UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural stability and surface sealing as related to organic matter depletion of a shallow organic… Bonsu, Mensah 1987

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

Item Metadata

Download

Media
831-UBC_1987_A1 B66.pdf [ 5.54MB ]
Metadata
JSON: 831-1.0097102.json
JSON-LD: 831-1.0097102-ld.json
RDF/XML (Pretty): 831-1.0097102-rdf.xml
RDF/JSON: 831-1.0097102-rdf.json
Turtle: 831-1.0097102-turtle.txt
N-Triples: 831-1.0097102-rdf-ntriples.txt
Original Record: 831-1.0097102-source.json
Full Text
831-1.0097102-fulltext.txt
Citation
831-1.0097102.ris

Full Text

S T R U C T U R A L S T A B I L I T Y A N D S U R F A C E S E A L I N G A S R E L A T E D TO ORGANIC  MATTER DEPLETION  OF A S H A L L O W ORGANIC  by MENSAH  A THESIS SUBMITTED  BONSU  IN PARTIAL F U L F I L M E N T OF  THE REQUIREMENTS DOCTOR  FOR THE DEGREE OF  OF  PHILOSOPHY  in F A C U L T Y OF GRADUATE STUDIES Department of Soil Science  We accept this thesis as conforming to the required  THE UNIVERSITY  O F BRITISH  April  ©  standard  COLUMBIA  1987  Mensah Bonsu, 1987  SOIL  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the The University of British Columbia, I agree that the Library shall make it freely available for reference for  and study. I further  agree that permission  extensive copying of this thesis for scholarly, purposes may be granted by the  Head of my Department or by his or her representatives. copying  or  publication of  this  thesis  without my written permission.  Department of Soil Science The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V 6 T 1W5  Date: April  1987  for  financial  gain  It is understood that shall  not  be allowed  ABSTRACT A physically based model describing the mechanism of surface  sealing of  soil was evaluated in the context of aggregate stability. The intent of the model study was to better understand the effect of mixing Fine-textured mineral subsoil with organic surface soil on structural stability and surface seal formation. The mixing results from tillage and harvesting operations,  and management practices  such as levelling. The index derived from the model showed that sealing of the shallow organic soil increased with an increase of mineral matter content. The mathematical formulation of the model was based on the principle of conservation of mass and Darcy's law for flow of water through a layered soil column. Assuming convective flow, it was shown theoretically that the rate of surface  seal formation is proportional to the flux density of the filtrate,  assumed by Scheidegger (1974). In the model it was further pore necks at the soil surface  assumed  clog first before the seal develops. The  as  that the assumption  that convective flow alone was responsible for the movement of the suspension is likely incorrect for suspensions derived from medium or coarse textured soils, since sedimentation  does influence the movement of larger particles. However,  introducing a constant sedimentation  parameter into the convective flow model did  not improve the model. Therefore, it is likely a non-constant  sedimentation  parameter could improve the model considerably. The model showed that for sufficiently large times the flux density of a filtrate flowing through a soil column at a constant hydraulic head is proportional to inverse square root of time. Testing the model experimentally showed a good agreement between  theory and experiment.  A highly significant correlation  between  the soil stability factor derived from the model and aggregate stability suggests that the index is a soil structural  attribute. The soil stability factor  was  exponentially related with aggregate stability and mineral matter content.  ii  However, whereas the relationship between the soil stability factor  and aggregate  stability gave a positive exponent, a negative exponent was obtained with mineral matter content. Further  studies showed that structural  stability and saturated hydraulic  conductivity of the aggregate beds were positively and significantly  correlated  exponentially. However, saturated hydraulic conductivity and mineral matter content were negatively  and significantly correlated  exponentially. Collateral to the  results of the model, the strong negative correlation between wet-sieved  aggregate  stability and mineral matter content confirmed the deleterious effect of mixing fine-textured  mineral soil on the structure of the shallow organic soil.  It was theorized that aggregates stabilized through clay-organic complexing are likely to be much stronger than aggregates stabilized through other mechanisms.  This implies that whenever the mineral matter content is much  higher than the organic matter content, the surplus mineral matter that does not interact with organic matter will be most dispersible. The high silt content of the mineral matter fraction is likely to be an important factor contributing to the decrease in structural  stability with increasing mineral matter content. Once the  clay and the organic colloids have interacted, the silt that remains is not capable of forming stable aggregates without colloids (Baver et al. 1972). From measurements of the air to water permeability ratio, the decrease in saturated hydraulic conductivity of the aggregate beds with increasing  mineral  matter content was attributed to slaking of the mineral matter fraction.  However,  it is possible for the soil with high mineral matter content to be stable if the mineral matter is allowed to be in contact with the organic matter for a long period of time.  iii  Table of Contents ABSTRACT  ii  LIST O F T A B L E S  vi  LIST O F FIGURES  viii  LIST OF S Y M B O L S  xi  ACKNOWLEDGEMENT  xiii  1.  INTRODUCTION  1  2.  A P H Y S I C A L L Y BASED M O D E L FOR S U R F A C E SEALING OF A S H A L L O W O R G A N I C SOIL  5  2.1 I N T R O D U C T I O N  5  2.2 M O D E L D E S C R I P T I O N  9  2.2.1 Surface Seal Formation Under Convective Flow 2.2.2 Surface Seal Formation under Convective Flow and constant Sedimentation 2.3 M A T E R I A L S A N D M E T H O D S  3.  9 17 19  2.3.1 Experimental Verification of the Model  19  2.3.2 Application of the Model  23  2.4 R E S U L T S A N D D I S C U S S I O N  30  2.4.1 Validity of the Theory  30  2.4.2 Relating The Soil Stability Factor To Aggregate Stability,and Mineral Matter Content of a Shallow Organic Soil  56  2.4.3 Validity of the Procedure  72  2.4.4 Variability within Replicates of The Soil Stability Factor for The Shallow Organic Soil  74  2.5 C O N C L U S I O N S  75  HYDRO-PHYSICAL IMPORTANCE OF ORGANIC M A T T E R D E P L E T I O N O F A S H A L L O W O R G A N I C SOIL  77  3.1 I N T R O D U C T I O N  77  3.2 M A T E R I A L S A N D M E T H O D S  81  3.2.1 General Procedure  81  iv  3.2.2 Determination of 0. I N N a O H Extractable Humic Substances. 3.2.3 Determination of A i r to Water Permeability Ratio  82  3.3 R E S U L T S A N D D I S C U S S I O N S  86  3.3.1 Relating Aggregate Stability to Mineral Matter Content 3.3.2 Relating Saturated  Hydraulic Conductivity of the  86  Aggregate  Beds to Aggregate Stability and Mineral Matter Content  4.  ...81  94  3.4 C O N C L U S I O N S  110  SUMMARY  Ill  LIST O F R E F E R E N C E S  114  APPENDIX  1  119  APPENDIX  2  121  APPENDIX  3  122  APPENDIX  4  123  APPENDIX  5  124  v  LIST OF T A B L E S Tables Page 2.1  Regression parameters for the relationship between log q s-  and log t 2.2  The parameters used to estimate q  32 33  S  2.3  2.4  2.5a  q  values measured and predicted (convective flow only) for s the Haney clay suspension without dispersing agent values measured and predicted (convective flow only) for s the kaolinite suspension with dispersing agent  35  q  q  s  36  values measured and predicted (convective flow only and  combined convective flow and constant sedimentation) for the Westham silt loam suspension without a dispersing agent 2.5b  q  s  values measured and predicted (convective flow only) for  the Westham silt loam suspension with dispersing agent 2.6  q  s  38  39  values measured and predicted (convective flow only and  combined convective flow and constant sedimentation) for the Abbotsford loam suspension without dispersing agent 2.7  40  Equivalent spherical diameter and corresponding Stokes velocity and time required for particles to fall 5 cm in a suspension  2.8  50  Non-parametric Mann-Whitney U-test for comparing measured and predicted (convective flow only) q  2.9  The factor 2 b K , ( H - H )  2.10  Mean aggregate stability, organic matter and mineral  2  1  2  t/L K  values s  2  z  matter content along the transects 3.1  Bulk density of aggregate beds, particle density, and vi  52 54  60  porosity of the aggregate beds 3.2  A i r and water permeabilities, air to water  99 permeability  ratio, and relative swelling of the aggregate beds  vii  101  LIST OF FIGURES  Figures Page 2.1  A schematic diagram illustrating surface sealing of soil in a column. 10  2.2  A schematic diagram of the apparatus used for testing the model.... 20  2.3  Scheme of sampling and location of sampling site on the V a n Halst F a r m  2.4  24  A schematic diagram of the apparatus used for determining soil stability factor of the shallow organic soil  2.5  26  Variation of flux density (q ) with constant sedimentation s rate ( k ) at different times for the Westham silt loam 2  suspension without dispersing agent 2.6  41  Variation of flux density (q ) with constant sedimentation rate ( k ) at different times for the Abbotsford loam 2  suspension without dispersing agent 2.7  The log-log relationship between flux density measured  43 and  predicted by convective flow only and time for the Westham silt loam suspension 2.8  The log-log relationship between flux density (measured  44 and  predicted by convective flow only) and time for the Haney clay suspension 2.9  45  The log-log relationship between flux density (measured  and  predicted by convective flow only) and time for the kaolinite suspension 2.10  46  The log-log relationship between flux density (measured  and  predicted by convective flow only and combined convective viii  flow and constant sedimentation) and time for the Abbotsford loam suspension without dispersing agent 2.11  The log-log relationship between flux density (measured predicted by combined convective flow and  48 and  constant  sedimentation) and time for the Westham silt loam suspension without dispersing agent  49  2.12  Organic layer depth along the three transects  57  2.13  Mean aggregate stability, organic matter content, and mineral matter content along the transects  2.14  59  Variation of flux density with inverse square-root of time for sample no. 4 which was 25m from the mineral soil ridge  2.15  61  - Variation of flux density with inverse square-root of time for sample no.8 which was 175m from the mineral soil ridge  2.16  Mean r  62 2  values and standard deviations associated with q s  versus 2.17  1//1  63  Probability distribution for which r  2  is significant for the  relationship q 2.18 2.19  versus 1//1 s Variability within E of the model a The scatter diagram showing the variation of E  64 65 a  with  distance from the mineral ridge Actual soil stability factor (E ) versus expected soil stability a factor (E ) x  67  2.21  Soil stability factor versus wet-sieved aggregate stability  70  2.22  Soil stability factor versus mineral matter content  71  3.1  Schematic diagram of the apparatus for the determination  2.20  ix  68  of air permeability  83  3.2  H C l digestible ash versus organic matter content  87  3.3  0.1N N a O H extractable fraction of 3g subsample versus organic matter content  89  3.4  Aggregate stability versus mineral matter content  90  3.5  Variation of K transects  of aggregate beds with distance along the s  95  3.6  K  3.7  K  3.8  Variation of air to water permeability ratio with distance  s s  of aggregate beds versus aggregate stability  97  of aggregate beds versus mineral matter content  98  from the mineral soil ridge  103  3.9  A i r to water permeability ratio versus aggregate stability  104  3.10  A i r to water permeability ratio versus mineral matter content  3.11  105  Organic matter content versus relative swelling of the aggregate beds  107  x  LIST OF SYMBOLS Symbols  Explanation Units  A  Area of column  m  a  Constant coefficient of proportionality  m/s  b C  A dimensionless constant given by (dp ) b A n integration constant  c  Concentration of suspension  d  = bK(H -H ) z 1  E E  + LK k K z 1  _  2  1/2  (m/s)  Actual soil s t a b i l ^ factor  a  3  m /s  1  2  Soil stability factor defined as (K /2b) z  (m s)  1/2  1/2  /day  1/2  E  Expected soil stability factor  g H  2  kg/m  2  (m s)  Acceleration due to gravity  9.81  Head of suspension at t = 0  0  11,-112  /day m/s  2  m  Constant hydraulic head difference  across the column  m  K, K z k  Saturated hydraulic conductivity of soil below the seal Saturated hydraulic conductivity of the seal  m/s m/s  A i r permeability  m  2  k  Water permeability  m  2  Constant sedimentation rate  m/s  k  w 2  L  Length of soil below the seal  m  L,  Length of soil before swelling  m  L  Length of soil after swelling  m  dM  Elemental mass of particles trapped in the soil  kg  dM  Elemental mass of solids lost from suspension  kg  Elemental mass of solids accumulated  kg  2  dM p  z  Porosity of the porous medium  xi  in the seal  Pa  p  Atmospheric pressure  Q  Volume outflow of water  m  q s t  Flux density of filtrate  m/s  3  s  Time of flow  dt  Elemental change in time  V  Volume of tank  z  Thickness of seal  s m  m m  Elemental thickness of seal  dz AGG MN  3  %  Wet-sieved aggregate stability  %  Mineral matter content  %  Bulk density of seal  kg/m  3  P  Particle density  kg/m  3  kg/m  3  s  P  Density of water  w  \  Viscosity of air  kg m"  1  s"  1  K  Viscosity of water  kg m"  1  s"  1  AH  T  Hydraulic head drop across the seal  xii -  m  ACKNOWLEDGEMENT The  funds used to carry out this work were provided by the British  Columbia Ministry of Agriculture and Food. The financial assistance received is fully acknowledged. I am very grateful to Dr. Jan de Vries, my research supervisor, for his interest and contribution that helped to make this work a success. The  contribution and suggestions provided by the other members of my  research supervisory committee, Drs. T. A . Black, M . D. Novak, and J . C. Keng, that helped to make this work meet its present form, are well appreciated. Many thanks go to M r . V a n Halst, whose farm was used as the study site for this research. His interest in this research and the generosity he showed at all times are very much appreciated. The  bursary I received from the Award Office of the University of British  Columbia, at a time when I was in financial difficulty, enabled me to stay on to complete this work. Many thanks to the Director of Awards for this great help. I thank M r . Bernie Von Spindler, who drew some of the diagrams in this thesis.. His willingness to help at any time needs to be commended. I am extremely grateful to Dr. L . M . Lavkulich, the Head of the Department of Soil Science, University of British Columbia, for the interest he always showed in resolving my personal problems whenever I was confronted with one. His approachability is a trait that is highly commendable. I express my gratitude to my wife and children for their patience that encouraged me to complete this work. I give thanks to God for His help and guidance throughout the course of  xiii  1. I N T R O D U C T I O N The shallow organic soils of the Serpentine-Nicomekl area of the Lower Fraser Valley are important to the economy of British Columbia. These organic soils cover an area of about 1,000 ha, comprising about 0.1% of the total area of improved farmlands in British Columbia. Yet about 5% of the British Columbia farm cash receipts is generated from these organic soils (De Vries, 1983). Poor structural and hydrologic characteristics are a constraint to the management  of these shallow organic soils (Bonsu, 1984). Therefore there is a  need for a better understanding of the processes leading to their undesirable structural and hydrologic characteristics. Soil structure may be defined as the arrangement of soil primary particles into stable aggregates and the associated pore spaces. The main types of pore systems in the soil are the macropores, micropores, and features  such as planar  voids due to cracks and biopores from dead root channels and wormholes. The macropores, planar voids, and biopores are considered as the water transmission pores, and the micropores as the water retention pores. A good balance of air and water in soil is critical for good performance of most arable crops. During rainfall excess water must drain in order to be replaced by air. Therefore the agronomic importance of good soil structure cannot be overemphasized. Good soil structure provides good soil tilth and improves the soil's resistance to erosion. A soil may be classified as having good structure if it is aggregated  and  stable. Aggregate or structural stability is a relative term that has been used to describe the resistance of soil structure to some destructive forces. For example, resistance to dispersion by water (Yoder, 1926), and resistance to waterdrop impact (McCalla, 1944). Information on the different methods used for assessing aggregate stability of soil is well documented ( e.g. Williams et al. 1966). To date, all the known methods used for assessing the stability of soil aggregates to  1  2 water are empirically-based. Recognizing the importance of structural stability of soil, there is a need for a physically based method to quantify the stability of soil  structure. When the soil structure is degraded the production potential of the soil  decreases because the optimum physical conditions are not satisfied. Ponding of water during spring, at the beginning of the growing season, is notorious on the shallow organic soils of the Serpentine-Nicomekl area of the Lower Fraser Valley due to structural degradation. Ponding that occurs in spring leads to a delay in seedbed preparation and consequently, late seeding and a loss of "opportunity days" to the farmer. Therefore, in 1984 a project funded by the British Columbia Ministry of Agriculture and Food was initiated to study the ponding problem. Bonsu (1984) in his M.Sc. thesis recognized two types of structural degradation that were associated with ponding of water on these shallow organic soils. The existence of compacted zones below the cultivation layer and the tendency of the organic soils to form a surface seal of low infiltrability were the two major structural problems observed (Bonsu, 1984). The particular soil studied in this work occurs on the V a n Halst Farm near Cloverdale in British Columbia. The geographical location and the aerial photograph showing the V a n Halst Farm are presented elsewhere (Bonsu, 1984). The soil belongs to the Vinod series which has an organic layer that is either less than or equal to 0.40 m deep. The soil is classified as Rego Gleysol (peaty phase) in the Canadian System of Soil Classification (Luttmerding, 1981). A detailed description of the soil, geology, and hydrology of the area is presented elsewhere (Bonsu, 1984). The soil is intensively used for growing vegetables. A t the site where the soil was sampled for this work, the depth of the organic layer ranges from zero on a mineral soil ridge to about 40 cm in the adjacent depression. The mineral  3 soil ridges were formed surficially by marine deposits that gave rise to hummocks and depressions before the organic soil began to form on top (Lavkulich, personal communication). Subsidence, long-term cultivation, and in some cases, levelling were responsible for the exposure of the ridges. In areas where the organic layers are shallow, the underlying mineral soil has been brought up during tillage and harvesting operations, and mixed with the organic layer. Also when the soil is levelled mineral matter is moved from the mineral soil ridges to the organic depressions, increasing the mineral matter content of the organic layers in the depressions. The overall objectives of this thesis are three-fold: 1.  to formulate and test experimentally a physically based model describing the mechanism of surface sealing of soil in a column,  2.  to relate the index derived from the model to the structural stability of a shallow organic soil, and  3.  to explore the effects of the mixing of mineral matter with the organic layer on structural stability and saturated hydraulic conductivity of a shallow organic soil located in the Serpentine -Nicomekl area of the Lower Fraser Valley of British Columbia. Chapter 1 reports on the mathematical formulation of the model,  experimental verification of the model, and the regressions and correlations of the index derived from the model with structural stability, and mineral matter content. In chapter 2 , the declining stability of the structure of the shallow organic soil due to management-induced mixing of mineral matter is described by establishing relationships between aggregate stability and mineral matter content using a regression approach. Also, the hydrologic importance of organic matter depletion due to management-induced mineral matter mixing is discussed by  4 establishing a relationship between saturated hydraulic conductivity of the aggregate beds and aggregate stability and mineral matter content using a regression approach. The differences in saturated hydraulic conductivity values of the aggregate beds are discussed relative to their slaking tendencies using the concept of air to water permeability ratio.  2. A P H Y S I C A L L Y B A S E D M O D E L F O R S U R F A C E S E A L I N G O F A S H A L L O W O R G A N I C SOIL  2.1 I N T R O D U C T I O N The importance of stable soil structure for optimum soil tilth (Baver et al. 1972) cannot be disputed. A n unstable soil tends to slake in water, the dispersed soil particles blocking the pores and ultimately sealing the soil surface. When the soil surface is sealed the infiltrability of the soil is reduced,. which can lead to overland flow and soil erosion on sloping land. When the land surface is horizontal and relatively flat, ponding is often the consequence of surface sealing (De Vries, 1983; Bonsu, 1984). Ponding of water on arable lands during the growing season is a major cause of a loss of "opportunity days" to the  farmer.  In the Lower Fraser Valley, this problem is particularly serious in spring. In addition, when the surface seal dries it forms a crust. The formation of a crust resulting from surface sealing can inhibit seedling emergence  (Richards, 1953).  Hence a knowledge of the mechanism of surface sealing is important for the management  of soil in terms of its structure  and hydrology.  The mechanism of pore-clogging and surface sealing in soils has not been extensively explored. Conceptually, pore-clogging may be defined as the blocking of the pore networks of the soil due to movement of dispersed soil particles. Pore-clogging, indeed, is a precursor of surface sealing. The soil pores must be clogged before a surface seal can form. Dispersed soil particles may migrate with the flow of water to clog the pore-necks in the soil body. The particles are likely to move to clog the pores in the soil body only if the dispersed particles are smaller relative to the size of the pores immediately at the soil surface. Otherwise, energy is required to move the particles down (Bertrand and Kamil, 1962). Using clays labelled with  5  Rb  8 6  6  Bertrand and Kamil (1962) observed that about 1 percent of the clay migrated to a depth of 3 cm in soil columns following artificial application of rainfall. A t about a depth of 7.5 cm the radioactivity of the soil was equal to background, indicating the absence of migrated clay. In addition they found that the clay content, specific surface, aggregate stability, and organic matter content of the surface soil decreased with the artificially applied rainfall. In a similar study Kamil and Bertrand (1962) observed that the application of artificial rainfall did not cause any change in the hydraulic conductivity of a quartz sand, but the hydraulic conductivity of the top 1.5 cm of a silt loam soil decreased appreciably. They attributed the reduction in hydraulic conductivity of the silt loam soil to the breakdown of the soil aggregates, pore-clogging, and  surface  sealing due to particles moving with water through the soil. Suspended clay particles following ploughing and puddling of field soils can result in the formation of cutans (Bouma 1969).[Cutans are coatings of clay on peds, which are natural aggregates separated  from each other by natural  planes  of weakness (Fairbridge and Finkl,1979)]. Under field conditions Bouma (1969) observed that when a suspension of clay particles was sucked into the soil matrix, the clay settled on the walls of the pores and blocked them. Whenever a dispersing agent is passed through a soil, the dispersed particles resulting from aggregate collapse can block the larger pores to result in reduced saturated  hydraulic conductivity (Chen and Banin, 1975). This continuous  network of dispersed particles can be made visible with a scanning electron microscope observation (Chen and Banin 1975). Shainberg and Singer (1986) reported on the effects of suspension concentration and concentration of chloride solutions having sodium adsorption ratio (SAR) of 4.0 on hydraulic conductivity of depositional crusts on soil columns. They used the ratio of suspension flux to the flux of clear solution as a measure of the effect of sediment deposition on the  7  hydraulic properties of the column. They observed that the ratio decreased as the concentration of solids increased, indicating a decreasing flux. In addition, they noted that the ratio depended on both the crust thickness and structure.  The  hydraulic conductivity of the depositional crust formed when the solids were dispersed with 0.001M and 0.005M chloride solutions having S A R of 4.0 was respectively 2 and 3 orders of magnitude less than the bulk soil. The physics of pore-clogging and surface sealing in soils has not been studied to any appreciable detail. However, direct observations have shown that dispersed soil particles that migrated with the flow of water are responsible for pore -clogging and surface sealing in soils. The quantity of particles dispersed due to some sort of mechanical abrasion (Dong et al. 1983) and the extent of pore-clogging and surface sealing (Bertrand and Kamil,1962) can be related to the structural stability of the soil. Some empirical equations expressing hydraulic conductivity reduction during surface sealing of soil have been proposed (Van Doren and Allmaras, 1978; Linden, 1979; Moore et al.1980; Brakensiek and Rawls, 1983). Van Doren and Allmaras (1978) proposed the following empirical exponential equation: K = K . exp(— S I X S U ) , where K is the hydraulic conductivity of the sealed surface  (cm/hr), K . is the initial hydraulic conductivity  of the surface soil (cm/hr), SI is the structural stability constant (cm /J), and 2  S U is the cumulative rainfall kinetic energy (J/cm ). 2  Linden (1979), using a modified form of the equation due to V a n Doren and Allmaras (1978), expressed empirically the change in hydraulic conductivity of a surface seal with time as: K = K exp(— SE) +0.02, where K is the t b t hydraulic conductivity of the surface seal at time t (cm/hr), K is the initial hydraulic conductivity of the soil surface (cm/hr), S is the soil stability factor (cm /J), and E is the cumulative rainfall energy (J/cm ). 2  2  8 Moore et al. (1980) also proposed an exponential decay function to express the change in hydraulic conductivity of a surface seal with time as: (K.—Kj)exp(—  at)  =  where K^. is the final saturated hydraulic conductivity of  a well-established stable seal (cm/hr), and a is a constant (hr" ' ) , and t is time (hr). Brakensiek and Rawls (1983) combined the equations due to Linden and Moore et al. (1980) to yield the empirical equation: K c K^)exp( — CE), where K  = K + f  (1979)  (K — o  is the hydraulic conductivity of the seal at time t, K  is the initial hydraulic conductivity of the surface soil, E is the cumulative rainfall energy (J/cm ), and C is a constant (cm /J). 2  2  The information obtained from the literature indicates deficiencies in our knowledge of the physical mechanism of surface  sealing as part of the soil water  hydraulic system. Most approaches to describing the mechanism of surface  sealing  of soil have been based on empiricism. Brakensiek and Rawls (1980) have already expressed concern about these gaps. Therefore, there is a need for a physically based equation to describe the mechanism of surface sealing of soils. In addition, the empirical relationships presented in the literature suggest that soil structure stability is a key factor when considering the mechanism of surface sealing of soils. As pointed out in the introduction to this thesis, surface sealing is an important aspect of the problem associated with the management of the shallow organic soils of the Serpentine-Nicomekl area of the Lower Fraser Valley. This study was therefore 1.  undertaken  with the objective of:  formulating and testing experimentally a physically based model describing the mechanism of surface sealing of soil in a laboratory column, and  2.  relating the index derived from the model with structural stability and mineral matter content of the shallow organic soil.  9  2.2 M O D E L  DESCRIPTION  2.2.1 S U R F A C E S E A L F O R M A T I O N U N D E R C O N V E C T I V E Consider a soil as a  filtering  FLOW  medium through which a suspension of  dispersed soil particles is flowing in response to a constant hydraulic gradient across a column (Fig.2.1). Depending on the mass concentration of the suspension, a surface  seal will develop as the suspended particles reach the soil  surface. Let the following assumptions be proposed with respect to the mechanism of surface sealing: The pore necks at the soil surface  1.  are clogged first by the dispersed solids  before the seal starts to develop on top of the soil surface, giving rise to a two-layered system. This implies the layer below the seal is unaltered. The suspended particles reache the soil surface by convection only. This  2.  implies sedimentation is negligible. It is implicit from this assumption that during the process of sealing, the concentration of the suspension is constant. Consider that an element of seal thickness dz has formed on a  constant  surface area A of a uniform soil in a laboratory column in time dt, due to the movement of filtrate in response to a constant hydraulic gradient (Fig.2.1). If dM dM  z s  is the elemental mass of solids accumulated in the seal in time dt, and is the elemental mass of solids that has been lost from the suspension in  time dt, then by the principle of conservation of mass, we can write  d M /dt z  = dM /dt s  (1.1)  Eq.1.1 means that the rate of mass accumulation of solids in the seal is equal  Fig. 2.1. A schematic diagram i l l u s t r a t i n g s u r f a c e s e a l i n g of s o i l i n a c o l u m n .  to the rate of mass depletion of solids in the suspension. Equivalently, Eq. 1.1 can be written as  d M /dt = cdV /dt z sus  (1.2)  where c is the mass concentration of the suspension (kg/m ), and dV /dt is sus the volume flow rate of the suspension ( m s~ ). 3  3  1  If p is the bulk density of the seal that has formed (kg/m ), then we b 3  can write Eq. 1.2  as  p,Adz/dt = cdV /dt b sus  (1.3)  Assuming the volume of solids in the suspension is negligible compared to the volume of water (V ), then we can write w  dV  sus  /dt = dV /dt w  (1.4)  where dV /dt is the volume flux of the filtrate. It follows from Eq. 1.3 that w  dz/dt =  where q  s  is the instantaneous  (c/pj q b s  (1.5)  volumetric flux densitv of the filtrate. In Eq.1.5  dz/dt is the rate of seal thickness increase. From the assumption that the  12 concentration c of suspension is constant, we can replace (c/p ) by a constant b. b Hence, we finally see that  dz /dt =  (1.6)  bq  s  so that the rate of surface seal formation is proportional to the flux density of the  filtrate. Eq.(1.6) was assumed by Scheidegger (1974) in his theory of  filtration.  Scheidegger (1974) proposed that "the rate of surface deposition of material on a filtering  medium is proportional to the throughput of the  filtrate".  Swartzendruber and Uebler (1982) studied pore-clogging and surface sealing in sand and sand-silt media due to the passage of kaolinite and  sewage  suspensions by assuming that the mass rate of particle trapping is directly proportional to the mass flow rate of the suspension and inversely proportional to the mean pore velocity. The assumption they used in formulating their theory was mathematically stated as  dM/dt = [ac(dV/dt)]/[(dV/dt)/Ap]  (1.7)  where dM/dt is the mass rate of particle trapping in the soil, c is the mass concentration of suspension, A is the cross-sectional area of the soil, p is the porosity of the porous medium at the point of particle trapping, and a is a constant coefficient of proportionality, and [(dV/dt)/Ap] is the mean pore velocity. On cancelling like terms Eq. 1.7  becomes  13  (1.8)  dM/dt = acAp  A l l the terms on the right-hand side of Eq. 1.8 are constant, implying dM/dt is a constant according to Swartzendruber =  and Uebler (1982). Since dM/dt  p Adz/dt, comparison of Eqs. 1.5 and 1.8 implies that for surface b  sealing  (1.9).  ap  From Eq. 1.9 we see that either a or p should be a variable, since during surface  seal formation q  is a variable decreasing over time. Therefore  the  s theory developed by Swartzendruber  and Uebler (1982) for surface  seal formation  requires a modification. Dimensional analysis of Eq. 1.6 reveals that mathematical^  b is  dimensionless, but physically it represents the volume of seal formed per unit volume of suspension depleted. Under conditions where an eroding force is applied to cause soil dispersion, c could be used to designate the dispersibility of the soil. Under conditions where a constant reproducible, eroding force is applied to cause dispersion of different soils, c could be used to compare dispersibilities. The larger the value of c the lower will be the  their structural  stability. With the formation of a seal of thickness z in time t, the column can now be viewed as two-layered, and the flux density of the filtrate through  the  composite column at time t can be written as (Hillel,1980)  q  s  =  ( H , - H ) / [z/K z 2  +L/KJ. 1  (1.10)  14  in which ( H , K  z  H ) is the constant hydraulic head difference across the column; 2  is the saturated hydraulic conductivity of the seal; K  1  is the saturated  hydraulic conductivity of the soil below the seal; and L is the length of the soil below the seal. Combining Eq.1.6 and Eq.1.10 we can write  ( H , - H ) K / ( z + L K IK) z z 1  =  2  (l/b)(dz/dt)  (1.11).  Therefore,  ( z + L K / K )dz = b ( H , z 1  H , ) K dt z  (1.12).  Integrating Eq.1.12 gives  z / 2 + Lz(K / K ) = b ( H , - H , ) K t + C z 1 z  (1.13)  2  where C is the integration constant. The value of C is zero, since at t=0,  z + 2Lz(K/K) z 1 2  -  z=0, so that Eq. 1.13  2b(H,-  Eq.1.14 is quadratic; solving for positive z gives  H )K t = z 2  0  becomes  (1.14).  15 z = L K K' z 1  1  {(l + 2 b K ( H 1  H ) t/L K ) z  2  1  2  1  1/2  2  -1}  (1.15).  Substituting for z in Eq. 1.10 using Eq. 1.15 we obtain  q  s  1/2  =  ( H , - H ) / t(L K ' ) { l + 2bK 1  2  1  H )t/L  K  2  2  be approximated  (H, - H ) t / L K } z 2  2  ]  (1.16).  For t sufficiently large and K (H,-  2  1  z  much smaller than K  z is expected to be much greater than  the factor 2bK 1 1 1, so that Eq.1.16 can 2  as  q  s  =  (K/2b) z  1 / 2  (H,-H )  1 / 2  2  /t  (1.17).  1 / 2  With reference to Eq.1.16, we see that at t=0,  q  = K ( H , - H )/L, 2  s  as  1  expected physically. If Eq.1.16 is valid, then at large times a plot of q  against  1/2 1/t should give a straight line passing through the origin, but abruptly deviating from the horizontal line K ( H - H ) / L as t tends to zero. 1  Since K  z  and b are constants, Eq.1.17 can be written as  q  where  2  s  = E(H,~ H ) 2  1/2  It  1/2  (1.18)  16  E  = (K /2b) z  We may rewrite Eq. 1.18  log q  1/2  = (K pJ2c) z b  (1.19).  m  in a logarithmic form as  = log { E ( H , - H ) 2  1 / 2  }~  1/2 log t  (1.20).  Again, if the theory is valid, a plot of log q against log t should give a s straight line with a slope of -0.5. The magnitude of the constant E depends on the saturated hydraulic conductivity of the seal, the bulk density of the seal, and the concentration of the suspension. Smaller values of E imply higher c , lower K , and higher p . z b A very low K implies a more compact seal, and, therefore, a relatively higher z bulk density. Assuming that at very low K values, the occurrence of p in the z b numerator of Eq. 1.9 becomes redundant. Then it is logical to infer that the finer the dispersed particles the lower the value of K  z  is expected to be when  the dispersed particles move to form the seal. The more impervious the seal that forms and the higher the concentration of the suspension the less stable the soil. Therefore, the constant E is an attribute of soil structure and it may be designated as "soil stability factor". If an eroding force is applied to cause the dispersion of soil in a column, and the resulting suspension allowed to flow through the soil at a constant 1/2 hydraulic head, a plot of q against 1/t can be used to calculate E. Thus s provided a constant reproducible eroding force is used to disperse a soil, E can be used to characterize its structural stability assuming K  is known.  17 2.2.2 S U R F A C E S E A L F O R M A T I O N U N D E R C O N V E C T I V E F L O W A N D CONSTANT  SEDIMENTATION.  Assuming that there is a constant sedimentation rate represented  by k  2  during surface seal formation in a laboratory column, then Eq. 1.6 may be written as  dz/dt = bq + k  (1.21).  2  In Eq.1.21, it is important to keep in mind that b = dp,.  b  Hence, c is not  constant over time i f sedimentation mechanism is considered. Also,  sedimentation  rate is never constant for a suspension consisting of different particle sizes. However, these assumptions were made in order to make it possible to show mathematically the sensitivity of q Substitution of q  s  to sedimentation. s in Eq.1.21 using Eq.1.10, and further manipulation  yields  dz/dt = {bK (H i z  H , ) / (z+LK  z  K ' ^+ kj 1  (1.22)  or  dz/dt = {bK ( H , - H ) + z 2  Eq. 1.23 can be written as  k (z+LK 2  K , " ) } / (z + L K K' ) 1 z 1 1  z  1  (1.23).  18  dt = {(z + L Kz  z  k (z+LKz K/' Sftdz 1  K " ) / [bK(H -H )+ 1 1  1  2  2  (1.24).  Or  dt = {(z+a,)/(k z+d)} dz  (1.25)  2  where  a,  = LK K, z 1  (1.26)  1  and  d  = bKOl,z  H ) +LK k z 2  2  K," 1  (1.27).  1  Integration of Eq.1.25 yields  t+C  =  k  2  "  1  [  k  2  z + d  -  d  l ^ k z z + d^ + aTkj,"  1  [  ln(k z+d)] 2  (1.28)  or  t+C = k-(kz+d) + (a,k - -dk )[ln(kz + d)] 1  2  1  2  2  2  2  (1.29)  19 where C is the constant of integration. Substituting for a , ,  and d in Eq.1.29 using Eqs. 1.26 and  respectively,and evaluating C on the condition that at t=0,  k  2 2  t  =  k z 2  -  bK ( H , z  -  z=0  1.27  gives  H )ln{(k z+d)/d} 2  (1.30)  2  Eq.1.30 is an implicit equation for z as a function of t. B y taking the second derivative of E q . 1.30 with respect to k , it can be shown using 2  L'Hospital's Rule that as k  tends to zero, E q . 1.30 reduces to E q . 1.14,  2  as  expected. The contribution to seal formation due to constant sedimentation can be assessed by numerically solving for z in Eq. 1.30 for given values of the other parameters and calculating q  using E q . 1.10. s  2.3 M A T E R I A L S A N D M E T H O D S  2.3.1 E X P E R I M E N T A L V E R I F I C A T I O N The following surface of known concentrations  OF T H E MODEL  sealing experiment was performed using suspensions  to test the validity of the theory. The apparatus used in  this experiment is illustrated in Fig. 2.2. The apparatus consists of a cylindrical column 37 cm long and 5.8 cm internal diameter made of acrylic plastic. A bent glass tube flexibly connected to a constant head system was used to maintain a constant hydraulic head difference  across the column.  A lower layer of medium sand (0.25-0.50 mm) and an upper layer of very fine sand (0.05-0.10 mm) were formed in water while vibrating the column. The layers were formed in water to ensure saturated condition. The medium sand was used to support the very fine sand. Gravel ( >2 mm) placed on a  20  F i g . 2 . 2 . A s c h e m a t i c d i a g r a m of the a p p a r a t u s used f o r t e s t i n g the model.  21 screen in the cylinder formed the base supporting the sand. The lengths of the medium sand layer and the very fine sand layer were 24.5 cm and 1.5 cm, respectively. A porous cup (PC) was installed in the very fine sand layer. Two piezometers  were also installed in the medium sand layer 12 cm apart, with the  top one being 3 cm from the bottom of the very fine sand layer. The porous cup and the piezometers  were connected to manometers mounted on a manometer  board using flexible tubing. The saturated hydraulic conductivities of the very fine sand and the medium sand were always determined  prior to initiating the sealing  experiment to make sure that the system was working satisfactorily. Under saturated condition and positive pressure, the porous cup was used as a piezometer. The piezometers  made it possible to calculate the saturated hydraulic  conductivities of the very fine sand and the medium sand using outflow volume measurement. The surface sealing experiment was performed using suspensions generated from Westham silt loam, Haney clay, kaolinite, and Abbotsford loam, respectively. The suspensions were made by adding water to 30 g of air dry soil (<2  mm)  to make a total volume of 300 c m . The soil was allowed to hydrate overnight 3  and then dispersed either with or without a dispersing agent for a period of 10 minutes in a mechanical milk shaker.  The concentration of sodium  hexametaphosphate (Calgon) used as a dispersing agent was 2.5 g/1. The concentrations  of the suspensions were determined by taking two 25 c m  3  aliquots  of the suspensions using a burette, and evaporating the suspension to dryness at a temperature of 105° C for 30 hours. To start the sealing process, the siphon connecting the column to the constant head unit was clamped, and then moved up away from the surface  water  in the column. The water table was raised close to the top of the very  fine sand layer by raising the top of the riser of the outflow unit to the same  22  elevation as the surface of the very fine sand layer. This was done to create a zero hydraulic head gradient at t = 0  as the suspension was being added. The  flexible tubing connecting the porous cup to the manometer  board was clamped.  The outflow unit was clamped and water on top of the very fme sand was removed. The suspension was thoroughly mixed again and a 130 c m  3  aliquot was  taken and gently added to the surface of the very fine sand . A plastic sheet which partially covered the very fine sand surface prevented erosion of the surface of the very fine sand as the suspension was being added. The outflow unit was kept clamped when the suspension was being added. The siphon connecting the column to the constant head system was placed back such that the bent tip of the siphon, which served as the exit for the water, was near the surface of the suspension. (Fig. 2.2). Next, the outflow unit was moved down to establish a hydraulic head gradient. A l l the clamps were removed, and the volume outflow of the filtrate was measured  cumulatively as a function of time using a balance as the  filtrate  flowed through the column with the solids remaining at the surface forming a seal. The outflow volume was measured  until the boundary between  the  suspension and the clear water above had reached the seal surface. For dispersed clay suspensions,  it took about 2 days for all the suspension to reach the  sealing surface, whereas for the loam it took about 15 minutes. In the case of the dispersed clay suspensions, the effective time used to calculate the flux densities was as long as 6 hours. The saturated hydraulic conductivity of the seal was determined by measuring the flux after the suspension had cleared. The saturated hydraulic conductivity was calculated by the formula: K  = q  (z/AH ), where K is the z s T z saturated hydraulic conductivity of the seal, q is the volumetric flux per unit  23 area per unit time, z is the thickness of the seal, and  is the hydraulic  head difference across the seal as indicated by the piezometer with a porous cup (PC). The thickness of the seal was measured with a millimeter rule. The saturated hydraulic conductivities of the very fine sand and the medium sand were 2.1X10" and 4.6X10" "m s" , respectively, as compared 5  1  with the saturated hydraulic conductivity values of the seal, which ranged between 1.3X10"  7  m s" ' for the loam to 3.7X10"'  0  m s"  1  for the dispersed  clay. Therefore, with the formation of the seal, it was appropriate to assume that the head drop due to the very fine sand and the medium sand was negligible relative to the head drop due to the seal. The flux densities of the filtrate during the sealing process were calculated as a function of time.  2.3.2 APPLICATION OF THE MODEL i. Sampling and Sample preparation.  The scheme of sampling and the location of the sampling site on the Van Halst Farm are shown in Fig. 2.3. Sampling was done on April 11, 1985 when it had not rained for 6 consecutive days. Sampling was done systematically on three transects 200 m long at a spacing of 25 m beginning from a mineral soil ridge north to an organic depression. Thus soils of different organic matter contents and, hence, of different structural and hydrologic characteristics could be identified. The horizontal distance between adjacent transects was 10 m. Bulk samples were taken with a shovel from a depth of 0-15 cm and transported to the laboratory in plastic bags, making sure that no sample lay on top of another. In total, 27 samples were taken. The samples were air-dried by spreading them thinly on the laboratory bench covered with brown paper.  VANCOUVER  CLOVERDALE  HIGHWAY 10  SCALE 1  272625  -A000  AAA  242322 A A A  0 20 40 60 m  .212019 AAA  181716 A A A  15H13 A A  «  121110 AA  A  9 8 7 A A A  654 A A «  FARM HOL'^E  321 A A A  RIDGE  T T T  ' T II  I  1  I  COAL RAILWAY  1 1  T I I IT  F i g . 2.3. Scheme o f s a m p l i n g and l o c a t i o n of s a m p l i n g s i t e on t h e Van H a l s t Farm. (The numbers r e p r e s e n t t h e s a m p l e s i n t h a t o r d e r ) .  25  The air-dry samples were passed through a nest of sieves, 2 mm (upper), and 0.5 mm (lower). The aggregates that passed through the 2 mm sieve but retained on the 0.5 mm sieve were saved for the study. The aggregates  larger  than 2 mm were further crushed with a wooden rolling pin and passed through the sieves again.  ii. Determination of Soil Stability Factor (E) for organic Soil Aggregates.  A sketch of the apparatus used for determining the soil stability factor of the soil aggregates is presented in F i g . 2.4. The permeameter consists of an acrylic plastic cylinder 30 cm long and 5 cm internal diameter. Two air pressure valves were installed in the system. The first valve was used to step down the air pressure from the main, while the second valve was used to control the air pressure entering the system. A T-joint installed in the set-up served as the air pressure distribution system. A water manometer was used to monitor the gauge pressure of the air entering the column. Flexible tubing connected to a bent glass tube served as the siphon which connected the constant head device to the water on top of the soil. The diameter of the orifice of the air entrance to the column was 4 mm. The outflow unit was connected to the base of the permeameter  through a rubber stopper with a hole bored through the centre.  The open end of the hole inside the base of the cylinder was covered with a 1.4 mm squared-hole plastic screening. The column was filled with gravel (<4 mm, >2 mm) to a depth of 3 cm from the screen by gently tapping on the side of the cylinder. The top of the gravel was uniformly covered with a layer about 0.5 cm thick of coarse sand (0.5-1.0 mm).  F i g . 2.4. A s c h e m a t i c d i a g r a m o f the a p p a r a t u s used f o r d e t e r m i n i n g s o i l s t a b i l i t y f a c t o r of the s h a l l o w o r g a n i c s o i l .  27  The aggregates were split into four parts using the sample splitter, and the column was filled with the aggregates by gently tapping the side of the cylinder 80 times with a handle of a screw-driver. In the process, a column of dry aggegates 9 cm long was formed above the sand surface. The bulk density of the aggegate column was determined by weighing the column before and after adding the aggregates.  The air-dry mass was corrected for water content to  obtain the oven-dry mass. The outflow end of the column was connected to a constant head device via flexible tubing. The outflow tubing was clamped initially. The constant head device was moved down so that the top of the riser of the constant head device was level with the bottom of the soil. The clamp blocking the flexible tubing connecting the constant head device to the base of the column was removed and the soil was wetted by capillarity. After wetting the soil by capillarity, the overflow of the constant head device was raised to the same elevation as the soil surface so that water began to collect on the surface of the soil. When the soil was fully saturated, a constant hydraulic head difference of 13 cm was established on the column. The outflow clamp was removed and water was made to flow through the soil for 30 minutes before the initial saturated hydraulic conductivity of the  aggregates  was determined by the constant head method. Having determined the initial saturated  hydraulic conductivity, the outflow  end of the column was clamped. The siphon connecting the column to the constant head device was removed. The air pressure control valve was opened gradually until the pressure registered by the water manometer was 13 cm of water. The clamp blocking the air entrance to the column was removed, and the air pressure was re-adjusted to 13 cm of water. The entrance of the air to the column was located 5 mm from the surface of the soil.  28 The movement of air within the water caused the water to move relative to the soil surface to create a reproducible eroding force causing dispersion of soil particles. A free space of 17 cm was maintained between the water  surface  and the top of the cylinder in order to minimize spilling during dispersion. Because of turbulence which was observed when air was used to cause dispersion, it was not possible to calculate the magnitude of the eroding force. The dispersion was continued for 10 minutes. The siphon connecting the column to the constant head device was re-installed and the hydraulic head difference adjusted  to 15 cm. The clamp blocking the outflow was removed and  the outflow volume was measured  cumulatively as a function of time as a seal  formed on the soil surface. Measurement was continued until the  interface  between the suspension and the clear water had reached the soil surface. the effective saturated hydraulic conductivity after sealing was measured.  Finally Every  sample was replicated three times using a "fresh" sub-sample in each case.  iii. Determination of Wet-Sieved Aggregate Stability.  The method used for the determination of the wet-sieved aggregate stability was similar to that described by Kemper (1965). The wet-sieving equipment consisted of a motor-driven mechanical device that would raise and lower the sieve holder through a distance of 2.5 cm, and at a frequency of 30 strokes per minute. The motion of the system has both an upward stroke and an oscillating action through an angle of 3 0 ° . A special sieve holder capable of receiving 12 separate sieves in each determination was ,used. The sieves had an inside diameter of 7.5 cm and 0.25  29 m m mesh openings. 4 g of air dry aggregates placed on the sieves were pre-wetted with an atomizer spray (Swartzendruber et al. 1954; Baver et al. 1972; Chaney and Swift, 1984). The samples were wet-sieved with the whole set of sieves completely immersed in a basin of water for 20 minutes. The aggregates  retained on the  sieves after wet-sieving were oven-dried at 105°C for 24 hours and weighed. Each sample was replicated three times. Because partly decomposed plant particles greater than 0.25 mm were also part of the organic soil, no correction was made for particles greater than 0.25 mm. The aggregate stability was expressed as the ratio of the oven-dry mass of the stable aggregates  after  wet-sieving to the oven dry mass of a 4g sub-sample. The temperature of the water used for the wet-sieving was 12 °C. The temperature of the water did not change appreciably during the 20-minute sieving period.  iv. Determination of Organic Matter and Mineral Matter Content.  , The organic matter content was determined by igniting 2 g of oven-dry aggregates  (0.5-2.0 mm) in the muffle furnace at a temperature of 400°C for 8  hours. The-loss-on-ignition was taken as a measure of the organic matter  content.  The major drawback of using loss-on-ignition as an estimate of organic matter of a noncalcareous soil is the error associated with loss of clay mineral water (Ball, 1964). This loss of adhered water is important in the  temperature  range of 450-600°C (Ball, 1964). Thus, provided the temperature is kept below 450°C, the loss-on-ignition method is sufficiently accurate for estimating organic matter content of a noncalcareous soil (Ball, 1964). Boggie and Robertson (1972),  30  Maas (1972), and several others have used loss-on-ignition as a measure of organic matter content of organic soils. The residue left after igniting the soil contained the mineral ash from organic matter, and sand, silt, and clay from soil management-related mixing. In the context of this work, the sand, silt, and clay fraction is referred to as the mineral matter fraction. To determine the mineral matter content, the residue left after ignition was digested by adding 20 c m  3  of 2 N HC1 to the residues in the  crucibles. The crucibles were kept overnight in a constant temperature  room at  25°C, and then placed on a sand-bath for 2 hours. The end-reaction was indicated by yellowish-green colour. The residues were filtered using Whatman No. 4 2 filter paper, washed several times with hot distilled water, and dried in an oven at 105°C. The weight of the residue after HCl-digestion was taken as an estimate of the mineral matter content. (Lowe, personal communication).The difference between the original weight of the residue and the weight after HC1 treatment represented HCl-digestible ash.  2.4 R E S U L T S A N D D I S C U S S I O N  2.4.1 V A L I D I T Y O F T H E T H E O R Y Regression analyses of log q filtrate  s  on log t were carried out for the flow of  generated from Haney clay, Abbotsford loam, Westham silt loam, and  kaolinite through a very fine sand layer, using the General Linear Model (GLM) of Statistical Analysis System (SAS), (SAS Institute Inc., 1982). The Abbotsford loam is composed of 46 percent sand, 42 percent silt, and 12 percent clay. The Westham silt loam is composed of 3 percent sand, 75 percent silt, and 22 percent clay. The dominant clay minerals of of the Westham silt loam are vermiculite, chlorite, and mica (Luttmerding, 1981). The dominant clay minerals  31 of Haney clay are vermiculite and chlorite (personal communication with Daryl Hockley, Department of Civil Engineering, U.B.C.). Table 2.1 shows the results of the regression analyses of log q  on log t s  using S A S . Except for Westham silt loam with a dispersing agent and Haney clay with a dispersing agent, for which the slopes of the regression lines were -0.72 and -0.62, respectively, all slopes were between -0.47 and -0.51. From the slopes of the regression lines, it appears the theory accounts satisfactorily for the mechanism of surface sealing with suspensions  generated from a silt loam  without a dispersing agent, a loam without a dispersing agent, Haney clay without a dispersing agent, and kaolinite with a dispersing agent, since the slopes of the regression lines are approximately -0.50 as predicted by the In order to show whether q predict the measured q  derived from the model (Eq. 1.18) can  s or not, the constant b (Eq. 1.19) has to be known. The  s constant b was defined as c/p.  b  (Eqs. 1.5 and 1.6). For a thin seal, the bulk  density is difficult to measure by the standard used in predicting q  s  theory.  methods. Hence, the value of b  from Eqs. 1.17 and 1.18 was estimated using the concept  of mass balance. Assuming that all the solids in the suspension go into forming a seal on the soil surface, then the mass balance of the surface sealing system may be written as A H c 0  = A z p , , where A is the area of the soil over which the seal b  is forming, H  0  is the head of suspension at t=0,  c is the concentration of the  suspension, and z is the thickness of the seal that has formed. From the mass balance approach we see that b = c/p  = z/H . 0  D  Table 2.2 shows the parameters used to calculate q  s  from Eqs. 1.17 and  1.18. The error associated with measuring the thickness of the seal with a millimeter rule was estimated  to be about 10 percent. For Haney clay with a  dispersing agent, the remaining suspension was siphoned off after  a 0.2 cm thick  Table 2 . 1 . Regression parameters for the r e l a t i o n s h i p between l o g q a n d l o g t . g  Dispersing  Soil  1 agent  Intercept  Slope  r  2  Sig. level  (m/s)  Westham S i L  Wo  6.9  -0.470  0.997  0.0001  Westham S i L  W  12.9  -0.72  0.995  0.0001  Haney  clay  Wo  13.1  -0.51  0.976  0.0002  Haney  clay  w  3.39  -0.62  0.953  0.004  Kaolinite  w  5.02  -0.51  0.991  0.0001  Abbots. Loam  Wo  61.8  -0.49  0.903  0.05  Wo  -  W -  without with  Table  Soi i  2.2.  parameters used to e s t i m a t e  q  .  Dispersing  Measured  Expected  Seal  Head of  agent  cone, of  cone.  thickness  susp.  susp.  susp.  (z)  at t = 0  3 (g/cm )  (cm)  of  b=c/  K  =z/H  p  o  Head  difference  at t=0 3 (g/cm )  H (cm) o  (10  -8  (m)  m/s) Westham S i L  Wo  0. 103  0. 10  0.55  Westham S i L  W  O. 104  0. 10  0.50  Haney c l a y  Wo  0. 1 10  0. 10  0. 70  Haney c l a y  O. 108  O. 10  K a o l i n t te  0. 104  0. 10  0. 40  0.059  0. 10  0.20  Abbots.  Loam  Wo  5.0 5.0 5.0 5.0 5.0 5.0  0.11  0. 36  0.049  0. 10  O. 38  2.5  0. 14  0.47  0.92  0. 48  0.039 0.049  0.08  0.49  13.0  0.04  0.45  34 seal had formed and replaced by pure water in order to measure the  saturated  hydraulic conductivity of the seal. This procedure was used because of the excessive length of time required for the entire suspension to move through the column. Therefore, the value of b was not estimated for the chemically dispersed Haney clay. Even though the same amount of soil was dispersed in each case, the concentration of the loam suspension was about one-half of the expected value (Table 2.2). This could be attributed to the coarse nature of the loam, which. resulted in significant sedimentation of the coarse particles to the bottom of the container as sampling was being done using a 25 ml burette. Therefore  the  exact textural composition of the loam used to run the experiment was unknown. It was interesting to note that the saturated  hydraulic conductivity values of the  seals formed from the Westham silt loam and the Haney clay suspensions treated with a dispersing agent were, respectively, about 20 and 64 times lower than those formed without a dispersing agent. Similar results were reported by Shainberg and Singer (1986). Tables 2.3 and 2.4 show the values of q  measured and predicted from s Eqs. 1.18 and 1.19 (assuming convective flow only) for the suspensions derived from the Haney clay and the kaolinite, respectively. In the case of the Haney clay, the predicted q  s  values by convective flow only were higher than  measured values by a factor ranging between kaolinite , however, the predicted q  s  the  1.5 and 2.0. In the case of the  values by convective flow only were lower  than the measured values by a factor ranging between 0.81 and 0.98. Because the Abbotsford loam and the Westham silt loam contain considerable proportions of silt and sand particles, it was postulated that sedimentation would make a significant contribution to the movement of the suspended particles. Therefore in addition to q  measured and predicted assuming  35  T a b l e 2.3. q v a l u e s m e a s u r e d and p r e d i c t e d ( c o n v e c t i v e f l o w ) f o r t h e Haney c l a y s u s p e n s i o n without d i s p e r s i n g agent.  Time  Measured q  Predicted q (Mass  flow only)  (10"  m/s)  ('sec.)  (10"  60  14.4  28.0  300  8.7  13.0  600  4.S  9.0  1200  3.4  6.4  1900  2.5  5.1  2900  9  4.1  ?  6  m/s)  6  T a h l e 2 4. a v a l u e s m e a s u r e d a n d p r e d i c t e d _ (convective flow) f o r the k a o l i n i t e suspension d i s p e r s i n g agent.  Measured q  Predicted q  s  (Mass flow only) (1Q-  7  m/s)  (10'  7  36.0  32.0  21.0  17.0  12.0  10.0  m/s)  37  convective flow only, q  was also predicted assuming combined convective flow s  and different constant sedimentation rates ( k ) . z was calculated for each t by 2  using a standard  routine available through U . B . C . Computing Centre. This routine  uses linear and rational interpolation and bisection to determine the zero of the left-hand-side  of Eq. 1.30 subtracted  from the right-hand-side of E q . 1.30  (Appendix 1). The "new" z calculated was used in E q . 1.10 to calculate q s which accounts for both convective flow and constant sedimentation. The q  s  values (measured  and predicted) for the Westham silt loam  suspension without dispersing agent and with dispersing agent are presented in Tables 2.5a and 2.5b, respectively. Table 2.6 gives the q  values measured  and  s  predicted by convective flow only and combined convective flow and  arbitrary  constant sedimentation rates for the Abbotsford loam suspension without dispersing agent. For the Westham silt loam without dispersing agent (Table 2.5a), the q s values predicted by convective flow only were higher than the measured by  a factor ranging between  at k  2  value of 2.5X10"  measured  q  values  1.33 and 1.56. In Table 2.5a it was observed that  values were higher than the s values at small times (250-3000 s) by a factor ranging from 2.1 to 7  m s" , 1  the predicted q  s A t large times (5000-20000 s), the predicted q values closely estimated the s measured q values by a factor ranging between 0.95 and 1.7. A t k value of s 8.9X10" m s " , the predicted q values far underestimated the measured q s s values. F i g . 2.5 illustrates the variation of q with k at different times for the s Westham silt loam suspension without dispersing agent. In F i g . 2.5, it appears a 3.6.  2  6  1  2  k  exists. This behaviour regarding k and q s s relationship for the Westham silt loam was difficult to explain since the critical 2  value which gives a maximum q  value of k  2  2  that would give the maximum value of q  was difficult to evaluate s  from E q . 1.30. In the case of the Westham silt loam with dispersing agent (Table 2.5b), the predicted q  values by convective flow only did not follow any  T a b l e 2.5a. q v a l u e s m e a s u r e d a n d p r e d i c t e d ( c o n v e c t i v e fSow o n l y a n d c o m b i n e d c o n v e c t i v e f l o w a n d c o n s t a n t s e d i m e n t a t i o n ) f o r t h e Westham s i l t loam s u s p e n s i o n w i t h o u t d i s p e r s i n g a g e n t .  Measured q  Predicted  q  k =0 2  Predicted q  Predicted  at k 2  at  of -7  2.5*10 (10  m/s)  (10  m/s)  (10  m/s m/s)  k  2  9.0*10 (10  q  predicted q  at  of  m/s m/s)  k  2  8.9*10 (10  50.0  78.0  180  110.0  15.0  22 .0  32.0  54 .0  22 .0  2.5  15.0  22.0  3 1.0  12.0  1. 2  12.0  17.0  20 . 0  7. 1  0. 74  8. 1  12.3  11.0  3.6  0.37  7.5  10.0  7. 8  2. 4  0. 33  6.3  8 . 8  6.0  1. 8  of  m/ m/s)  39  T a b l e 2.5b. q v a l u e s measured and p r e d i c t e d (convective ow o n l y ) f o r t h e Westham s i l t loam suspension with d i s p e r s i n g agent.  Time  Measured q  Predicted q  s  (Mass flow onlyi  (10"  7  m/s)  ao-  50  76.0  43.0  150  34.0  25.0  400  19.0  15.0  800  10.0  11.0  1200  6.8  s.s  1700  5.7  7.4  2400  4.9  6.2  4000  3.5  4.8  4600  2.8  4.5  22000  1.3  2.1  31000  1.1  1.7  T a b l e 2.6. q v a l u e s measured and p r e d i c t e d ( c o n v e c t i v e f l o w o n l y and c o m b i n e d c o n v e c t i v e f l o w and c o n s t a n t s e d i m e n t a t i o n ) f o r t h e A b b o t s f o r d loam suspension without d i s p e r s i n g agent.  T 1me  Measured q  Predicted  (at  q  k =0) 2  Predicted q  Predicted  at k  at  2  of  at k  k of 2  m/s  8.9*10  m/s  ( 10  m/s)  (10  2  of  m/s)  (10  3.6*10  m/s  -5  -5 m/s)  Predicted q  -5  -7 2.5*10  q  m/s)  (10  (s)  (10  40  12 O  13.5  13 O  8.2  3.5  1 10  5.2  8.2  7 .9  3.9  1 .4  200  3.8  6 . 1  5.8  2.4  0.78  700  2.9  3.2  3.0  0 . 82  0. 23  m/s)  O  41  180  160 -  Legend  140 -  250 s 1 500 s  120-:  3000 s  7?\ 100  80  y  so  40  J  »  20-  20  40  60  80  SEDIMENTATION RATE (10~ m / s ) 7  F i g . 2.5. V a r i a t i o n o f f l u x d e n s i t y , (q ) w i t h constant sedimentation rate ( k ) at d i f f e r e n t t i m e s f o r t h e Westham s i l t loam s u s p e n s i o n without d i s p e r s i n g agent. 2  100  42  consistent pattern with the measured values. This inconsistency could be due to the influence of the chemical used to cause the dispersion. With chemical dispersion, the measured q  dropped steeply with time as indicated by the  s magnitude of the slope of log q  against log t (Table 2.1). s In the case of the Abbotsford loam (Table 2.6), the q  values predicted s by convective flow only were higher than the measured values by a factor ranging between k  2  1.10 and 1.60. As was expected, the q  value of 2.5X10"  7  m s"  were almost equal to those predicted by  1  convective flow only (that is k  values predicted at low s  2  =  0 ). A t k  2  of 8.9X10"  6  m s" , 1  the  predicted values estimated closely the measured values for three out of four cases. Fig. 2.6 illustrates the variation of q  with k for the Abbotsford loam s suspension at different times. Unlike the Westham silt loam suspension, q s versus k  2  2  relationship for the Abbotsford loam suspension decreased consistently  with increasing k . 2  In order to show the behaviour of the model, the log-log plots of q s against t were also obtained. Figs. 2.7, 2.8, and 2.9 illustrate the log-log plots of q against t for both the measured and the predicted cases (convective flow s only) for the suspensions derived from the Westham silt loam, the Haney clay, and the kaolinite, respectively. The linear log-log relationships between q  s  and t  for the measured cases were almost parallel to those predicted by mass flow only for the suspensions derived from the Westham silt loam without dispersing agent, the Haney clay without dispersing agent and the kaolinite with dispersing agent. In contrast, the plot of log q against log t for the measured case of the s Westham silt loam suspension with a dispersing agent was too steep and the parallel arrangement obtained (Fig. 2.7).  between the measured and the predicted case was not  14-i  Legend  12-  40  s  11 0 s 10 I  o  200  s  700  s  >-  <5  \  LiJ Q  X ID  \ \  Q UJ  I— O Q  Cr: Q.  \  4  V\  100  200  300  SEDIMENTATION RATE ( 1 0 ~ m 7  /s)  F i g . 2.6. V a r i a t i o n of f l u x d e n s i t y (q with constant sedimentation rate (k ) at d i f f e r e n t t i m e s f o r t h e A b b o t s f o r d loam s u s p e n s i o n without d i s p e r s i n g agent (Wo) 2  400  1000  M e a s u r e d (Wo)  A  •  Predicted (Wo) •  \ •  M e a s u r e d (W)  \  Predicted (W)  • 10 H  10  i  i  i r I i i 11  100  i  1—i—i M M  1000  1  1—i—i i i i 11  10000  Time (s) F i g . 2.7. The l o g - l o g r e l a t i o n s h i p between f l u x d e n s i t y m e a s u r e d and p r e d i c t e d by c o n v e c t i v e f l o w ( E q . 1.18) and t i m e f o r t h e Westham s i l t loam s u s p e n s i o n w i t h (W) and w i t h o u t (Wo) d i s p e r s i n g a g e n t .  1—  X 1000-  X  00 I  X  o  X  A  >'Tn  c  CD  Q  x  100-  A  3  Legend A  Measured (W)  X  Measured (Wo)  A  A  Predicted (Wo)  1010  -1  1  1 — I — l l l l  100  -\  1—i—rr-  1000  Time (s)  F i g . 2.8. The l o g - l o g r e l a t i o n s h i p between f l u x d e n s i t y (measured a n d p r e d i c t e d by c o n v e c t i v e f l o w ) and t i m e f o r t h e Haney c l a y suspension.  A  V  A  in  O  V  A  100%  X  A  *V A  >~  c <u Q X  Legend A  Measured (W) Predicted (W)  10100  -I  1  1  1  1  1  ! I  - 1 — i — i — i — i  1000  T i m e (s)  F i g . 2.9. The l o g - l o g r e l a t i o n s h i p between f l u x d e n s i t y (measured and p r e d i c t e d by c o n v e c t i v e f l o w ) and t i m e f o r t h e k a o l i n i t e suspension.  47 The log-log plot of q  s  against t (for q  s  measured, predicted by convective  flow only, and predicted by combined convective flow and different  constant  sedimentation rates) with respect to the Abbotsford loam suspension without dispersing agent is illustrated in Fig. 2.10. It was observed that if the last data point of the measured case was discarded, a linear log-log relationship between q  s  and t could be obtained. F i g . 2.11 shows the log-log plot of q  measured q  s  and q  s  and t for  s  predicted by combined convective flow and different  k  2  values for the Westham silt loam suspension without dispersing agent. With combined convective flow and constant sedimentation rate ( k ) , there were some 2  deviations from the expected log-log linearity between q arrangement  expected between the predicted q  and t and the parallel  and the measured q s  was not s  obtained. A t this juncture, it is appropriate to distinquish between a "colloidal solution" or "sol" and a suspension. According to V a n Olphen (1963), the particles of a suspension are relatively larger and are subject to sedimentation. The particles of a sol, however, are comparatively smaller and are subject to Brownian motion. The border line between a sol and a suspension is usually set at an "equivalent spherical diameter" or "Stokes diameter" of 2 um (Van Olphen, 1963). Table 2.7 shows the "equivalent spherical diameter (d)" and the corresponding Stokes velocity and the time required for the particles to fall 5 cm in a column. Table 2.7 was derived on the assumption that k to Stokes velocity. In Table 2.7, it is evident that k  2  2  was equivalent  of 2 . 5 X 1 0 '  7  m/s  corresponds to Stokes diameter in the colloidal size range, and therefore,  the  particles are not subject to sedimentation (Van Olphen, 1963). The Stokes diameter corresponding to k  2  of 8 . 9 X 1 0 "  6  is 3.16Mm (Table 2.7), which is in  silt size range. Particles in silt size range are subject to sedimentation (Van  48  Legend A  M e a s u r e d (Wo) P r e d i c t e d (Wo) k2=0 k2=3._6»JCT m / s 5  k2=8.9 « 1 0 " m / s  )_}  10  1  ,  1  1—,—i  i  i |  1  r—-i  1  i  100  T i m e (s)  F i g . 2.10. The l o g - l o g r e l a t i o n s h i p between f l u x d e n s i t y measured and p r e d i c t e d u s i n g Eq. 1.18 ( c o n v e c t i v e f l o w ) a n d E q s . 1.30 w i t h 1.10 (combined c o n v e c t i v e flow and c o n s t a n t s e d i m e n t a t i o n r a t e ) and time f o r the A b b o t s f o r d loam s u s p e n s i o n w i t h o u t d i s p e r s i n g agent.  I I I I  1 0 0 0  49  1000-  A  .A 100-  A A A  E OD  I  o  >~  "in C <D  10-  Q X  \ Legend  A  1-  0.1 100  -1  Measured  (Wo)  k  2  of 8 . 9 * 1 0 "  k  2  of 9 . 0 * 1 0 "  k  2  of 2 . 5 * 1 0 -  6  m/s  7  m/s  7  m/s  1—!  1000  10000  Time (s)  F i g . 2.11. The l o g - l o g r e l a t i o n s h i p between f l u x d e n s i t y m e a s u r e d and p r e d i c t e d u s i n g E q s . 1.30 w i t h 1.10 (combined c o n v e c t i v e f l o w and c o n s t a n t s e d i m e n t a t i o n r a t e ) and t i m e f o r t h e Westham s i l t loam s u s p e n s i o n w i t h o u t d i s p e r s i n g agent.  50  T a b l e 2 . 7 . E q u i v a l e n t s p h e r i c a l d i a m e t e r and c o r r e s p o n d i n g S t o k e s v e l o c i t y and t i m e r e q u i r e d f o r i n d i v i d u a l p a r t i c l e s t o f a l l 5 cm i n a s u s p e n s i o n .  Time  Stokes velocity  required  (10-  to f a l l  7  m/s)  5 cm  2.5  55.5  9.0  15.4  36  3.9  90.0 360 .0 1800  K  5  4  0.38 ° -  0  7  8  51  Olphen, 1963). Considering the case of the Abbotsford loam (Table 2.6), a comparison of predicted q q  s  values at k  2  of 8 . 9 X 1 0 "  6  m s"  1  with the corresponding measured  values using the Mann-Whitney U test (Siegel, 1956) showed no significant  difference between the two sets of q  s  values, (the calculated U value =  lower tail U value at 1 percent level of significance = 0). From  5; the  the  Mann-Whitney U test analysis, it can be deduced that sedimentation of particles in the silt size range contributes to an increase in the seal thickness with respect to the Abbotsford loam suspension, even though the last predicted data point deviated considerably from the measured value (Fiq. 2.10). In reality, particles in the sand size range are also subject to sedimentation, but large k values are required for sand. A t large k underestimates  2  values, however, the model seriously  q . Therefore, the use of a constant k s  would be appropriate to consider k  2  2  2  is not appropriate.  It  as time-variant.  Paired comparisons of the measured and predicted q  s  values (convective  flow only) were sought using the non-parametric Mann-Whitne3 U-test (Siegel, r  1956). The calculated U values and the lower tail and the upper tail U values at 1 percent level of significance for all the suspensions are given in Table 2.8. Since the calculated U values were generally greater than the lower tail values and less than the upper tail values, it is statistically accepted that there are no significant differences between the measured and the predicted values. With reference to the Abbotsford loam suspension, the analysis showed that sedimentation of particles in the silt size range did influence q^. However, the actual textural composition of the "loam" suspension that was used to run the test was not known, since there was settling of particles to the bottom of the container during sampling. In contrast, the model failed to predict the influence of sedimentation for the Westham silt loam, despite the high silt  52  T a b l e 2.8. N o n - P a r a m e t r i c Mann-Whitney U - t e s t f o r comparing measured and p r e d i c t e d ( c o n v e c t i v e flow only) q values.  Sample  Disp.  Calc.  Lower  Upper  agent  U  Tail  Tail  U  U value  value  value  (1% sig.)  (1% sip.)  Remarks  Westham S i L  Wo  16  7  42  n.s.  Westham S i L  W  54  26  95  n.s.  Haney clay  Wo  9  4  32  n.s.  Kaolinite  W  15  4  32  n.s.  Abbots. Loam  Wo  5  0  16  n.s.  53  content of the silt loam. Relative to E q . 1.21, it is important to realize that for a silt loam soil during a sealing event, particles reach the surface of the seal by convective flow and variable sedimentation. Thus, both c and k  2  are variable because over the  silt particle diameter range the sedimentation velocity is variable (Table 2.7). In arriving at E q . 1.30, it was assumed that k  2  and c were constant. These  limitations inherent in the assumptions used to derive E q . 1.30 may be responsible for the poor sensitivity of the model when sedimentation is taken into account. The deviation from the linear log-log relationship between measured q  and s  t for the "loam" suspension stems from the last data point (Fig. 2.10). Because the measured concentration of the loam suspension was low, and the seal formed from the loam particles was relatively more permeable than the other seals (Table 2.2), it took less than  15 minutes for the dispersed solids of the "loam"  to reach the surface of the seal. There was very little change in flux density of the loam suspension towards the end of the sealing process. 1/2 It may be recalled that for the q versus 1/t relationship to be valid, s the factor 2 b K ( H , - H ) t / L K must be much greater than 1 (Eq. 1.16). z 2  1  2  2  The calculated values of this factor for the different suspensions at time t (in seconds) are presented in Table 2.9. It can be observed that even for the minimum time of 40 seconds used for the Abbotsford loam, the factor is much greater than  1. Therefore, it is reasonable to accept that during surface seal  formation of soil in a laboratory column under convective flow, the flux density of the filtrate is proportional to inverse square-root of time. Testing the model experimentally revealed a good agreement  between  theory and experiment. However, some discrepancies were observed due to some limitations inherent in the theory. Two of the assumptions made in formulating  Table  Sample  2.9.  The  factor  Dispersing  2bK, (H,-H )t/L K 2  2  2  2bK ,(H,  -  t  2  H )t/L K 2  2  agent  Westham  SiL  Wo  (16.9 s" )  Westham  SiL  W  (304.0 s ' ' )  Wo  (10.3 s  W  (313.6 s ' )  Wo  (0.54 s" )  Haney  clay  Kaolinite Abbots.  Loam  1  - 1  t  ) t t  1  1  t  .  z  55 the theory were that (i) the sediment reaches the sealing surface by convective flow and constant settling and that (ii) al) the solids are deposited on the soil surface and none pass through to clog the pores in the soil body. The latter assumption may not be quite correct because in the case of the kaolinite, it was noticed that the "milky material" did pass through the very fine sand layer. In this instance, the mass balance approach used to estimate b was incorrect. This may be the reason why it was only in the case of the kaolinite suspension that the predicted values underestimated the measured values. In the case of the other suspensions, the suspended particles did not move into the sand layer. Personal communication with R. J . Southard by Shainberg and Singer (1986) showed similar findings. Sedimentation can influence the rate of increase of seal thickness for suspensions derived from coarse or medium textured soils. However,introducing a constant sedimentation parameter  into the convective flow  model did not improve the model. Therefore, it is necessary to consider sedimentation rate to be  time-dependent.  Another factor observed in the study was that when a suspension was treated with a dispersing agent, the saturated  hydraulic conductivity of the  resulting seal was about 20 or 64 times lower than using water alone. The geometrical arrangement  of the constituent mineral particles, including voids,  during the process of surface seal formation can play a significant role in the nature of the seal that develops. Considering clay particles alone, the arrangement  final  of the clay particles forming the surface seal may be described as  either dispersed or flocculated (Yong and Warkentin, 1975). In a flocculated condition, the clay particles tend to settle with random orientation and open structure  (Yong and Warkentin, 1975; personal communication with R. J .  Southard by Shainberg and Singer, 1986). In a dispersed state, due to interparticle repulsion, the clay particles tend to settle with parallel orientation  56 into a relatively compact seal (Yong and Warkentin, 1975; personal communication with R. J . Southani by Shainberg and Singer, 1986). The suspension of Haney clay without a dispersing agent was observed to have formed floes. Similar floes were not observed for the clay treated with a dispersing agent. The deviation in log q  versus log t plot whenever  s suspension is treated with a dispersing agent suggests that K time-dependent  z  the  is likely to be  whenever the suspension is treated with a dispersing agent.  The relationship between the soil stability factor (E) (Eqs. 1.18 and  1.19)  and aggregate stability of a shallow organic soil in transition towards becoming a mineral soil will be discussed later in this chapter.  2.4.2 R E L A T I N G T H E SOIL S T A B I L I T Y F A C T O R TO A G G R E G A T E STABILITY,AND MINERAL MATTER CONTENT OF A SHALLOW ORGANIC  SOIL.  i. Background Information of the  Soil.  Field water content of the soil at the time of sampling varied from 32 percent by mass on the mineral soil ridge to 118 percent by mass in the organic depression (Appendix 2). The water content of the air dry samples varied from 4 percent by mass for the mineral soil on the ridge to 20 percent by mass for the organic soil in the depressions. The bulk densities of the aggregate columns are presented in Appendix 4. The organic layer depth profiles of the three transects are illustrated in Fig. 2.12. Some variations in the depth of the organic layer were observed for samples taken from the same horizontal grids, indicating nonuniformity of the organic depth profile. F i g . 2.13 and Table 2.10 illustrate the mean aggregate stability, the mean organic matter content, and the  57  0.5-  A  0.4  on UJ  _l o  x  ts- o  0  X—V-X  V\ / /  0.3H  z <  o  oc o  LL.  0.2  o X  I—  Q_ UJ  o  o.H  0.0  Legend 200  A  TRANSECT 1  DISTANCE FROM THE MINERAL RIDGE (m)  X  TRANSECT 2  •  TRANSECT 3  0  25  F i g . 2.12. O r g a n i c three t r a n s e c t s .  50  75  layer  100  125  150  depth along  175  the  58 mean mineral matter content for the samples as a function of distance from  the  mineral soil rid^e to the organic depression. Aggregate stability and organic matter content increased with increasing distance from the ridge. A s expected, the mineral matter content decreased with distance from the ridge since organic matter content and mineral matter content are almost complementary to one another (Fig.2.13).  ii. The Soil Stabiliy Factor (E) of the Shallow Organic Soil Aggregates.  Figs.2.14 and 2.15 show the regression of q  s  on l/j/t for two samples  taken at a distance of 25 m and 175 m, respectively, from the mineral soil ridge. These two samples represent soils of low organic matter content and high organic matter content, respectively. A s predicted by the theory, the relationship between q  s  and 1/i/t for a flow of suspension through a column was  approximately linear with zero intercept. Extrapolation of the regression lines showed a slight deviation from the zero intercept. The mean coefficient of determination ( r ) values and the standard deviations of the r 2  2  values for q s  against ll\/1 for all the samples are illustrated in Figs.2.16a and 2.16b, respectively. The mean r  2  values ranged from 0.88 to 0.99, and the  deviations of the mean r  2  values ranged from 0.01 to 0.06. The distribution of  the probability levels at which the r samples is presented  2  standard  values are significant for all the replicated  in F i g . 2.17. The probability levels at which the r  2  values  were significant ranged from 0.0001 to 0.008. The E values obtained through regression analyses were designated  as  "actual soil stability factor (E )". The mean E values and their standard a a deviations are given in Appendix 3. The coefficients of variation for the mean E  values are plotted in Fig.2.18. Apart from a single case where the coefficient  59  F i g . 2.13. m a t t e r and transects.  Mean a g g r e g a t e m i n e r a l matter  s t a b i l i t y , organic c o n t e n t a l o n g the  60  T a b l e 2.10. Mean a g g r e g a t e and m i n e r a l m a t t e r c o n t e n t  s t a b i l i t y , organic of the t r a n s e c t s .  Distance  Mean  Mean  Mean  from the  aggregate  organic  mineral  mineral  stability  matter  matter  content  content  soil ridge (m)  (%)  (%)  (%)  0  67.211.6  14.111.1  77.911.7  25  66.713.8  18.2+2.4  73.912.6  50  73.519.5  24.213.1  67.213.6  75  80.513.3  36.616.5  54.0116.7  100  81.911.2  39.310.8  50.511.1  125  84.613.3  44.716.9  45.017.5  150  91.414.2  60.119.2  28.719.3  175  96.410.5  67.215.1  22.615.2  200  93.311.2  60.010.9  29.010.2  matter  61  F i g . 2.14. V a r i a t i o n of f l u x d e n s i t y w i t h i n v e r s e s q u a r e - r o o t of t i m e f o r sample no. w h i c h was 25 m from t h e m i n e r a l r i d g e .  4  62  F i g . 2.15. V a r i a t i o n of f l u x d e n s i t y w i t h i n v e r s e s q u a r e - r o o t of t i m e f o r s a m p l e no. w h i c h was 175 m f r o m t h e m i n e r a l r i d g e .  8  0.87 0.86-  0.85 -I—;—i—i—i—i—:—i—i—i—i—i—i—i—i—i—i—i—l—i—l—l—i—i—i i i 1 2 3 4 5 6 7 6 9 10 11 12 13 14 15 16 17 18 1920 21222324252627  SAMPLE NUMBER  deviation; 7T ^ d e v i a t i o n s (a) a s s o c i a t e d relationships. V  1  U  6  S  (  b  )  a  n  d  with g s M  standard v e r s u s 1 A/t '//t v  e  r  b  U  S  38  36H 34H 32 H  30H 28H O  A  I  26  o  20-  O  (ti  n S3 P  24H  18161412108 6420  0 01 I r p i { II  I  la  '",""T  e © L E V E L OF PROBABILITY  Fia 2 17. P r o b a b i l i t y d i s t r i b u t i o n for r ' l s s i g n i f i c a n t f o r the r e l a t i o n s h i p versus 1 /v 1.  which  9  7  %  65  F i g . 2.18. model.  Variability  within  E  a  of  the  66  of variation was 0.70, the coefficients of variation were always below 0.50, with the majority ranging between 0.06 and 0.30. F i g . 2.19 illustrates the diagram for the variation of E  scatter  with distance from the mineral ridge. The E  a a values are low between the ridge and 125 m away from the ridge. Visually, it appears the curve relating E  a  with distance from the ridge may be described as  exponential. To ascertain whether or not the E values obtained through regression analysis were real, the E values were again obtained by constraining the curves of q  s  against 1/j/t  to have zero intercepts as predicted by the model. The E  values obtained in this way were designated as "expected soil stability factor (E )". The E values and their standard deviations are given in Appendix 3. x x The distribution of the coefficients of variation in E  x  was similar to that of E . a  With the exception of one case where the coefficient of variation was than 0.60, the remaining coefficients of variation in E  greater  were always below 0.50,  with the majority ranging between 0.01 and 0.33. The regression of E  a  on E  x  is presented in Fig.2.20.The regression  equation relating E  to E was obtained as E = -15.44 + 1.22 E (r = 0.905). a x a x The linear relationship indicated that 82 percent variation in E could be a ascribed to E ( r =0.819). From the regression of E on E , it was observed x a x 2  that the error in obtaining a unity slope was 22 percent. Inspite of this error, the regression of E  a  on E  x  was highly significant (P=0.0001). Therefore, the E  values obtained through linear regression analysis were good enough to  represent  the model.  ii. Relating the Soil Stability Factor (E) to Soil Structural Stability.  67  300-  + 250-  O 33  O  200H  or O i— CJ <  150H  o m  in  +  100-  A  O  in  +  o  O  +  +  4  504  ~r  0  25  50  — 75 i— 75  100  —i  125  Legend 1  1  1  150  175  200  DISTANCE FROM THE MINERAL RIDGE (m)  F i g . 2.19. The s c a t t e r d i a g r a m showing t h e v a r i a t i o n of E w i t h d i s t a n c e from t h e mineral r i d g e . a  A  Transect 1  +  Transect 2  O  Transect 3  350-, L a  =-15.44 + 1.22  E X  r=0.905 •a  300-  S.E.  = 0.1152  E  OC o  250-  I—  o  L<  >-  tr  _j CO i<  200-  00  O  150 -  A  <  100-  A A/k^A 'A  A  50-  50  100  150  200  250  EXPECTED SOIL STABILITY FACTOR [/(  300 m  s)  F i g . 2.20. A c t u a l s o i l s t a b i l i t y f a c t o r (E ) v e r s u s e x p e c t e d s o i l s t a b i l i t y f a c t o r (E )? (The l i n e i s t h e r e g r e s s i o n l i n e w i t h t h i given equation).  /da  69  The data for aggregate stability determined by the wet-sieving method presented in Appendix 4. The regression  on aggregate stability a (AGG) was obtained using the G L M of S A S . The linear correlation between E  analysis of E  and A G G was positive and significant at a probability level of 0.0001 ( r 0.57). When E substantially E  a  a  was log-transformed  improved ( r  =  2  the correlation between E  0.66). The substantial  due to A G G in the log-transformed  between E  a  and A G G was  a  increase of the variation in  case suggested that the  against A G G is presented in F i g . 2.21. The equation relating E AGC  a  =  2  relationship  and A G G could be exponential. The semi-logarithmic plot of E  expressed as E =(4.41)(1.04) a  are  a  a  to A G G was  (r = 0.812; P = 0.0001).  The data for organic matter content as determined by loss-on-ignition, and mineral matter content as estimated by HC1 digestion, are presented in Appendix 4. The linear correlation coefficient between E  and mineral matter content was a negative and significant at a probability level of 0.0001 ( r = 0.63). On 2  log-transforming E , the correlation between log E and mineral matter content a a (MN) was improved ( r  2  =  0.70). Therefore  the relationship between E  was described as exponential. The correlation between loq E  a negative and highly significant (P = 0.0001). Fig.2.22 gives the  E  a  semi-logarithmic  and M N . The equation describing the relationship between  — MN and M N was obtained as E =(309.03)(1.02) " . a a Since mineral matter content and organic matter content are  complementary variation in E  and M N  and M N was  5  relationship between E  a  almost  to one another (Fig.2.13), organic matter explains the same a  as mineral matter content. The positive and highly significant  correlation between E  a  and aggregate stability suggests that E  of soil structure. It is important  to note that  a  is an attribute  and the concentration of  dispersed solids (c) also indicate the consequence of structural instability. The lower the structural  stability the smaller are the particles that are eroded  and  70  F i g . 2.21. S o i l s t a b i l i t y f a c t o r v e r s u s wet-sieved aggregate s t a b i l i t y . ( T h e l i n e i s the r e g r e s s i o n l i n e with the g i v e n e q u a t i o n ) .  F i g . 2.22. S o i l s t a b i l i t y factor versus m i n e r a l m a t t e r c o n t e n t . (The l i n e i s t h e r e g r e s s i o n l i n e w i t h the g i v e n e q u a t i o n ) .  72 the more impervious the seal that forms by the eroded particles. However, the reason for using E instead of K  and c as an index of structural stability is z that E is easily determined without having to know K and c explicitly. The highly significant negative correlation between mineral matter content and E  a  suggests that the mineral matter fraction is likely to be easily eroded  whenever the soil is subjected to an eroding force. This implies that the structural stability of the shallow organic soil is likely to be low at higher mineral matter content. The reverse will be true with higher organic matter content. In F i g . 2.13 it appears the arbitrary critical level of mineral matter content of the shallow organic soil could be around 45 percent. This arbitrary critical level occurs 125 m from the ridge. The E  values are also lowest a  between the ridge and 125 m away from the ridge (Fig. 2.19). Therefore, between the mineral ridge and 125 m away, the structure of the soil is likely to be weak. Data not reported in the thesis indicated that the thickness of the seal measured with a millimeter rule varied from approximately 1.5 mm for samples taken from 150 m away from the ridge to 4 mm for samples taken directly on the ridge. Beyond 150 m away from the ridge the seals were so thin that they could not be measured using the millimeter rule. The thickness of the seal depends directly on the concentration of the suspension. A decrease in water infiltration rate with an increase in seal thickness was reported by Shainberg and Singer (1986). 2.4.3 V A L I D I T Y O F T H E P R O C E D U R E . The validity of the procedure was assessed by evaluating E  a  derived from  the model against aggregate stability as determined by the standard wet-sieving method (Kemper, 1965). The highly significant correlation between E  and  73 aggregate stability suggests that E  a  is an attribute of soil structure. E  a  that the 'awer the saturated hydraulic conductivity of the seal and the  shows higher  the concentration of dispersed solids the less stable is the soil. A n important aspect about the E-test worthy of note is that the whole procedure for the determination approximately suspension  of E using the air dispersion technique may take  1 to 3 hours, depending on the concentration of the resulting  and the permeability of the seal. The E-test is, therefore, more  efficient than the wet-sieving test which may require approximately 2 days. In absence of planar voids and biopores, the key soil physical property that controls water movement in soils is pore size distribution (Marshall, 1958). The pore size distribution is also a function of soil structure (Childs, 1940; 1942). A strong dependence of hydraulic conductivity on soil structure was experimentally reported by Swartzendruber  et al. (1954) and Sharma and Uehara  (1968). The ease of dispersion of soils has, in isolation, been used as an index of soil erodibility (Middleton, 1930). The so-called dispersion ratio was defined by Middleton (1930) as the ratio of silt+clay when the soil is shaken without a dispersing agent to silt+clay in a sample that was previously dispersed with a dispersing agent. Conceptually, the dispersion ratio should be smaller for a more stable soil. Woodburn and Kozachyn (1956) observed negative correlations  between  dispersion ratio and aggregate stability and soil splashability. In reality, dispersibility, per se, cannot be used as an absolute measure of soil erodibility, since under natural rainfall condition in the field detachment of undispersed aggregates can occur (Bryan, 1969; Alberts et al. 1980).  Nevertheless,  dispersibility can be related to structural stability of soils provided the size of the dispersed solids is specified. Therefore, an index that involves both saturated hydraulic conductivity of a seal and soil dispersibility must relate to soil  74  structure. The procedure has revealed that structural stability of the shallow organic soil is likely to decrease with increasing mineral matter content. The tendency for mineral matter to cause structural instability in the shallow organic soil will be discussed in detail in chapter 2 of this thesis.  2.4.4 V A R I A B I L I T Y W I T H I N R E P L I C A T E S O F T H E SOIL S T A B I L I T Y  FACTOR  F O R T H E S H A L L O W O R G A N I C SOIL. Even though the agreement between theory and experiment was good, variability within sample populations of replicates of the soil stability factor derived from the model were high for some of the samples. The mode of dispersion of the aggregates and the extent of sample homogeneitj>  are  important  in this regard. The effect of the time of dispersion on aggregate stability determination by the wet-sieving method was first studied by Russel and Feng (1947). They observed that the order of aggregate stability of soils could change depending on the length of time used for dispersion. They realized that the mass of stable aggregates retained on the 0.25 mm sieve as a function of time of dispersion was described by the relation log W = a - b log t, where W is the mass of water stable aggregates, t is the time of dispersion, and a and b are parameters.  Russel and Feng (1947) described the two parameters a and b as  the "initial stability" and the "rate of disintegration", respectively. Dong et al. (1983) measured dispersibility ratio as the ratio of clay-sized particles dislodged by hand shaking in water to the total clay-sized fraction determined by ultrasonic dispersion. They observed that the correlation between the dispersibility ratio and organic carbon became less significant for longer periods of dispersion (32 to 128 hours).  75  In reality, the order of aggregate stability could similarly be reversed by also using a larger force of dispersion, even though the time of dispersion could be short. In this study, air was used to induce dispersion, making the mode of dispersion too vigorous. As a matter of fact, the validity of using air as a natural process of causing dispersion is debatable. The method of dispersion and the fact that the mineral matter brought up during tillage and harvesting operations has not been completely homogenized with the organic soil could explain some of the variability associated with some of the samples.  2.5  CONCLUSIONS The theory described contributes to our understanding of the mechanism of  surface sealing of soil, even though the model does not describe the actual field situation. The soil stability factor (E) derived from the model is shown to be related to structural stability, even though E , per se, is not an absolute measure of structural stability of soil. Since the saturated  hydraulic conductivity of the  seal and the concentration of dispersed solids are the key factors controlling E , the practicality of this procedure lies in its use for assessing the sensitivity of soil to dispersion and sealing. On dispersion, the finer the eroded particles the more impervious the seal that forms from the eroded particles. In the management  context, the least stable soil is likely to form the most impermeable  seal or crust when subjected to rainfall impact. It can therefore be argued that E can be used as an indicator of structural stability of soil. In terms of management  of these shallow organic soils, the procedure  has  amply shown that mixing the fine-textured mineral subsoil with the organic layer is not an appropriate management  strategy  for these shallow organic soils. Apart  from the fact that mixing mineral soil with organic soil hastens biological oxidation of the organic materials,and subsequently, tending to increase  subsidence  76  of the organic surface (Broadbent, I960), it has been shown through the procedure that dispersibility and surface sealing are apt to increase as the mineral matter content of the organic soil increases. Surface sealing contributes in part to the frequent ponding of water on the shallow organic soil during rainfall events. There is a need for changes in land use strategy of the shallow organic soil as the depths of the organic layers become shallower.  3. H Y D R O - P H Y S I C A L I M P O R T A N C E O F O R G A N I C M A T T E R D E P L E T I O N O F A S H A L L O W O R G A N I C SOIL  3.1 I N T R O D U C T I O N The role of organic matter as a stabilizing agent of mineral soil aggregates has long been recognized (Harris et al. 1966). In an unstable soil, the  destructive  forces related to air entrapment can shatter the aggregates during a wetting process to result in slaking (Emerson, 1977). The slaked particles may reduce the effective porosity of the soil and consequently diminish the ability of the soil to conduct water. The deleterious effect of weak soil structure is a decline in soil productivity. De Boodt et al. (1961) observed a significant correlation between aggregate stability and crop yield. The literature is full of information on the structural state of mineral soils. In contrast, the structural state of organic soils has not been studied extensively. Yefremova (1984) attempted  to estimate the degree of structure  formation in organic soils from their morphological characteristics. He reported that the degree of structure  formation in organic soils depended directly on the  total amount of cold 0.1N N a O H extractable humic substances. Using cold 0.1N N a O H extractable fraction, Yefremova (1984) observed that the quality of structure of forest peat soils decreased with depth. He termed the cold 0.1N N a O H extractable fraction as "organo-mineral complexes". The standard method of reclaiming organic soils for agricultural use is drainage. Long-term drainage increases the rate of oxidation of the organic materials because the aeration status of the organic soil is improved. Decomposition and mineralization occur concurrently when organic soil is reclaimed for agricultural purposes. (Mineralization may be defined as the biological oxidation of the organic matter to water and carbon dioxide with liberation of  77  78 the mineral nutrients). Mineralization of organic matter in organic soils results in substantial changes in both the physical, chemical, and hydrologic characteristics of the organic soil. A comprehensive review on the effect of mineralization in organic soils on the physical and hydrologic properties was presented by Bonsu (1984). Total pore volume and pore sizes appear to be the key properties  worthy  of note as far as the effect of mineralization on the ability of organic soils to conduct water is concerned. Mineralization tends to decrease the total pore volume and pore sizes in organic soils (Lishtvan and Zuyev, 1982), resulting in a corresponding decline in the ability of the soil to conduct water (Bonsu, 1984). Another important effect of mineralization in organic soils is subsidence (a decrease in surface elevation). A comprehensive review of subsidence in organic soils was presented by Bonsu (1984). A n important aspect to note about subsidence is that continued subsidence after  reclaiming an organic soil is a  process of biochemical oxidation. Therefore, under favourable conditions as  pertains  in arable lands, biochemical oxidation of the organic material may progress the underlying mineral material is reached, if no control measures are (Schothorst,  until  taken  1977).  Tillage hastens the rate of mineralization of the organic materials, because tillage opens up the pores to initially improve the aeration of the soil. In addition, tillage brings up new organic material from below to face biological oxidation. In reality, organic substrate is important for continued microbiological activity in soils. Beck (1984) obtained a highly significant correlation (r=0.93) between microbiological activity and organic matter content in agricultural soils. In mineral soils, microbiological activity associated with organic matter decomposition tends to produce stable soil aggregates (Meredith and Kohnke, 1965; Harris et al. 1966). Humus that is produced through microbiological activity in soil is bound to clay to form stable aggregates. Clay-bound humus has been observed to  79 be more resistant to further microbiological attack than humus alone (Sen, 1961). However, when clay is present in excess compared with humic materials, logically, the surplus unbound clay will have the tendency to be less stable than the bound clay. Hydraulic conductivity of soil is important in the application of drainage and irrigation systems. Hydraulic conductivity depends in part on soil structure. Swartzendruber et al. (1954), Taylor and Henderson (1959), and Sharma and Uehara (1968), used saturated hydraulic conductivity as a measure of soil structural stability. Swartzendruber et al. (1954) observed that for disturbed soil samples the saturated hydraulic conductivity and aggregate stability as determined by the wet-sieving method always increased with improved soil structure. However, for undisturbed samples, capillary intake rate and aeration porosity were better indices of soil structure. Sharma and Uehara (1968) also observed similar results, but their findings suggested that soils with larger aggregates should have higher saturated hydraulic conductivities. Not only is organic matter important in the stabilization of soil aggregates,  but also essential for the maintenance of a high hydraulic  conductivity of soil. Organic matter was observed to decrease the slaking sensitivity of silt soils (Boekel, 1963). The coating of silt by organic colloids might have been responsible for the decrease. This could be possible if organic matter was present in abundance (Hamblin, 1977). Poomia and Pal (1979) reported that saturated hydraulic conductivity of a sandy loam soil was higher after longterm treatments with farm yard manure than without manure treatments.  Becher and Kainz (1983) also reported similar results with manuring.  In organic soils saturated  hydraulic conductivity may decrease by an order of  magnitude from partly decomposed to a well decomposed state (Walmsley and Lavkulich , 1975; Bonsu, 1984).  80  Longterm cultivation and subsidence of the organic soil of the Serpentine-Nicomekl area have decreased the depth of the organic layer to the extent that the mineral subsoil has been brought up and mixed with the organic layer during tillage and harvesting operations. Laboratory studies have shown that mixing mineral soil with an organic soil increases the rate of decomposition of the organic material (Broadbent, 1960). The prime goal of an}' soil management  strategy is to maintain good soil  tilth, of which structural stability and hydraulic conductivity form an important integral part. From the literature, we see that the structural characteristics and saturated  hydraulic conductivity of mineral soils, as influenced by organic matter,  have been explored to some appreciable extent. However, similar information on organic soils is scanty. More importantly, a knowledge of the structural and hydrologic characteristics of an organic soil in transition towards becoming a mineral soil is essential for the future land use plan of the organic soil in question. The idea of using air to water permeability ratio as a measure of structural stability of soils was initiated by Reeve (1953). Hamblin (1932) used air to water permeability ratio as an index of soil structural stability. In a perfect stable medium such as sand, air to water permeability ratio is expected to be unity. The larger the ratio the less stable the soil structure. The gradation of organic matter content from mineral soil ridges to the depressions of some of the shallow organic soils of the Serpentine-Nicomekl area provides a suitable avenue to derive information on the relationship between saturated  hydraulic conductivity and structural stability. The objectives,  therefore,  of this study were to explore the relationships between aggregate stability, saturated  hydraulic conductivity, and mineral matter content of natural aggregates  derived from a shallow organic soil using a regression approach, and,  further,  81 explain the influence of a fine-textured mineral matter, resulting from mixing of the organic surface layer with the mineral subsoil, on the slaking tendencies of the aggregates using air to water permeability ratio. Mineral matter, in this context, is defined as the percent sand, silt, and clay present in the organic soil.  3.2 M A T E R I A L S A N D M E T H O D S  3.2.1 G E N E R A L  PROCEDURE  A brief description of the site, sampling scheme, and method of sample preparation were given in chapter  1 of this thesis. The procedures for  determining aggregate stability, organic matter content, mineral matter  content  and HC1 digestible ash of the aggregates were also described in chapter this thesis. The laboratory values of saturated  1 of  hydraulic conductivity of the  aggregates predetermined by the constant head method prior to dispersing the aggregates (Appendix 5) were used for this chapter. The particle density of the aggregates was determined by the pycnometer method using deaired distilled water. Every determination was replicated three times. The p H of the soil was measured using a standard p H meter (pHm 62) at the ratio of 4 parts of water to 1 part of soil that was passed through a 2 mm sieve.  3.2.2 D E T E R M I N A T I O N O F 0.1N N A O H E X T R A C T A B L E  HUMIC  SUBSTANCES. On March 22 1984, 20 samples were obtained at the study site along one transect. Sampling was done systematically at a spacing of 10 m. On October 17, 1984 another  10 samples were obtained at the site . A l l samples  were air dried and prepared as described in chapter  1 of this thesis and used  82  for pilot studies. The humic materials of these 30 samples were extracted by adding 60 ml of 0.1N N a O H to 3 g of the air dried samples (<0.5 mm) in 275 ml plastic bottles. Extraction was done by shaking gently for 14 hours using a mechanical shaker, and centrifuging at 2,200 rpm for 20 minutes. The extracts were decanted and the residues washed with distilled water and then recentrifuged. The washing and recentrifugation were repeated two times with distilled water and then with 60 ml 0.05N H C l , and lastly with distilled water. The residues after extraction of the humic substances were transferred into evaporating dishes, evaporated to dryness, and dried in an oven at 105°C for 24 hours. The oven dry mass of a 3 g subsample was obtained for each sample. The difference between the oven dry mass of the 3 g subsample and the oven dry mass of the residue was used to represent the 0.1N N a O H extractable humic substances. The extract was expressed as a percentage of oven dry mass of the -soil.  3.2.3 D E T E R M I N A T I O N O F A I R TO W A T E R P E R M E A B I L I T Y  RATIO.  A sketch of the apparatus used for the air permeability measurement is shown in F i g . 3.1. The laboratory air permeameter was similar to that as descibed by Reeve (1953). The capacity of the tank was 0.1846 m . The tank 3  gauge pressure was measured with a water-filled manometer. The tank was connected to the laboratory air source through a valve system. The air permeameter consisted of a 30 cm long and 5 cm inside diameter acrylic plastic cylinder. The soil aggregates were packed in the permeameter  and  then vibrated with an air-driven vibration inducer at a constant pressure of 10 psi (69 kPa) for 20 seconds. The aggregates were supported by a 3 cm deep layer of gravel (>  2 mm, <4 mm) supported by a screen. A 5 cm thick layer  Compressed  air  Thermometer  Water manometer  Acrylic plastic  Soil  cvlinder  aggregates  Coarse  sand  Gravel Screen Rubber s t o p p e r  fc\r\L x f o r the  F  ]  ? ? d i a g r a m of the a p p a r a t u s d e t e r m i n a t i o n of a i r p e r m e a b i l i t y . c  h  e  i  a  t  i  c  84  of coarse sand (0.5-1.0 mm) supported the aggregates against the gravel surface. The length of the aggregate columns ranged from 10.5 to 14 cm. The column was attached to the outlet of the tank through a rubber stopper. Initially the outlet valve to which the column was attached was closed. The valve connecting the air source to the tank was opened to increase the pressure in the tank to 40 cm of water gauge pressure. The outlet valve to which the column was attached was then opened, and the time required for the manometric pressure to drop from 25 to 5 cm of water was recorded. The run was repeated three times for precision. The air permeability was calculated by the formula: k = [ln(y a  1  /y )](Tj 2  LV)/(APt), in which k is the air permeability a a  (m ), T? is the dynamic viscosity of air at the air temperature (1.84X10" a 2  m"  1  5  kg  s" ), L is the length of soil column (m), V is the volume of the tank 1  (m ), A is the area of soil (m ), P is the atmospheric air pressure (Pa), t is 3  2  the time (s), y  t  is the manometric pressure at time zero, and y  2  is the  manometric > pressure at time t, both in cm of water. The formula used for computing the air permeability is based on the assumption that the pressure distribution throughout the soil column corresponds to the steady state pressure distribution existing in the tank (Kirkham, 1946). This condition is not exactly true, because a finite time is required for any pressure change in the tank to be reflected at points in the sample (Kirkham, 1946). However, inspite of the admittedly non-steady state flow condition in the soil column, Kirkham (1946) has shown that if the pressure variations of the manometer used for the test are small compared to one atmosphere, this assumption is justified. In reality, the formula used for computing the air permeability is an approximate form of an exact but relatively complicated formula (Kirkham, 1946). If the gauge pressure (in the manometer) in excess of atmospheric pressure is  85  small, then the error associated with the formula used for computing the air permeability can be represented as: (y ~ 1 )/[2Pln(y,/y )] (Kirkham, 1946). Under V  2  2  the condition of the experiment, the error was only 0.60 percent. Using the constant head method, the laboratory permeability of the aggregate columns to water was determined on the same samples used for the air permeability measurements. The method was similar to that described for the laboratory determination of the saturated hydraulic conductivity in chapter 1, except that the temperature of the water was taken into consideration. The thermostat controlling the temperature of the tap water was faulty but was unnoticed. Therefore, the aggregates were wetted by capillarity overnight with tap water whose temperature was 13 °C under the laboratory temperature condition of 20°C . The error associated with using viscosity of water at 13°C was estimated to be around 10%, if it is assumed that the estimated mean temperature of the water in the column after allowing the tap water to flow for 5 minutes was 16° C. The error is expected to be smaller for the more permeable samples, since the possibility for the temperature of water in the column to be close to that of the tap in this instance is expected to be high. The water permeability was calculated by the formula: k = (Q L r\ )/ (A t A w w H p . g), where k is the water permeability (m ), Q is the volume outflow of w w 2  water (m ), L is the length of the soil column (m), TJ 3  water at its temperature (1.2X10"  3  kg m~  1  s" ), 1  is the viscosity of w A is the area of the soil  (m ), t is the time of flow (s), A H is the rrydraulic head difference across the 2  column (m), p is density of water at its temperature (1000 kg m~ ),and g is w 3  the gravitational acceleration (9.81 m s~ ).  From the k and k values, air to a w water permeability ratios were obtained for samples taken on transects 1 and 2. 2  86 3.3 R E S U L T S A N D DISCUSSIONS  3.3.1 R E L A T I N G A G G R E G A T E S T A B I L I T Y TO M I N E R A L M A T T E R  CONTENT.  The data for aggregate stability, organic matter content, mineral matter content, and HC1 digestible ash and their respective standard deviations are presented in Appendix 4. The data represent means of three replicates. The data for soil p H are given in Appendix 2. The p H represents a mean of two replicates. For the aggregate stability data, the coefficient of variation within the replicates in general ranged between 0.002 and 0.126. There was, however, one value of 0.21. The soil p H ranged from 4.50 to 5.44. The low p H values suggested that the soil had insignificant carbonate content, and, therefore, the ignition method used for the determination of the organic matter content was justified. The p H tended to decrease with increased distance from the mineral soil ridge, but not consistently. A l l regression analyses were done using the General Linear Model (GLM) of Statistical Analysis System (SAS Inc. 1982). Fig. 3.2 shows the linear regression of HCl-digestible ash on organic matter content. The sum of percent organic matter content, percent mineral matter content not digested by HC1, and HC1 digestible ash must be 100%. The regression equation relating HC1 digestible ash and organic matter content was expressed as D A =  7.56 +0.047 O M ,  where D A is the percent HC1 digestible ash, and O M is the percent organic matter content. Organic matter content accounted for 61.5 percent variation in HC1 digestible ash. The correlation coefficient between HC1 digestible ash and organic matter content was positive (r = 0.784), and highly significant (P = 0.0001). To a reasonable approximation, the results show that HC1 digestible ash predicts the mineral ash arising from igniting organic matter. The low correlation between  F i g . 3.2. H C l d i g e s t i b l e a s h v e r s u s o r g a n i c m a t t e r c o n t e n t . (The l i n e i s t h e r e g r e s s i o n l i n e w i t h the given e q u a t i o n ) .  88 H C l digestible ash and organic matter content suggested that HCl-extractable elements such as Ca, M g , M n , and others that were not directly linked with organic matter, but linked with clay minerals and not lost during ignition, might have had a considerable influence on the relationship. To a good approximation, the mineral matter content may be estimated as the difference between the total ash  and H C l digestible fraction. The relationship between 0.1N N a O H extractable humic substances and  organic matter content for the samples used for the pilot studies is shown in Fig.3.3. The regression of 0.1N N a O H extractable humic substances on organic matter showed that 93 percent variation in 0.1N N a O H extractable humic substances can be explained by organic matter alone ( r  2  =0.927). The regression  equation describing the relationship between 0.1N N a O H extractable humic substances and organic matter content was expressed as N A =  1.8 +  0.537  O M , where N A is the percent 0.1N N a O H extractable humic substances in a 3 g subsample, and O M is the percent organic matter content in the sample. The correlation coefficient was positive and highly significant (P = 0.0001). The regression shows that 0.1N N a O H extractable fraction is a constant fraction of the organic matter content. However, the 0.1N N a O H extractable substances may also contain mineral substances. Besides, the 0.1N N a O H extractable fraction has been described as an organo-mineral complex (Yefremova, 1984). Fig.3.4 shows the linear regression of aggregate stability on mineral matter content. The mineral matter content accounted for 93 percent variation in aggregate stability ( r  2  =0.934). The regression equation describing the  relationship between aggregate stability and mineral matter content was expressed as A G G =  108.25 -  0.532 M N , where M N is the mineral matter content in  percent dry mass, and A G G is the aggregate stability in percent by dry mass. The correlation coefficient was negative and highly significant (P=0.0001). The  40 NA=1.80 + 0 . 5 3 7 OM r = 0.963 * * *  S.E.  1 0  1  i 10  1  1 20  = 0.0282  1  1 30  ORGANIC MATTER  1  1 40  1  1 50  1  CONTENT(%)  F i g . 3.3. 0.1N NaOH e x t r a c t a b l e f r a c t i o n o f a 3 g subsample v e r s u s o r g a n i c m a t t e r c o n t e n t . (The l i n e i s t h e r e g r e s s i o n l i n e w i t h t h e given equation).  1 60  F i g . 3.4. Aggregate s t a b i l i t y versus mineral m a t t e r c o n t e n t . (The l i n e i s t h e r e g r e s s i o n l i n e w i t h the g i v e n e q u a t i o n ) .  91 results showed a linear negative relationship between aggregate stability and mineral matter content. The texture of the mineral matter on the mineral ridge is silty clay (54% clay; 46% silt). A s the thickness of the organic surface layer decreases due to oxidation loss, tillage and harvesting operations tend to bring up mineral matter from the shallow organic areas. (In Fig.2.13 of chapter  1 of this thesis , the  systematic decline of mineral matter from the mineral soil ridge to the organic depression is well illustrated). The mineral matter of weak structure may be eroded from the mineral soil ridges and deposited in the organic depressions during rainfall events to result in the formation of surface seal in the depressions. The foregoing results have confirmed the importance of the effect of organic matter content on the stabilization of the natural aggregates of a well -decomposed organic soil that has been influenced by a fine-textured mineral subsoil. Increasing mineral matter content was associated with decreasing aggregate stability. In Table 2.10, the lowest organic matter content is 14 percent, which, for ordinary mineral soils, is high enough for the aggregates to be highly stable if the organic matter and the mineral matter are allowed to be in contact for a long period of time. This does not happen because "new" mineral matter is brought up every year with tillage and harvesting operations. Some recent studies in mineral soils also showed a significant increase in aggregate stability with increasing levels of organic matter content (Zilva et al. 1982; Chaney and Swift, 1984; Mishack et al. 1985). It is not the intent of this thesis to go into detailed discussions on the mechanism of clay-organic complex formation, which chemically explains in part the mechanism of aggregate stabilization. The reader interested in the detailed information on this topic may refer to Harris et al. (1966), Schnitzer and Khan  92  (1972), Theng (1979), and Sposito (1984). However, it is of interest to seek an explanation for the consistent decrease in the aggregate stability of the organic soil with increasing mineral matter content. One mechanism of aggregate stabilization involves an interaction of humic substances with clay to form organo-clay complexes (Schnitzer and Khan, These clay-organic complexes protect the humic substances from further attack (Sen, 1961; Schnitzer and Khan,  1972).  biological  1972). Chaney and Swift (1984)  extracted humic substances from mineral soils with 0.1M sodium and 0.1M N a O H , respectively, and determined the carbohydrate  pyrophosphate content by  succesive hydrolysis with 12M and 0.5M sulphuric acid. They observed that these organic extracts were highly correlated with aggregate stability. Chaney and Swift (1984) again realized that it was not possible to distinguish whether or not one organic component extracted was more important than the other. They  therefore  concluded that organic matter alone was sufficient as a diagnostic criterion to identify structural instability of the soils they studied. These studies also showed that for the shallow, well-decomposed organic soil that had been mixed with a fine-textured mineral soil, 0:1N N a O H extractable humic substances and organic matter content could explain the same variation in aggregate stability. Because of the colloidal properties of humic substances and their high affinities for cations, the alkali extracts of humic substances in soils are frequently in association with large amounts of finely dispersed clay (Andreaux, 1982). Besides, alkali extracts of humic substances in soils have been described as "organo-mineral complexes" (Yefremova, 1984). Yefremova (1984) used the humic substances extracted with 0.1N N a O H in characterizing the quality of structure of forest peat soils in the Soviet Union. Some conclusive evidence exists that montmorillonite has a greater potential of forming aggregates of higher water stability than kaolinite (Peterson,  93  1946). The mineral fraction of the organic soil used for this study  contains  montmorillonite as the dominant clay mineral (Clark et al. 1961; Bonsu, 1984). Therefore it is possible to achieve high water stability in this soil provided sufficient organic matter is available to form clay-humic complexes. • It may be theorized that clay-organic complexing can go on as long as the clay mineral has a component organic colloid to interact with. When the amount of clay mineral is much higher in proportion than the organic colloid, there is a possibility that surplus clay would be left unbonded. Discounting the influence of sesquioxides, aggregates formed by clay alone without organic matter interaction can be less stable (Sen, 1961). The theory propounded may explain in part why aggregate stability of the shallow organic soil decreased  consistently  with increasing mineral matter content. The review work of Greenland (1965) showed that about 52 to 98 percent of the total soil carbon was in the form of clay-organic complexes. It is known that the role of organic matter in stabilizing aggregates is less pronounced in soils containing large amounts of clay. Acton et al. (1963) observed that polysaccharides produced aggregates of higher stability in soils low in clay than in soils of high clay content. Another possible explanation accounting for the decrease in aggregate stability with increasing mineral matter content as observed in this study may be sought from the high silt content of the mineral fraction. Stable aggregate formation does not take place in sand or silt in absence of organic or inorganic colloid (Baver et al. 1972). This implies that the silt that remains  after  clay-organic colloids interaction would be most dispersible. Another role of organic matter that may be applicable in this study is that humic substances may coat the surfaces  of the mineral particles, which  upon irreversible dehydration, may lead to interparticle bonding (Mitchell, 1976).  94 The presence of hydrophobic coatings on mineral soil particles due to organic matter was reported by Roberts and Carbon (1972), and Miller and Wilkinson (1977). Giovannini et al. (1983) observed that a hydrophobic fraction  extracted  with benzene from a naturally occurring water repellent soil in Italy decreased the water stability of the soil aggregates as determined by the  wet-sieving  method.  3.3.2 R E L A T I N G S A T U R A T E D H Y D R A U L I C C O N D U C T I V I T Y O F T H E A G G R E G A T E B E D S TO A G G R E G A T E S T A B I L I T Y A N D M I N E R A L MATTER  CONTENT.  The data for the saturated hydraulic conductivity of the aggregate beds are shown in Appendix 5. The variation of the saturated hydraulic conductivity of the aggregate beds with distance from the mineral soil ridge is presented in Fig.3.5. The saturated hydraulic conductivity increased almost by 20 times from the mineral soil ridge to the organic depression. In addition, it can be visually observed that saturated hydraulic conductivity of the aggregate beds increased exponentially with distance from the ridge. The correlation between saturated hydraulic conductivity of the aggregate beds and aggregate stability was sought using G L M of S A S . The linear correlation coefficient between saturated hydraulic conductivity and aggregate stability was positive (r = 0.757) and highly significant (P=0.0001).  The linear  equation describing the relationship between saturated hydraulic conductivity and aggregate stability was expressed as K  =  —2.52 +  0.03 (AGG), where K  s  is s  the saturated hydraulic conductivity (m /day), and A G G is the aggregate stability expressed in percent dry mass. When saturated hydraulic conductivity was log-transformed,  and the regression  of the log-transformed  hydraulic conductivity  on aggregate stability obtained, the correlation was highly improved (r = 0.918).  20  18-  • X  A X A 6-  A  X  A  Legend A  25  A • X 50  • A  8 X  75  100  125  150  175  200  DISTANCE FROM MINERAL RIDGE (m)  Fig. beds  3 . 5 . V a r i a t i o n o f K of with distance a l o n g t h e s  the aggregate transects.  A  TRANSECT1  X  TRANSECT2  •  TRANSECT3  96  Fig. 3.6 shows the semi-log relationship between K  s  and wet-sieved aggregate  stability. In the log-transformed case aggregate stability alone accounted for 84.3 percent variation in saturated hydraulic conductivity of the aggregate beds. Thus an exponential relationship shown in Fig. 3.6 better describes the variation of saturated hydraulic conductivity with aggregate stability. A n exponential relationship between saturated hydraulic conductivity and bulk density following compaction was reported by Taylor and Henderson (1959). The linear correlation coefficient between saturated hydraulic conductivity and mineral matter content was negative (r =-0.758), and highly significant (P = 0.0001). The linear relationship accounted for 57.4 percent variation in saturated hydraulic conductivity due to mineral matter content. When saturated hydraulic conductivity was log-transformed and used in the regression analysis, the correlation between the log-transformed K  and mineral matter content greatly s improved (r=-0.911). The semi-logarithmic relationship between K and mineral s matter content is illustrated in Fig.3.7. The exponential relationship shown in Fig. 3.7 accounted for 83 percent variation in K  s  due to the influence of mineral  matter content. These results show that K  s  of the aggregate beds decreases with  increasing mineral matter content. Logically, it follows that  of the aggregate  beds must increase with increasing organic matter content, since organic matter and mineral matter are almost complementary to one another (Fig.2.13). If the interaggregate porosity is the dominant factor controlling K ^ , then K  s  values resulting from aggregates of the same size and roughness should be  expected to be the same. The data for the particle density, the bulk density, and the porosity of the aggregate beds, versus distance from the ridge are presented in Table 3.1. The porosity of the aggregate beds ranged between percent and 58.8 percent, with a mean and a standard  48.4  deviation of 53.3 percent  F i g . 3.6. K of a g g r e g a t e b e d s v e r s u s a g g r e g a t e s t a b i l i t y . (The l i n e i s t h e r e g r e s s i o n l i n e with the given e q q u a t i o n ) .  F i g . 3.7. K of a g g r e g a t e b e d s v e r s u s m i n e r a l m a t t e r c o n t i n t . (The l i n e i s t h e r e g r e s s i o n l i n e with the g i v e n e q u a t i o n ) .  T a b l e 3.1. B u l k d e n s i t y of a g g r e g a t e beds and p o r o s i t y ( f ) of t h e a g g r e g a t e beds.  f a  particle  ),  Transect 2  Transect 1  D i st.  (p  density ( p ) , g  Transect 3  p  t from r i dge (m)  (kg/m )  (kg/m  853.3+15.3 25  )  (%)  (kg/m )  (kg/m  1 ,842+8.5  53.7  846.7+30.6  816.7132.1  1 .708+11.3  52.2  50  723.3± 1 1. 6  1.650+7.8  56.2  75  730.0+17.3  1,506+44 . 6  5 1.5  100  693  1.490+5.0  53.4  125  616.7125.2  1,336±17.0  53  150  526.7+15.3  1 . 164±14 .8  54  175  516.7+5.8  1 . 17417 .8  20O  540.0+10.0  3+15  3  1.156±17.0  )  (%)  (kg/m  (kg/m  1.839+15.6  54.0  890.0126.5  1 .8521 14.8  51.9  880.0110.0  1,822110.6  5 1.7  906.7125.2  1.80611.8  49 . 8  796.7  1,710154.4  53.4  770.O+10.0  1,648130.4  53 . 3  683.3+25.2  1 ,46211 1 . 3  53.3  633.3+11.6  1,43914 1 .0  56.0  686.7+23.1  1 . 4 14 + 7 . 8  5 1 4  693.3115  1,34414 . 2  48 .4  8  703.3120.8  1,376110.6  48  9  690.0117.3  1,417126.9  51.3  8  560.0±17.3  1 .228119. 1  54.4  600.0110.0  1,30114.2  53 . 9  56.0  553.3+15:3  1 , 168124.8  52.6  540.0169.3  1, 182 + 22 . 6  54 . 3  53.3  526.7+11.5  1 .278+12.0  58.8  540.0117.3  1,250+12.0  56 . 8  ±58.6  3  )  (%)  SO  100 and 2.35 percent, respectively. The coefficient of variation from the mean porosity of the aggregate beds was only 4.4 percent. Therefore, it was  reasonable to  assume that the interaggregate porosity of the aggregate beds was  not  significantly different. The relationship between permeability and porosity may  be explained by  the Kozeny-Carman equation (Lagerwerff et al., 1969). The Kozeny-Carman equation may  be written as (Lagerwerff et al., 1969) k = e / (M t^, S ), in 3  2  2  which k is the permeability, e is the total pore fraction of the medium, M is the empirical factor representing the shape and size of the particles, t^, represents the empirical tortuosity factor of the flow path, and S is the specific surface of the flow bed. The Kozeny-Carman equation is semi-empirical, and it is supposed to be valid for viscous flow through unconsolidated granular media in which the pores are of uniform size and evenly distributed, and the soil is not subject to swelling (Lagerwerff et al., 1969). Re-examining the Kozeny-Carman equation , the other important factor which may  bring about differences in permeability is the tortuosity factor, which  can be influenced by swelling. Swelling was  observed in the aggregate beds  during wetting. Table 3.2 shows the data for air and water permeabilities, air to water permeability ratios, and the relative swelling of the aggregate beds for two transects. The relative swelling was  defined as ^ - L ^ / L , , where L  2  is the  length of the column after swelling, and L, is the length before swelling has occurred. The differences in K  s  values of the aggregate beds may  be attributed to  the swelling and slaking tendencies of the aggregates. Air does not react with soil to cause slaking, but water being polar may  react with soil to cause  swelling and slaking. The consequence of the swelling-slaking phenomenon would be a decrease of permeability due to blocking of pores by the dispersed particles.  T a b l e 3.2. A i r (k ) a n d w a t e r (k ) p e r m e a b i l i t i e s , a i r t o w a t e r p e r m e a b i l i t y r a t i o (k /k ) , a n d R e l a t i v e s w e l l i n g (L -L,/L,) of t h e aggregates. 2  a  D1st .  k  f rom  ( 10  (ni  a  )  k  )  { 10  - 10  w  (m - 12  )  k  a  /k w  )  w  Re 1 a t 1ve  k  Swe1 l i n g  ( 10  (m  a  - 10  )  k  )  ( 10  w  (m  )  k  a  /k w  - 12  Re 1 a t i v e  Swe11i ng  )  r 1 dge (m) 1 .97  2 .25  87 .6  4 .3  2 .95  4.15  71.1  4 .5  25  3.90  17 .00  22.9  5 .5  3.82  3 .08  124 . 0  4 .3  50  2.90  4 . 30  67 .4  4.3  3.43  6 . 18  56 . 2  4 .5  75  2 .49  10. 70  23 . 3  4.3  2 . 97  14.01  2 1.2  3 .4  100  3.28  9 . 44  34 . 8  4 .2  3 . 36  14 . 3 0  23 . 5  8 .3  125  2 . 45  6 . 22  39.4  9.5  4 . 55  1 1 .60  39 . 2  3.8  150  3.53  65 . 30  5.5  13.0  4 . 53  12 . 0 0  37 .8  15.4  175  4 .03  51 . 9 0  7 .8  17.4  4 . 85  52 .80  9.2  12.5  200  2.45  4 3 . 70  5.6  8 . 7  4 . 85  56.80  8 .2  10.7  102 The extent of slaking will depend on the stability of the aggregates. Conceptually, in a perfect stable medium such as sand, air to water  permeability  ratio is expected to be unity. But this usually does not happen because of air entrapment with the passage of water. From Table 3.2, the mean air permeability of the aggregate beds was 3.46X10 ^ ^ m , with a standard deviation of 8.71X10 ^ m , giving a coefficient 2  2  of variation of 25.2 percent. Also, the mean water permeability of all the aggregate beds together was 2.11X10^^ m , with a standard deviation of 2  2.16X10  -1  1  m , giving a coefficient of variation of 101 percent. Thus, it was 2  noticed that whereas the air permeabilities of the aggregate beds did not vary much, the water permeabilities of the aggregate beds differed significantly. The lowest water permeability was obtained for samples on or near the mineral soil ridge. The organic matter content was also lowest on or near the ridge (Fig. 2.13). The variation of air to water permeabilit}  7  ratio with distance from the  mineral soil ridge is illustrated in Fig. 3.8. The ratio was observed to be relatively high on or near the ridge. Fig. 3.9 shows the relationship between  air  to water permeability ratio and wet-sieved aggregate stability. The correlation between  air to water permeability ratio and wet-sieved aggregate stability was  negative and highly significant (r = -0.762; P=0.0004).  The regression  equation  describing the relationship between air to water permeability ratio and aggregate stability was k /k a w k  w  =  232.51— 2.34 A G G , in which k  a  is the air permeability,  is the water permeability and A G G is the wet-sieved aggregate stability. The  results indicated that 58 percent variation in air to water permeability ratio could be ascribed to aggregate stability. Fig. 3.10 illustrates the linear regression of air to water permeability ratio on mineral matter content. Mineral matter content explained 54 percent variation  103  F i g . 3.8. V a r i a t i o n of a i r t o water p e r m e a b i l i t y r a t i o w i t h d i s t a n c e from mineral s o i l ridge.  the  F i g . 3 . 9 . A i r t o water p e r m e a b i l i t y r a t i o versus aggregate s t a b i l i t y . * ( T h e line is r e g r e s s i o n l i n e w i t h the g i v e n e q u a t i o n ) .  H O  A  ka/kw=1.13 MN -17.68  120-  r=0.737; p=0.0005  g <C cr  100-  S  -  E  -  = 0.2600  >i—  A  _j  CD  80-  UJ  ~l  10  I  I  I  I  20  30  40  50  1  60  1  70  1  80  MINERAL MATTER CONTENT (%)  F i g . 3.10. A i r t o water p e r m e a b i l i t y r a t i o v e r s u s m i n e r a l m a t t e r c o n t e n t . (The l i n e i s the r e g r e s s i o n l i n e w i t h t h e g i v e n e q u a t i o n )  106 in air to water permeability ratio. As expected, the correlation between air to water permeability ratio and mineral matter content was positive (r = 0.737). The linear regression equation describing the relationship between air to water permeability ratio and mineral matter content was k /k = a w  1.13 M N — 17.63.  The positive correlation between air to water permeability ratio and mineral matter content suggested more slaking and reduction in permeability with increasing mineral matter content. The air to water permeability ratio of a coarse sand used as a check was 1.5. The deviation from unity of air to water permeability ratio of the structureless coarse sand could be attributed to air entrapment. Even though air to water permeability ratio and aggregate stability were significantly correlated the correlation coefficient was not as high as expected. The sensitivity of using air to water permeability ratio as a measure of structural stability of soil was questioned by Hutson (1982). Perhaps, air to water permeability ratio can better be used for assessing the sensitivity of soils to slaking. Fig. 3.11 shows the relationship between relative swelling and organic matter content of the aggregates.  The correlation between relative swelling and  organic matter content was positive (r = 0.883) and highly significant (P<0.001). Organic matter content explained 78 percent variation in relative swelling of the aggregates in the columns. Therefore if swelling was the dominant cause of slaking of the aggregates,  we would expect air to water permeability ratio to  increase with increasing organic matter content. But this was contrary to the results. The higher levels of organic matter in some of the samples are due to the presence of partly decomposed plant materials in the organic soil. Two types of pore systems may be distinguished in the aggregate beds: the internal pores within the partly decomposed plant tissue, and the interaggregate  pores. The  F i g . 3.11. R e l a t i v e s w e l l i n g v e r s u s o r g a n i c m a t t e r c o n t e n t o f t h e a g g r e g a t e s . (The l i n e i s the r e g r e s s i o n l i n e w i t h the g i v e n equat i o n ) .  108 internal pores give rise to intracellular moisture content (Romanov, 1961). Therefore in organic soils, the water storage capacity can be substantially higher than the total porosity calculated from particle density and bulk density (Romanov, 1961). Thus the greater swelling observed in the samples containing higher levels of organic matter was predominantly due to intracellular absorption of water by the partly decomposed plant tissue. For  swelling to occur in mineral soils, water molecules must be adsorbed  by the active clay. The three mechanisms accounting for the adsorption of water molecules by clays may be described as (Hillel,  1980) electrostatic attraction of  the dipolar water molecules to charged sites on the clay surface and in the intermicellar spaces; hydrogen bonding to exposed oxygen atoms on the clay crystal; and the hydration of adsorbed cations associated with the clay. In the first layer of water molecules on the clay, the adsorption of water may be dominantly controlled by electrostatic attraction, hence the water molecules are held with great tenacity. The second and the subsequent layers of water molecules may be held by hydrogen bonding, causing the attractive force field to decrease with distance from the clay. Swelling pressures may be developed as a body of clay in a confined state sorbs water. The swelling pressures are related to the osmotic pressure difference between the double layer and the external solution (Hillel, 1980). As each clay micelle expands, its negatively charged ions repel those of the  adjacent  micelle, and thus the micelles tend to push each other apart. Since the body of clay is confined, the internal effect of this intermicellar repulsion is closing of the larger pores. The swelling pressures are opposed by the inter-particle cohesive forces (Berezin et al. 1983). If the swelling pressures dominate the cohesive forces between the particles, the external effect of the intermicellar repulsion is swelling, which results in aggregate collapse and slaking (Emerson, 1977; Berezin  109  et al. 1983). The swelling-slaking phenomenon can also be explained in terms of organic anions-metal ions compbxing (Oades, 1984). The complexing of trivalent and divalent ions by organic anions reduces the positive sites available for bonding by the clay lattice. In addition, the adsorbed organic anions increase  the  negative charges on the colloidal surfaces, thus increasing the thickness of the diffuse double layer. The repulsive effect of the negative charges favours slaking. The slaked particles block some of the water transmission pores. Thus the overall consequence of the swelling-slaking phenomenon is the reduction of permeability of the soil to water. Montmorillonite is the dominant clay mineral of the mineral subsoil of the organic soils of the Serpentine-Nicomekl area (Clark et al. 1961; Bonsu, 1984). Since montmorillonite is a swelling clay, slaking of mineral particles in the samples of higher mineral matter content cannot be discounted. The blocking of water transmission pores by the slaked particles was likely responsible for the decrease in the saturated hydraulic conductivity of the aggregate beds with increasing mineral matter content. In the samples containing higher levels of organic matter , swelling was markedly due to intracellular absorption of water by the partly decomposed plant tissue. Thus, slaking was minimized considerably and saturated hydraulic conductivity was not drastically affected by swelling. In Chapter  1 of this thesis, a procedure developed to test the sealing  behaviour of the shallow organic soil revealed that the sealing tendencies of the organic soil increased with increasing mineral matter content. The index derived to quantify the susceptibility of the organic soil to dispersion and sealing correlated significantly with aggregate stability. Collateral to the in chapter  1, the results in chapter  findings  reported  2 of this thesis have confirmed the  deleterious effect of high mineral matter content on the structural and hydrologic behaviour of the shallow organic soil.  110  3.4  CONCLUSIONS These studies confirm the importance of good structural characteristics for  saturated water flow in the shallow organic soil of the Serpentine-Nicomekl area of the Lower Fraser Valley. Maintenance of higher organic matter content is a prerequisite for soil structural management of the shallow organic soil. Higher levels of mineral matter leads to a weaker soil structure and a decrease in saturated hydraulic conductivity. The slaking tendency of the shallow organic soil increases with increasing mineral matter content. The blocking of pores by slaked particles whenever the soil is wetted is likely to be responsible for the low saturated hydraulic conductivity. Mineral matter eroded from the mineral soil ridges during rainfall events is deposited in the depressions to increase the mineral matter content of the organic depressions. The presence of mineral matter enhances surface sealing and ponding in the organic depressions. Spreading of mineral matter from the ridges to the organic depressions during levelling is likely an inappropriate management practice. The dispersion of the fine-textured mineral fraction during rainfall can result in surface sealing, reduced infiltrability and ponding in the organic depressions. The economic consequence of the sealing-ponding events is a loss of "opportunity days" to the farmer.  4. SUMMARY First, a physically based model was formulated and tested experimentally to describe the mechanism of surface seal formation of soil in a laboratory column. Next, an index derived from the model was related to aggregate stability and mineral matter content of a shallow organic soil with the aim of finding the effect of mixing mineral matter with the organic layer through management on the structure and the hydrologic behaviour of the shallow organic soil. Second, the relationships between saturated hydraulic conductivity and aggregate stability and mineral matter content of natural aggregates derived from the shallow organic soil were established using regression analyses. The relationship between aggregate stability and mineral matter content was similarly established. Furthermore, the concept of air to water permeability ratio was used to explain the slaking tendencies of the aggregates as influenced by mineral matter content due to mixing. By assuming convective flow only, the theory developed indicated that for a constant hydraulic head difference across a laboratory column, the flux density of a filtrate flowing through a soil column is proportional to the inverse square-root of time. The proportionality constant which was designated "soil stability factor (E)" was shown to be related to the saturated hydraulic conductivity and the bulk density of the surface seal that has formed, and the concentration of dispersed solids. It was observed that the value of the constant sedimentation rate (k ) that would bring the theoretical values close to the 2  measured values was generally small, suggesting that k  2  should be a  non-constant parameter. In testing the validity of the theory using suspensions of known concentrations, a good agreement was found between theory and experiment.  Ill  112  The soil stability factor was exponentially related with aggregate stability and mineral matter content. Whereas the relationship between the soil stability factor  and aggregate stability gave a positive exponent with a positive correlation,  the relationship between the soil stability factor and mineral matter content gave a negative exponent, with a negative correlation. The correlations were all highly significant (P = 0.0001). The relationships between saturated hydraulic conductivity of the aggregate beds and aggregate stability and mineral matter content were also described  as  exponential. Whereas the relationship between saturated hydraulic conductivity and aggregate stability gave a positive exponent, the relationship between saturated hydraulic conductivity and mineral matter content gave a negative exponent. The correlation coefficient between saturated hydraulic conductivity and aggregate stability was positive and highly significant (P = 0.0001). The correlation between saturated hydraulic conductivity and mineral matter content was highly significant (P=0.0001), but negative.  also  These results show that higher levels  of mineral matter are a limitation to good structural characteristics  coefficient  and hydrologic  of the shallow organic soil. However, it is possible for the soil  containing high mineral matter to be stable i f the mineral matter and  the  organic matter are allowed to be in contact for a long period of time. Using air to water permeability ratios, the decrease in saturated hydraulic conductivity of the aggregate beds with increasing mineral matter content was explained in the context of the slaking tendency of the mineral matter fraction. Even though swelling was more pronounced  in the aggregates containing  levels of organic matter, intracellular absorption decomposed  plant tissue was likely responsible  of water by the  higher  partly  for the marked swelling with  minimal slaking of the aggregates containing higher levels of organic matter.  113  The deleterious effect of mineral matter on the structure of the shallow organic soil was confirmed by the strong negative correlation between aggregate stability and mineral matter content. The decrease in aggregate stability with increasing mineral matter content was explained in the context of clay-organic complexing. It was theorized that when the amount of clay mineral was much higher than the organic colloid available for complexing, the surplus clay unbonded by the organic colloid would be most easily dispersible. The high silt content of the mineral matter fraction was also likely to be an important factor contributing to the decreasing aggregate stability with increasing mineral matter content, since in the absence of organic and inorganic colloids silt alone is incapable of forming stable aggregates. In managing the shallow organic soils, it is appropriate to minimize mixing of the fine-textured mineral subsoil with the organic layer, since higher levels of mineral matter are likely to have an adverse effect on both the structural and the hydrologic behaviour of the shallow organic soil.  LIST OF REFERENCES 1.  Acton, C . J . , D . A . Rennie, and E . A . Paul. 1963. The relationship of polysaccharides to soil aggregation. Can. J . Soil Sci. 43: 201-209.  2.  Andreaux, F . 1982. Genesis and properties of humic molecules. In: Bonneau and Souchier (Editors)."Constituents and Properties of Soils". Acadamic Press, London. P P 109-139.  3.  Alberts E . E . , W . C . Moldenhauer, and G.R. Foster. 1980. Soil aggregates and primary particles transported in rills and interrill flows. Soil Sci. Soc. A m . J . 44: 590-595.  4.  Ball, D . F . 1964. Loss-on-ignition as an estimate of organic matter and organic carbon in non-calcareous soils. J . Soil Sci. 15: 84-92.  5.  Baver L . D . , W . H . Gardner and W.R. Gardner. 1972. "Soil Physics". 4th Edition, John Wiley and Sons, Inc. New York, P P 179-180; 140-146.  6.  Becher, H . H . and M . Kainz. 1983. Effect of long-term manuring on soil structure. Soils and Fert. V o l . 46(8):Abst.7679.  7.  Beck, Th. 1984. Microbiological and biochemical characteristics of agricultural soils part II. Relations to the organic matter content. J . Plant Nutrition and Soil Sci. 147: 467-475.  8.  Berezin, P . N . , A . D . Voronin, and Y e . V . Shein. 1983. A n energetics approach to the quantitative evaluation of soil structure. Soviet Soil Sci. (Translated). 15: 103-109.  9.  Bertrand, A . R . and S. Kamil. 1962. The effects of rainfall intensity on soil structure and migration of colloidal materials in soils. Soil Sci. Soc. A m . Proc. 26: 297-300.  10.  Boekel, P. 1963. The effect of organic matter on the structure of clay soils. Neth. J . Agric. Sci. 11: 250-263.  11.  Boggie, R. and R . A . Robertson. 1972. Evaluation of horticultural peat in Britain. Proc. 4th International Peat Congress, Vol.3, P P 185-192.  12.  Bonsu, M . 1984. Hydrologic behaviour of some structurally degraded organic soils. M.Sc. Thesis, Faculty of Graduate Studies, the University of British Columbia.  13.  Bouma, J . 1969. Microstructure and stability of two sandy loam soils with different soil management. Agric; Research Report No.724, Dept. of Regional Soil Sci., Agric. University, Wageningen, Neth.  14.  Brakensiek, D . L . and W . J . Rawls. 1983. Agricultural management effects on soil water processes part II: Green and Ampt parameters for crusting soils. Trans. A . S . A . E . 26: 1753-1757.  15.  Broadbent, F . E . 1960. Factors influencing the decomposition of organic soils in the California Delta. Hilgardia 29: 587-612.  16.  Bryan, R . B . 1969. The development, use and efficiency of indices of soil erodibility. Geoderma 2: 5-26. 114  115 17.  Chaney, K . and R.S. Swift. 1984. The influence of organic matter on aggregate stability of some British soils. J . Soil Sci. 35: 223-230.  18.  Chen, Y . and A . Banin. 1975. Scanning electron microscope (SEM) observations of soil structure changes induced by sodium-calcium exchange in relation to hydraulic conductivity. Soil Sci. 120: 428-436.  19.  Childs, E . C . 1940. The use of soil moisture characteristics in soil studies. Soil Sci. 50:239-252.  20.  Childs, E . C . 1942. Stability of clay soils. Soil Sci. 53: 79-92.  21.  Clark, J . C . , C . A . Gobin, and P . N . Sprout. 1961. Yellow mottles in some poorly drained soils of the Lower Fraser Valley, British Columbia. Can. J . Soil Sci. 41: 218-227.  22.  De Boodt, M . , De Lenheer, L . and D. Kirkham. 1961. Soil aggregate stability index and crop yields. Soil Sci. 91: 138-146.  23.  De Vries, J . 1983. Degradation effects of soil compaction. Proc. 8th British Columbia Soil Science Workshop. P P 91-121. B . C . Ministry of Agric. and Food, Victoria.  24.  Dong, A . , G . Chesters and G . V . Simsiman. 1983. Soil dispersibility. Soil Sci. 136: 208-212.  25.  Emerson, W.W. 1977. Physical properties and soil structure. In: Russell and Greacen (Editors). "Soil Factors in Crop Production in a Semi-arid Environment". P P 78-104, university of Queensland Press.  26.  Fairbridge, R.W. and C W . Finkl. 1979. (Editors), "The Encyclopedia of Soil Science", Part I, Vol. 12. Dowden, Hutchinson and Ross Inc., Stroudsburg, Pennsylvania.  27.  Giovannini, G . , S. Lucchesi and S. Cervelli. 1983. Water repellent substances and aggregate stability in hydrophobic soil. Soil Sci. 135:110-113.  28.  Greenland, D . J . 1965. Interaction between clays and organic compounds in soils. Part 1. Mechanisms of interaction between clays and defined organic compounds. Soils and Fert. 28 : 415-425.  29.  Hamblin, A . P . 1977. Structural features of aggregates in some East Anglian silt soils. J . Soil Sci. 28: 23-28.  30.  Hamblin, A . P . 1982. Soil water behaviour in response to changes in soil structure. J . Soil Sci. 33: 375-386.  31.  Harris, R . F . , G. Chesters and O . N . Allen. 1966. Dynamics of soil aggregation. Adv. in Agron. 18: 107-169.  32.  Hillel, D . 1980. "Fundamentals  33.  Hutson, J . L . 1982. Structural stability of diagnostic horizons in South African soils. South African J . of Science. 78 :453-454.  34.  Kamil, S. and A . R . Bertrand. 1962. Effect of rainfall energy on the permeability of soils. Soil Sci. Soc. A m . Proc. 26: 293-297.  of Soil Physics". Academic Press, Tronto.  116  35.  Kemper, W . D . 1965. Aggregate stability. In: Black, C A . (Editor)."Methods of Soil Analysis". Agronomy Monograph No.9, part 1. Academic press, New York. P P 511-519.  36.  Kirkham, D . 1946. Field method for determination of air permeability of soil in its undisturbed state. Soil Sci. Soc. A m . Proc. 11: 93-99.  37.  Lagerweff, J . V . , F . S . Nakayama and M . H . Frere. 1969. Hydraulic conductivity related to porosity and swelling of soils. Soil Sci. Soc. A m . Proc. 33: 3-11.  38.  Linden, D.R. 1979. A model to predict soil water storage as affected by tillage practices. Unpub. Ph.D. Thesis, University of Minnesota, St. Paul, Minnesota.  39.  Lishtvan, N . I . and T. G . Zuyev. 1982. Effect of mineralization on the hydrophilic properties of peat-bog soils. Soviet Soil Sci. (Translated). 14: 92-96.  40.  Luttmerding, H . A . 1981. "Soils of the Langley-Vancouver Map Area", Vol. 3. R A B Bulletin 18, B . C . Ministry of Environment, Kelowna.  41.  Maas, E . F . 1972. Organic soils of Vancouver Island. Contribution No.231, Agriculture Canada, Sidney Research Station, B . C .  42.  Marshall, T . J . 1958. A relationship between permeability and size distribution of pores. J . Soil Sci. 9: 1-8.  43.  McCalla, T . M . 1944. Waterdrop method of determining the stability of soil structure. Soil Sci. Soc. A m . Proc. 7: 209-214.  44.  Meredith, H . L . and H . Kohnke. 1965. The significance of the rate of organic matter decomposition on the aggregation of soil. Soil Sci. Soc. A m . Proc. 29:547-550.  45.  Middleton, H . E . 1930. Properties of soils which influence soil erosion. U . S . Dept. Agr. Tech. Bull. 178. 16pp.  46.  Miller, R . H . and J . F . Wilkinson. 1977. Nature of the organic coating on sand grains of nonwettable golf greens. Soil Sci. Soc. A m . J . 41: 1203-1204.  47.  Mishack, B . M . , E.R. Page and L C . Grieve. 1985. A comparison of soil aggregate stability tests using soils with contrasting cultivation histories. Commun. in Soil Sci. Plant Anal. 16: 315-322.  48.  Mitchell, J . K . 1976. "Fundamentals of Soil Behaviour". John Wiley and Sons, Inc. New York. P 175.  49.  Moore, I.D., C . L . Larson, and D . C . Slack. 1980. Predicting infiltration and micro-relief surface storage for cultivated soils. W R R C, Univ. of Minnesota, Bulletin No. 102.  50.  Oades, J . M . 1984. Soil organic matter and structural stability: Mechanisms and implications for management. Plant and Soil 76: 319-337.  51.  Peterson, J . B . 1946. The role of clay minerals in the formation of soil  117  structure.  Soil Sci. 61: 247-256.  52.  Poomia, S.R., and R. Pal. 1979. The effect of organic manuring and water quality on water transmission parameters and sodification of a sandy loam soil. Agric. Water Management 2: 163-175.  53.  Reeve, R . C . 1953. A method for determining the stability of soil structure based upon air and water permeability measurements. Soil S c i . . Soc. A m . Proc. 17: 324-329.  54.  Richards, L . A . 1953. Modulus of rupture as an index of crusting of soil. Soil Sci. Soc. A m . Proc. 17: 321-323.  55.  Roberts, F . J . , and B . A . Carbon. 1972. Water repellence in sandy soils in south west Australia: 2. Chemical characteristics of the hydrophobic skins. Aust. J . Soil Res. 10: 35-42.  56.  Romanov, V . V . 1961. "Hydrophysics of Bogs". Israel program for scientific translations, Jerusalem, 1968.  57.  Russel, M . B . and C L . Feng. 1947. Characterization of the stability of soil aggregates. Soil Sci. 63: 229-304.  58.  S A S Institute Inc., 1982. " S A S User's Guide: Statistics 1982 Edition".  59.  Scheidegger, A . E . 1974. "The Physics of Flow through Porous Media", 3rd. Edition. University of Tronto Press, P P 96-98.  60.  Schnitzer, M . and S.U. Khan. 1972. "Humus Substances in the Environment". Marcel Dekker Inc., New York.  61.  Schothorst, C J . 1977. Subsidence of low moor peat soils in the Netherlands. Goederma 17: 265-291.  62.  Sen, B . C . 1961. Bacterial decomposition of humic acids in clay humus mixtures. J . Indian Chem. Soc. 38: 737-740.  63.  Shainberg, I. and M . J . Singer. 1986. Suspension concentration effects on depositional crusts and soil hydraulic conductivity. Soil Sci. Soc. A m . J . 50: 1537-1540.  64.  Sharma, M . L . and G. Uehara. 1968. Influence of soil structure on water relations in low humic latosols. II. Water movement. Soil Sci. Soc. A m . Proc. 32: 770-774.  65.  Siegel, S. 1956. "Non-Parametric Statistics For The Behavioral Sciences". Mc Graw Hill.  66.  Sposito, G. 1984. "The Surface Chemistry of Soils". Oxford University Press Inc., New York. P P 143-147.  67.  Swartzendruber, D . , M . F . De Boodt, and D . Kirkham. 1954. Capillary intake rate of water and soil structure. Soil Sci. Soc. A m . Proc. 18: 1-6.  68.  Swartzendruber, D. and R. L . Uebler. 1982. Flow of kaolinite and sewage suspensions in sand and sand-silt media. II. Hydraulic conductivity reduction. Soil Sci. Soc. A m . J . 46: 912-916.  western  118 69.  Taylor, H . M . and D . W . Henderson. 1959. Some effects of organic additives on compressibility of Yolo silt loam soil. Soil Sci. 88: 101-106.  70.  Theng, B . K . G . 1979. "Formation and Properties of Clay Polymer Complexes". Elsevier Scientific Publication Co. New York. P P 243-273; 283-326.  71.  V a n Doren, D . M . and R.R. Allmaras. 1978. Effect of residue management on the soil physical environment, microclimate and plant growth. In: "Crop Residue Management System". A m . Soc. Agron. Special Publication. P P 49-83.  72.  V a n Olphen, H . 1963. " A n introduction to clay colloidal chemistry". Interscience Publishers, a division of John Wiley and Sons, New York, pp 4-5.  73.  Walmsley, M . E . and L . M . Lavkulich. 1975. Chemical, physical and land-use investigations of organic terrain. Can. J . Soil Sci. 55 : 331-342.  74.  Williams, B . G . , D . J . Greenland, G.R. Lindstrom and J . P . Quirk. 1966. Techniques for the determination of the stability of soil aggregates. Soil Sci. 101: 157-163.  75.  Woodburn, R. and J . Kozachyn. 1956. A study of relative erodibility of a group of Mississippi gully soils. Trans. A m . Geophysics Union. 37 : 749-753.  76.  Yefremova, T.T. 1984. Macrostructure and humus characteristics of forest peat soils. Soviet Soil Sci. (Translated). 16 : 94-102.  77.  Yoder, R . E . 1926. A direct method of aggregate analysis of soils and a study of the physical nature of erosion. J . A m . Soc. Agron. 28 : 337-350.  78.  Yong, R . N . and B . P . Warkentin. 1975. "Soil Properties and Behaviour". Elsevier Scientific Publishing Co. Amsterdam, The Netherlands. P P 82-85.  79.  Zilva, S., B . Zannetti and A . Re Del. 1982. Relationship between the soil structural stability index and soil organic matter. Agrochimica 26 :167-181.  APPENDIX NLE) . IMPLICIT  REAL  DIMENSION  T(10)  EXTERNAL LOGICAL  *8  1.  Zero2  (A-H,  (A S u b r o u t i n e  ,Y(1).S(1)  F N 1.FN2  LZ TK2.TKZ.TK1,TL.B.H1MH2,DUMT  DATA  T  /  40.DO,110.DO,200.DO.700.DO,1200.D0.1700.DO.2400.DO,  4000.DO.4600.DO,22000.DO/ T K 2 = 2.5D-7 T K Z = 1.3D-7 T K l = 2.lD-5 T L = 1.5D-2 B = 4.D-2 H 1 M H 2 = 4.5D-1 X = 0.D0 Y(1) = 0.D0 1  1  DUMT  = 1,10  =  Td)  Z = 0.D0 ZMX  =  ERR  =  0.10D0 l.D-8  CALL  ZER02  QS  TKZ H1MH2/(Z + TL TKZ/TK1)  =  U.B.C.  0-Z)  COMMON  DO  from  (Z. Z M X , F N l . E R R X Z )  : :  5  D U M = 2.DO B * T K 1 * * 2 * H 1 M H 2 * D U M T 7 ( T L * * 2 *TKZ) i:  EPS=l.D-4 H = fDUMT-X)/4.D0 S(1) = 0.D0  120 3 CALL  DDIFSY(1,X,DUMT,Y,H,EPS,S,FN2,&4)  I F ( X . L T . D U M T ) GO TO 1 WRITE  (1,2)  T(I),  3  Z, QS , Y d ) ,  2 FORMAT(T,Z,QS,  Y(1):',5G15.5)  STOP 4 WRITE(1,6) D U M T 6 FORMATC NO  CONVERGENCE',Gl5.5)  STOP END FUNCTION  FN1(Z)  IMPLICIT  REAL 8iA-H,0-Z)  COMMON  TK2,TKZ,TK1,TL,B.H1MH2.DUMT  ::  F N 1 -- T K 2 " 2 D U M T - T K 2 Z i  DUM  =  B TKZ H1MH2 :  : ;  +  TL ' T K Z T K 2 T K 1 1  F N 1 = F N 1 + B* T K Z * H 1 M H 2 "DLOG((TK2 " Z + DUM)/UUM.! RETURN END SUBROUTINE IMPLICIT  FN2(X,Y,F)  REAL*8  (A-H.O-Z)  COMMON TK2.TKZ,TKl,TL,B,HlMH2,DUMT DIMENSION  Y(1),F(1)  QS = TKZ*H1MH2/(Y(1) + T L * T K Z / T K 1 ) F(1) = B*QS + T K 2 RETURN END  121  APPENDIX 2. F i e l d (6q ) and a i r d r y (Bq ) water c o n t e n t , b u l k d e n s i t y o f a g g r e g a t e beds ( p ) , p a r t i c l e d e n s i t y (p ) and s o i l u  Samp.  #  % (kg H 0 / 2  kg soil) (field)  1 9  o 4 5 6 7 8 9 10 11 12 13 14 15  16 17 18 19 20 21 22 23 24 25 26 27  0.37 0.35 0.32 0.43 0.39 0.36 0.45 0.41 0.44 0.48 0.49 0.58 0.51 0.53 0.51 0.63 0.56 0.58 1.00 0.96 0.71 1.18 1.03 0.98 0.83 0.90 0.S3  »*D  P  (kg H 0 / kg soil.) (air dry.) 2  0.042 0.044 0.042 0.055 0.047 0.042 0.070 0.070 0.078 0.087 0.133 0.133 0.102 0.146 0.117 0.143 0.127 0.102 0.170 0.153 0.153 0.187 0.194 0.201 0.205  0.156 0.163  pH  P s  b  (kg/m )  (kg/m.)  (1:4.)  353.3± 15.3 846.7 ±30.6 890.0±26.5 S16.7±32.1 S80.0±10.0 906.7125.2 723.3111.6 796.7158.6 770.0110.0 730.0117.3 6S3.3125.2 633.3111.6 693.3115.3 686.7123.1 693.3115.3 616.7125.2 703.3120.S 690.0117.3 526.7115.3 560.0117.3 600.0110.0 516.715.8 553.3115.3 540.0169.3 540.0110.0  L842 18.5 1,839115.6 1.852114.8 1,70S± 11.3 1.822110.6 1.S06 ±31.8 1,65017.8 1.710154.4 1.64S130.4 1.506144.6 1,462111.3 1.439141.0 1,49015.0 1.41417.8 1.34414.2 1,336117.0 1.376110.6 1,417126.9 1,164114.8 1.228119.1 1,30114.2 1.17417.8 1.168124.8 1,1S2±22.6 1,156117.0 1,278112.0 1,250112.0  5.28 5.26 5.44 5.14 5.42 5.44 5.18 5.36 5.14 4.50 5.06 4.80 4.90 4.95 4.S5 4.76 5.02 4.88  3  526.7111.5  540.0117.3  3  5.00  4.SO 4.S6 5.12 4.92 4.90 5.02 5.04 4.94  APPENDIX 3. The a c t u a l stability factor.  Sample  E 1 / 2  /day]  (#.t  [(m s)  1 2  52.44±7.03 95.61±11.90 59.32±3.88 55.27±4.39 67.87±18.67 53.4118.90 86.61119.42 58.03110.98 129.61191.29 86.50118.74 100.1117.53 70.27120.96 59.74+4.60 92.65 + 46.94 82.7817.25 86.42123.86 75.92 + 30.36 86.43121.00 1SS.47122.77 154.98110.93 152.19 + 9.41 295.53 + 127.78 268.15 + 75.43 210.19148.64 202.08132.66 223.34120.85 122.43136.07  4  5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27  soil  E a  O  and e x p e c t e d  x [tm sj  1 / 2  /day]  67.0125.81 91.13128.91 84.21+10.96 44.0+1.68 75.94116.45 82.5810.98 86.05+33.88 80.49126.61 132.38 + 42.53 75.66120.93 116.69 + 9.84 119.17 + 6.90 55.5819.35 100.S3 + 67.1S 85.40+22.40 64.52115.09 73.88+28.68 108.22110.96 113.70 + 23.54 140.S3! 13.54 105.09+11.78 251.96136.S5 245.121119.53 133.31137.24 186.82117.75 157.52120.23 89.1019.02  123  APPENDIX 4. A g g r e g a t e s t a b i l i t y , o r g a n i c m a t t e r c o n t e n t , m i n e r a l m a t t e r c o n t e n t , and HCl d i g e s t i b l e a s h .  m  Org. matt. (%)  Min. Matt.  ( i)  (%)  67.03±1.30 68.80±6.84 65.6312.06 71.07±6.46 65.2316.44 63.87113.60 84.23 + 0.55 66.4018.36 69.S316.3 81.4312.08 76.8711.70 83.2710.85 82.3712.S4 82.S017.86 80.5310.42 8S.3714.58 S2.0715.76 83.2714.04 94.7011.04 92.8010.17 86.7311.53 96.5010.17 96.8010.92 95.9010.46 94.7011.05 92.4010.96 92.73+1.04  14.5010.87 14.8311.15 12.8310.29 20.8310.76 17.6710.76 16.1710.76 26.3310.29 20.6711.04 25.5011.0 32.8310.58 32.8310.58 44.1711.04 38.5011.0 40.010.50 39.3310.58 52.6710.76 39.1711.04 42.1710.29 69.010.0 64.17 + 0.76 51.1711.26 72.8310.58 65.8310.29 62.8310.76 59.0+3.28 60.1710.58 60.8310.58  76.6711.04 77.1710.76 79.8310.73 71.1710.76 74.3311.26 76.3312.36 64.6710.29 71.3311.04 65.6711.44 57.8310.29 5S.010.50 46.3310.29 51.5010.87 49.3310.58 50.6710.29 36.5010.0 50.5010.50 48.010.50 21.5010.50 25.50+0.S7 39.1710.58 17.1711.44 23.1710.76 27.5010.50 29.1711.44 2S. 8311.44 29.1711.26  S.8310.76 8.010.50 7.3311.04 8.010.0 8.010.50 7.5011.73 9.010.0 8.010.50 8.83 + 0.76 9.3311.26 9.1711.04 9.5011.32 10.0+0.50 10.6710.76 9.8310.76 10.S310.58 10.3310.58 9.8310.58 9.5010.50 10.3311.04 9.6710.76 10.010.50 11.010.S7 9.6710.29 11.S312.31 11.010.87 10.011.32  Samp. Agg. Stab. (#)  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27  c  HCl Dig. • Ash  124  APPENDIX 5. Mean s a t u r a t e d hydraulic c o n d u c t i v i t y (K )and s t a n d a r d d e v i a t i o n (Sd) o f t h e a g g r e g a t e beds and e f f e c t i v e saturated hydraulic conductivi < J after dispersion. K  f  Sample  #  Mean K (m/s!  o  o  4 5 6 7 8 9 10 1] 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27  Sd of K (m/s)  s  Mean K „. eff (m/s)  Sd of (m/s)  (IO" )  (10- )  '(10" )  (IO'  9.29 13.10 7.54 19.50 7.85 6.71 21.90 7.97 14.70 18.70 19.00 27.70 30.30 11.70 33.80 50.50 24.90 23.40 83.00 51.80 29.50 68.90 97.20 174.00 104.00 140.00 159.00  4.08 2.26 0.35 5.90 1.80 0.96 6.50 1.82 0.36 4.0 5.40 10.20 4.20 1.40 8.10 1.70 3.20 2.70 2.30 24.20 2.80 18.00 37.20 34.00 10.20 21.00 12.00  2.70 3.40 3.40 5.20 3.90 3.20 3.50 3.70 4.55 3.60 8.60 5.80 2.20 5.20 6.40 3.20 5.60 6.50 13.40 11.00 12.60 21.50 12.00 17.50 14.00 18.50 16.60  1.20 2.10 1.10 2.20 0.50 0.60 ' 2.80 2.0 2.62 0.50 2.70 3.20 0.40 5.00 2.40 0.70 0.40 1.30 0.40 2.10 0.20 6.10 10.00 17.50 3.80 4.90 1.80  6  1 2  E  6  6  6  

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.831.1-0097102/manifest

Comment

Related Items