016/S0048-9697(00)00891-3 McKeague, J. A., & Day, J. H. (1966). Dithionite- and oxalate- extractable Fe and Al as aids in differentiation various classes of soils. Canadian Journal of Soil Science, 46(1), 13–22. Miyittah, M. K., Gadekar, S., Pullammanappallil, P., Stanley, C. D., Bonzongo, J. C., & Rechcigl, J. E. (2012). Application of polymath chemical equilibrium simulation model for struvite precipitation in soils. Water, Air, and Soil Pollution, 223, 1995–2005. doi:10.1007/s11270-011-1000-9 84  Neset, T. S. S., & Cordell, D. (2012). Global phosphorus scarcity: Identifying synergies for a sustainable future. Journal of the Science of Food and Agriculture, 92, 2–6. doi:10.1002/jsfa.4650 Pagliari, P. H., & Laboski, C. A. M. (2012). Investigation of the inorganic and organic phosphorus forms in animal manure. Journal of Environment Quality, 41, 901. doi:10.2134/jeq2011.0451 Pant, H. K., Nair, V. D., Reddy, K. R., Graetz, D. a, & Villapando, R. R. (2002). Influence of flooding on phosphorus mobility in manure-impacted soil. Journal of Environmental Quality, 31, 1399–1405. doi:10.2134/jeq2002.1399 Parton, W. J., Stewart, J. W. B., & Cole, C. V. (1988). Dynamics of C, N, P and S in grassland soils: A model. Biogeochem, 5(1), 109–131. Pellerin, A., Parent, L.É., Fortin, J., Tremblay, C., Khiari, L., & Giroux, M. (2006). Environmental Mehlich-III soil phosphorus saturation indices for Quebec acid to near neutral mineral soils varying in texture and genesis. Canadian Journal of Soil Science, 86(4), 711–723. doi:10.4141/S05-070 Pierzynski, G. M., Logan, T. J., & Traina, S. J. (1990). Phosphorus chemistry and mineralogy in excessively fertilized soils: Solubility qquilibria. Soil Science Society of America Journal, 54, 1589. doi:10.2136/sssaj1990.03615995005400060013x Piteau Associates. (2005). Hydrogeological Assessment for Location of Groundwater Abstraction Site, Sumas Prairie, BC. Quiquampoix, H., & Mousain, D. (2005). Chapter 5: Enzymatic hydrolysis of organic phosphorus. In B. L. Turner, E. Frossard, & D. S. Baldwin (Eds.), Organic phosphorus in the environment (pp. 89–112). Balboa, Ancon, Panama: Smithsonian Tropical Research Institute. Rayment, G. E., & Higginson, F. R. (1992). Australian Laboratory Handbook of Soil and Water Chemical Methods, Australian Soil and Land Survey Handbooks (Vol. 3). Melbourne: Inkata Press. Reijnders, L. (2014). Phosphorus resources, their depletion and conservation, a review. Resources, Conservation and Recycling, 93, 32–49. Ruttenberg, K. C. (2003). Global phosphorus cycle. In H. D. H. and K. K. T. William H. Schlesinger (Ed.), Treatise on Geochemistry, Volume 8 (pp. 585–643). Elsevier. Sakurai, K., Ohdate, Y., & Kyuma, K. (1989). Factors affecting zero point of charge (ZPC) of variable charge soils. Soil Science and Plant Nutrition, 35(1), 21–31. Schendel, E. K., Schreier, H., & Lavkulich, L. M. (2004). Linkages between phosphorus index estimates and environmental quality variables. Soil and Water Conservation Journal, 59, 243–251. Schreier, H., Bestbeir, R., & Derksen, G. (2003). Agricultural nutrient management trends in the Lower Fraser Valley of BC (Based on Agricultural Census 1991-2001). 85  Schreier, H., Hall, K. J., Brown, S. J., Wernick, B., Berka, C., Belzer, W., & Pettit, K. (1998). Agriculture: an important non-point source of pollution. In T. Gray, Colin and Tuominem (Ed.), Health of the Fraser River Aquatic System Volume II (pp. 83–99). Environment Canada. Shand, C. A., Cheshire, M. V., Bedrock, C. N., Chapman, P. J., Fraser, A. R., & Chudek, J. A. (1999). Solid-phase 31P spectra of peat and mineral soils, humic acids and soil solution components: influence of iron and manganese. Plant and Soil, 214, 153–163. Shand, C. A., Smith, S., Edwards, A. C., & Fraser, A. R. (2000). Distribution of phosphorus in particulate, colloidal and molecular-sized fractions of soil solution. Water Research, 34(4), 1278–1284. doi:10.1016/S0043-1354(99)00240-7 Shang, C., Huang, P. M., & Stewart, J. W. B. (1990). Kinetics of adsorption of organic and inorganic phosphates by short-range ordered precipitate of aluminum. Canadian Journal of Soil Science, 470, 461–470. Sharpley, A. N. (1995). Soil phosphorus dynamics: agronomic and environmental impacts. Ecological Engineering, 5, 261–279. doi:10.1016/0925-8574(95)00027-5 Sharpley, A. N., Daniel, T., Sims, T., Lemunyon, J., Stevens, R., & Parry, R. (2003). Agricultural phosphorus and eutrophication. Environmental Protection, ARS-149(September), 1–43. Retrieved from http://ddr.nal.usda.gov/dspace/bitstream/10113/26693/1/CAT30907360.pdf Sharpley, A. N., Krogstad, T., Kleinman, P. J., & Haggard, B. (2007). Managing natural processes in drainage ditches for nonpoint source phosphorus control. Journal of Soil and Water Conservation, 62(4), 197–206. Sharpley, A. N., Smith, S. J., Jones, O. R., Berg, W. A., & Coleman., G. A. (1992). The transport of bioavailable phosphorus in agricultural runoff. Journal of Environment Quality, 21, 30–35. Shead, R. (2004). Summary of surface water quality sampling on Sumas River and tributaries (pp. 1–26). Surrey, BC. Shen, J., Yuan, L., Zhang, J., Li, H., Bai, Z., Chen, X.,& Zhang, F. (2011). Phosphorus dynamics: From soil to plant. Plant Physiology, 156, 997–1005. doi:10.1104/pp.cgi/www.plantphysiol.org/Turner Shirvani, M., Shariatmadari, H., & Kalbasi, M. (2005). Phosphorus buffering capacity indices as related to soil properties and plant uptake. Journal of Plant Nutrition, 28(December 2014), 537–550. doi:10.1081/PLN-200049235 Sims, J. T., Maguire, R. O., Leytem, A. B., Gartley, K. L., & Paulter, M. C. (2002). Evaluation of Mehlich 3 as an agri-environment soil phosphorus test for the Mid-Atlantic United States of America. Soil Sci. Soc. Am. J, 66, 2016–2032. Singh, B. B., & Jones, J. P. (1976). Phosphorous sorption and desorption characteristics of soil as affected by organic residues. Soil Sci Soc Am J, 40(3), 389–394. 86  Sjoerd, E. A., Van Der Zee, T. M., & Van Riemsduk, W. H. (1988). Model for long-term phosphate reaction kinetics in soil. Journal of Environment Quality, 17, 35–41. Smeck, N. E. (1985). Phosphorus dynamics in soils and landscapes. Geoderma, 36(39), 185–199. doi:10.1016/0016-7061(85)90001-1 Smil, V. (2000). Phosphorus in the environment: Natural flows and human interferences. Annual Review of Environment and Resources, 25, 53–88. Smith, K. A., Chalmers, A. G., Chambers, B. J., & Christie, P. (1998). Organic manure phosphorus accumulation, mobility and management. Soil Use and Management, 14, 154–159. doi:10.1111/j.1475-2743.1998.tb00634.x Snape, I., Scouller, R. C., Stark, S. C., Stark, J., Riddle, M. J., & Gore, D. B. (2004). Characterisation of the dilute HCl extraction method for the identification of metal contamination in Antarctic marine sediments. Chemosphere, 57, 491–504. Sollins, P., Robertson, G. P., & Uehara, G. (1988). Nutrient mobility in variable- and permanent-charge soils. Biogeochemistry, 6, 181–199. doi:10.1007/BF02182995 Statistics Canada. (2015). 2011 Census of Agriculture. Retrieved May 1, 2015, from http://www.statcan.gc.ca/eng/ca2011/index Stewart, J. W. B., & Sharpley, A. N. (1987). Controls on dynamics of soil and fertilizer phosphorus and sulfur. In R. F. Follert, J. W. B. Stewart, & C. V. Cole (Eds.), Soil Fertility and Organic Matter as Critical Components of Production Systems (pp. 101–121). Madison, Wisconsin: Special Pub. No. 19, American Society of Agronomy. Syers, J. K., Johnston, a. E., & Curtin, D. (2008). Efficiency of soil and fertilizer phosphorus use (pp. 01 – 63). The Clay Minerals Society. (1987). Influence of relative humidity on the reaction products of phosphates and noncyrstalline hydroxides of aluminum and iron. Water, 35(3), 228–231. Tiessen, H., Stewart, J. W. B., & Cole, C. V. (1984). Pathways of phosphorus transformations in soils of differing pedogenesis. Soil Science Society of America Journal, 48, 853. doi:10.2136/sssaj1984.03615995004800040031x Tran, T. S., & Simard, R. R. (1993). Mehlich III-extractable elements. In M. R. Carter (Ed.), Soil Sampling and Methods of Analysis (pp. 43–49). Travis, C. C. (n.d.). Mathematical description of adsorption and transport of reactive solutes in soil: A review of selected literature. Turner, B. L., Papházy, M. J., Haygarth, P. M., & McKelvie, I. D. (2002). Inositol phosphates in the environment. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 357, 449–469. doi:10.1098/rstb.2001.0837 87  Ulén, B., & Snäll, S. (2007). Forms and retention of phosphorus in an illite-clay soil profile with a history of fertilisation with pig manure and mineral fertilisers. Geoderma, 137, 455–465. doi:10.1016/j.geoderma.2006.10.003 Van Kauwenbergh, S. J. (2010). World phosphate rock reserves and resources (pp. 1–48). Muscle Shoals, Alabama. Walker, T. W., & Syers, J. K. (1976). The fate of phosphorus during pedogenesis. Geoderma, 15, 1–19. doi:10.1016/0016-7061(76)90066-5 Wang, Y. T., Zhang, T. Q., Hu, Q. C., Tan, C. S., O’Halloran, I. P., Drury, C. F.,& Welacky, T. (2010). Estimating dissolved reactive phosphorus concentration in surface runoff water from major Ontario soils. Journal of Environment Quality, 39, 1771–1781. Whatcom County Public Works. (2012). Swift Creek sediment management action plan (pp. 1–70). Whittaker, R. H. (1954). The ecology of serpentine soils. Ecology, 35(2), 258–288. Yang, X., & Post, W. M. (2011). Phosphorus transformations as a function of pedogenesis: A synthesis of soil phosphorus data using Hedley fractionation method. Biogeosciences, 8, 2907–2916. doi:10.5194/bg-8-2907-2011 Yu, W., Ding, X., Xue, S., Li, S., Liao, X., & Wang, R. (2013). Effects of organic-matter application on phosphorus adsorption of three soil parent materials. Journal of Soil Science and Plant Nutrition, 13, 1003–1017. doi:10.4067/S0718-95162013005000079 Yuan, G., & Lavkulich, L. M. (1995). Environmental phosphorus indices in manure amended soils in the Fraser basin of British Columbia, Canada. J. Environ. Sci. Health, B30, 841–857. Zelazny, L. W., & White, G. N. (1989). Chapter 11. The pyrophyllite-talc group. In J. B. Dixon & S. B. Weed (Eds.), Minerals in Soil Environments (Second Edi, pp. 527–550). Madison, Wisconsin: Soil Science Society of America.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0166274/manifest

Comment

Related Items