Soil Physics: Understanding Water Movement, Hydraulic Conductivity, and Soil Structure, Manuais, Projetos, Pesquisas de Agronomia

Baixe Soil Physics: Understanding Water Movement, Hydraulic Conductivity, and Soil Structure e outras Manuais, Projetos, Pesquisas em PDF para Agronomia, somente na Docsity! 15 CHAPTER 2 SOIL PHYSICAL PROCESSES The main soil physical processes influencing soil formation are movement of water plus dissolved substances (solutes) and suspended particles, temperature gradients and fluctuations, and shrinkage and swelling. Here, a short outline of such processes will be given. For more details on the principles involved in the movement of water through soils, reference is made to introductory soil physics texts such as given in 2.7. 2.1. Movement of water To understand the behaviour of water in a porous medium such as a soil, two characteristics are of paramount importance: (1) the relationship between the pressure potential and the volumetric moisture content, and (2) the relationship between the pressure potential and the hydraulic conductivity. These are shown in Figs 2.1 and 2.2 for soil materials of different textures. Figure 2.1. pF curves of various soil materials (after Bouma, 1977). Figure 2.2. Hydraulic conductivity in relation to pressure potential (in cm) for various soil materials (after Bouma, 1977). 16 The force with which water is retained in the soil is expressed either by the pressure potential h, which is most conveniently expressed in the height of a water column (cm) relative to the groundwater level. In unsaturated soils above the groundwater level, the pressure potential h is negative, as a result of the capillary action caused by simultaneous adhesion between water and soil particles and cohesion between water molecules. In the unsaturated zone, the pressure potential is often indicated by the logarithm of its positive value, the pF value (pF = -log h; h in cm). So a pressure potential of -100 cm is equivalent to pF = 2. Question 2.1. Change the values on the X-axis of Figure 2.2 into pF values. Below which moisture content is the hydraulic conductivity less than 1 cm/day for a) sand, and b) clay? (Compare Figures 2.1 and 2.2). The volume of water that can be conducted per unit time by an individual tubular pore at a given hydraulic gradient increases with the 4th power of the pore radius. Therefore, soils with coarse interstitial pores, such as sands, have a much higher saturated (h=0) hydraulic conductivity than (structureless) finer-textured soils which have narrower interstitial pores. As shown in Figure 2.2, K decreases with decreasing water content () or pressure potential (h). The decrease in K with decreasing h is very steep in coarse-textured soils. Therefore, in coarse textured soils, water movement is particularly slow in non-saturated conditions. Question 2.2. Why does the hydraulic conductivity decrease with the water content of the soil? Question 2.3. Why is unsaturated water movement in coarse-textured soils particularly slow? Consider pore-sizes and pore geometry. DIRECTION OF TRANSPORT If the hydraulic head decreases with depth (e.g. if the soil surface is wetter than the subsoil), DARCY'S LAW The generally slow, vertical laminar flow of water in soils can be described by Darcy's law: q = -K . grad H = -K . (h + z) / z , in which: q is the flux density (volume of water conducted through a cross-sectional surface area of soil per unit time (m3.m-2.s-1 or m.s-1), K is the hydraulic conductivity (m.s-1), grad H (m/m) is the gradient in hydraulic head, H, in the direction of the flow, and z the height above a reference level. The hydraulic head is composed of the gravitational potential (numerically equal to z) and the pressure potential h. K depends on number and sizes of pores that conduct water. 19 Figure 2.4. The effect of water repellence. Water drops have difficulty penetrating the water-repellent soil (right). From Dekker and Jungerius, 1990. Infiltrating rainwater is low in solutes, and it is under-saturated with respect to many minerals. As a result, minerals dissolve (weather), and solutes percolate downward. Mineral weathering is discussed in more detail in Chapter 3. Some solutes do not react with the solid phase of the soil and move unhindered through the soil with the percolating water. Many solutes, however, interact with solid soil materials and therefore move more slowly than the water does. The selective retention of solutes in the soil and the resulting change of soil water chemistry with depth is called a ‘chromatographic’ effect. Part of the interaction between solid phase and solute is by ion exchange. This causes a change in the composition of the percolating water (and, at the same time, in the composition of the exchange complex) with depth. The principles involved in ‘ion exchange chromatography’ in soil columns have been well established. The models describing such processes are of particular relevance when exotic or ‘foreign’ solutes are added to the soil with percolating water, as when heavy doses of fertilisers are applied, or in case of pollution. Particularly in climates with a slight to moderate excess of rainfall over (actual) evapotranspiration, removal of water by roots of transpiring plants causes an increase in solute concentration with depth. As a result, super-saturation with respect to certain minerals may be reached at some depth, so that solutes may precipitate to form those minerals (see Chapter 3). 20 Question 2.7. What is meant with the 'ion exchange chromatography' of soil columns? 2.3. Temperature effects Because soils have a large heat capacity and low heat conductivity, temperature fluctuations are strongly buffered, which means that the amplitude of daily and seasonal temperature differences strongly decreases with depth. Temperature has four main effects on soil forming processes: 1) on the activity and diversity of biota; 2) on the speed of chemical reactions, 3) on physical weathering of rocks, and 4) on the distribution of fine and coarse fractions. The first two processes are discussed in Chapter 4 and 3, respectively. Daily temperature variations are strongest in rocks that are exposed at the soil surface in extreme climates. The difference between day and night temperature may be more than 50oC in some arid climates. Temperature differences have a strong influence on physical weathering of rocks. Heating causes mineral grains to expand. Most rocks consist of more than one kind of mineral, and each mineral has its own coefficient of expansion. This will cause fractures at the contact between grains that expand at a different rate. Individual mineral grains are loosened from the rock. In addition, the temperature at the surface and in the interior of the rock is not equal, and the surface layer expands more strongly than the internal part. This causes chipping off of the surface layer (exfoliation). Both processes lead to physical diminution of the parent material, be it a solid rock or a sediment. Frost action has a similar effect. The volume of ice at 0oC is larger than that of water at the same temperature, and because freezing starts at the surface of rocks, structure elements, or mineral grains, freezing water causes cracks to expand and the material to fall apart into smaller units, mainly of sand and silt size. Frost action also results in a redistribution of fine and coarse material. Frost heaval causes stones to gradually move to the surface because ice lenses at their bottom lift them (Plate D, p. 40). Coarse pavements of polar deserts are formed this way. Frost heaval in combination with frost polygons causes coarse material to accumulate in frost (shrinkage) cracks and to form stone polygons (Plate E, p. 40). The finer central parts of such polygons are favoured by the vegetation, which causes further differentiation. 2.4. Shrinkage and swelling of soil aggregates and clays Removal of water from, or its addition to a soil may result in strong changes in volume: shrinkage and swelling. We distinguish three phases of shrinkage processes. A first phase is restricted to irreversible removal of water from sediments that have been deposited under water. This shrinkage and the chemical processes that are associated with it, are usually called ‘soil ripening’. The second and third phases, which are related to cyclical drying and rewetting, occur in different intensity in all soils. SOIL RIPENING Clayey sediments deposited under water are soft and have a high water content. When 21 drainage or evapo-transpiration removes water, such sediments increase in consistency and undergo various chemical changes. These processes together are called soil ripening. This term was coined in analogy to the traditional term 'ripening' used for cheese. In this process, a firm cheese substance is formed by pressing moisture from an originally wet, very soft mass of milk solids, followed by a period of further moisture loss by evaporation. The nearly liquid starting material is called 'unripe'; the much firmer end product is called 'ripe'. Clayey underwater sediments have a high pore volume that is completely water-filled, of the order of 80 percent. This corresponds to about 1.5 g of moisture per g dry mass of sediment. The sediment normally has a very low hydraulic conductivity, less than 1 mm per day under a potential gradient of 10 kPa/m (1 m water head per m). Therefore, the material dries out very slowly by drainage alone. Question 2.8. Check the statement that water-filled pore volume of 80% corresponds to about 1.5 g of water per g of dry sediment. Assume that organic matter (with a particle density of 1 g/cm3) makes up 10% of the volume of all solid matter and that the particle density of the mineral fraction is 2.7 g/cm3. Remember that pore volume + volume organic matter + volume mineral matter = 100%. For such a sample, calculate the dry bulk density (= the mass of dry soil per unit volume of soil in the field) and compare your result with the bulk density of well ripened sediments, which is 1.2 - 1.4 g/cm3. Briefly describe the cause of the difference. Question 2.9. Why do clayey under-water sediments have a very low hydraulic conductivity, in spite of a high total porosity? EFFECT OF FAUNA AND VEGETATION Roots need oxygen to grow and to function. Water-saturated, unripened sediments lack oxygen, and most plants cannot grow on such sediments. Certain plants, such as reed (Phragmites) and alder (Alnus) in temperate climates, and wetland rice (Oryza), or mangrove trees (Rhizophora and others) in the tropics, can supply oxygen to the roots through air tissue (aerenchyma). They can grow on unripened, water-saturated sediments and extract moisture from it. Such plants are sometimes used to reclaim freshly drained lands. Saline or brackish tidal sediments in temperate climates lack trees, but also grasses and herbs contribute strongly to water loss by uptake through roots and by transpiration. Burrowing by soil animals may greatly contribute to vertical permeability in unripened sediments. Tropical tidal flats usually contain many channels formed by crabs. Such sediments have a very high permeability, and even ebb-tide drainage can contribute significantly to the ripening process. The first biopores may be formed while the sediment is still under shallow water, or lies between high and low tide. The mud lobster (Thalassina anomalis), for example, may produce large, mainly vertical tunnels in tidal areas in the tropics. Roots of swamp vegetation may leave, after decomposition, biopores of different diameters. Such wetland pore systems may become fossilised by precipitation of iron oxides along the pore walls, 24 unripened. Firm clay that cannot be squeezed out between the fingers is completely ripened. Several ripening classes, based on the ripening factor n, have been defined between these extremes, as shown in Table 2.2. The ripening factor is the amount of water bound to a unit mass of clay fraction. This amount cannot be measured directly. It can be estimated from the moisture content of a soil sample if the contents of clay and organic matter are known as well, and if assumptions are made about the amounts of water bound to other soil components. The organic matter fraction in a soil binds about three times as much water per unit mass as the clay fraction under the same conditions. The (sand + silt) fraction of a clayey soil binds about 0.2 g water per g (sand + silt) under wet conditions. From these empirical relationships, the following formula for the n value was derived: n = (A - 0.2R)/(L + 3H) (2.1) where n is the mass of water per unit mass of clay + organic matter, A the water content of the soil on a mass basis (%, with respect to dry soil) R the mass fraction of (silt + sand) (% of solid phase) L the mass fraction of clay (% of solid phase) H the mass fraction of organic matter (% of solid phase). The n-value can be used to more precisely quantify ripening. The higher the n-value, the less ripened the soil. The n-value is used in the Soil Taxonomy to distinguish unripe soils (Hydraquents) from other mineral soils. Hydraquents should have an n-value over 1 in all subhorizons at a depth of 20 to 100 cm. The n-value decreases rapidly upon drying, so Hydraquents may quickly change in classification once they are drained. RESIDUAL AND ZERO SHRINKAGE Ripened fine-textured soils may exhibit appreciable shrinkage and swelling upon drying and wetting. This is illustrated in the left-hand part of Figure 2.5. In such soils, we recognise two phases of shrinkage: a phase which includes further reduction of aggregate volume upon water loss (residual shrinkage), and a second phase in which further water loss does not affect the aggregate volume (zero shrinkage). Recent research has shown that normal and residual shrinkage can be appreciable in clay soils (15-60% particles <2 m) in the Netherlands (Bronswijk, 1991). Some clay soils show very strong normal shrinkage, i.e. aggregates shrink considerably without any entrance of air into the aggregates. This implies that large inter-aggregate pores (cracks) may be formed while aggregates remain water-saturated, with water in very fine intra- aggregate pores. Under such conditions, rapid bypass flow through shrinkage cracks can take place. Furthermore, changing void ratio does imply that the architecture of the pore system is not constant, but may change with the water content. Soils differ in swelling behaviour, depending on the nature of the solid phase and the ionic composition of the ion exchange complex and the soil solution. 2:1 clay minerals with a low charge deficit (see Chapter 3.2) have a high potential for swelling. Swelling increases with the relative amount of Na+ ions on the adsorption complex and with decreasing electrolyte level of the interstitial water. 25 Table 2.2. Field characteristics of ripening classes (Pons and Zonneveld, 1965). Field characteristics Ripening class limiting n-value very soft1, runs between the fingers without squeezing completely unripened 2.0 1.4 1.0 0.7 soft, is easily squeezed out between fingers practically unripened moderately soft, runs between fingers when squeezed firmly half ripened moderately firm, can just be pushed out between fingers when squeezed firmly almost ripened firm, cannot be squeezed out between fingers ripened 1 soft and firm in the sense of soft mud and firm, wet clay; not in the sense of the definitions for dry or moist material as in the Soil Survey Manual (Soil Survey Staff, 1951 or later editions. As shown by Tessier (1984) and Wilding and Tessier (1988) (Figures 2.6 and 2.7), individual (TOT) plates of smectite are stacked together face to face to form clay particles, or "quasi crystals", which comprise between 5 and 10 (Na-saturated) to more than 50 (Ca- saturated) plates. Under wet conditions (h= -10 cm) with low electrolyte levels, Na- saturated smectites have a swollen diffuse double layer (the space between individual plates containing exchangeable cations), up to 10 nm thick. Under the same conditions, Ca- smectite plates have a spacing of 1.86 nm (see also Table 2.1). The clay particles are arranged in a honeycomb structure (Fig 2.6, 2.7), with water-filled pores of up to 1 m wide. This structure is flexible as an accordion. As the wet clay dries out, water is lost from the bellows and the accordion closes. Upon rehydration, water re-enters the bellows. The ease with which the bellows open and close depends on the clay mineral and the physico- chemical conditions. In a Ca-smectite, the opening and closing of the bellows cause most of the shrinkage and swelling. In a Na-smectite the uptake and removal of interlayer water is important as well. Flexibility of the clay particles is highest in Na-smectites, lower in Ca-smectites, and still lower in the brittle, rigid clay particles ("domains") of illite and in the coarse particles ("crystallites") of kaolinites. This explains the decreasing shrinkage and swelling potential in that order. (More information on smectites, illites, kaolinites, and other clay minerals is given in Chapter 3.2). The data in Fig 2.6 refer to pure clay-water mixtures. In actual soils the swelling is much less because of the presence of physically inert coarser minerals, increased cohesion of mineral particles due to binding to various organic and inorganic substances and, at least at some depth, overburden pressure. 26 Figure 2.6. Schematic representation of the microstructure of smectite saturated with (A) NaCl, or (B) CaCl2 at 10-3 M chloride concentration. From Wilding and Tessier, 1988. Fgure 2.7. Microstructure of a smectite at 1M NaCl; water saturated. From Wilding and Tessier, 1988. 29 Problem 2.2 Table 2B gives calculated amounts of water passing the soil (in mm/yr) at 0, 10, 20 and 30 cm depth below the land surface for (1) three soil materials with different volumetric soil moisture content (%), (2) three levels of total annual precipitation, and (3) two regimes of rainfall intensity. Annual precipitation values with monthly distributions and monthly potential evapo-transpiration (PET) typical for De Bilt (Netherlands), Rabat (Marocco) and Kintapo (Uganda) were used in the calculations (taken from Feijtel and Meyer, 1990). Annual total PET values are 626 mm at De Bilt, 818 mm at Rabat and 1521 mm at Kintapo. In the low rainfall intensity regime, the monthly rainfall was evenly distributed; in the high intensity regime, every month had only three days with rain, providing 20 % of the monthly rainfall on day 2, 30 % on day 14 and 50 % on day 26 of each month. Water uptake by roots is assumed to be evenly distributed over a 30-cm deep root zone. Questions a. What is the total annual precipitation for each of the three locations? b. Draw graphs (depth on the y-axis, percolating water on the x-axis) for any three of the different climate-soil combinations to illustrate the effects of climate and soil water holding capacity on the amount of water percolating at various depths. c. Briefly explain the reasons for the effects of rainfall intensity and soil moisture holding capacity for the amount of water percolating at various depths. d. Provided percolating water would always completely displace previously existing soil solution, which rainfall regime would be most favourable for leaching of solutes, followed by chemical precipitation of translocated solutes at a certain depth? Which would be more conducive to leaching of the whole soil profile? e. How does your answer to question d have to be modified for a strongly pedal soil? Problem 2.3 Criticise the following statement: In dry climates, where potential evapo-transpiration exceeds annual precipitation, leaching of solutes to below the root zone does not take place. Problem 2.4 Figures 2C and 2D (from Bronswijk, 1991) show shrinkage characteristics for two subsoils of heavy clay soils from the Netherlands. The vertical axis gives the void ratio. In which soil do aggregates shrink strongly during drying, giving rise to formation of cracks between the aggregates, and allowing bypass flow? 30 Table 2B. Amounts of water passing through the soil (in mm/yr) at 0, 10, 20 and 30 cm depth below the land surface for (1) three different soil materials (as indicated by differences in available soil moisture, volumetric %), (2) three levels of total annual precipitation (De Bilt, Rabat and Kintapo), and (3) two regimes of rainfall intensity. location rainfall intensity available soil moisture (%) water (mm/yr) passing at depth (cm) actual evapo-tran- spiration (mm) 0 10 20 30 de Bilt low 25.0 765 452 296 243 522 10.0 765 452 309 261 504 1.5 765 452 318 270 495 high 25.0 765 698 658 630 135 10.0 765 702 665 640 125 1.5 765 702 665 640 125 Rabat low 25.0 497 290 200 147 350 10.0 497 302 218 176 320 1.5 497 310 225 189 307 high 25.0 497 410 389 370 127 10.0 497 440 413 398 99 1.5 497 448 424 409 87 Kintampo low 25.0 1517 856 566 392 1125 10.0 1517 868 581 407 1110 1.5 1517 877 589 416 1101 high 25.0 1517 1296 1183 1111 406 10.0 1517 1317 1213 1147 370 1.5 1517 1326 1218 1151 366 31 Figures 2C and 2D. Shrinkage characteristics of the C11g of profile Bruchem (left) and of the C22g horizon of profile Schermerhorn (right). Problem 2.5 Figures 2E and 2F refer to a soil reclaimed from the IJsselmeerpolders (The Netherlands). Fig. 2E indicates the elevations of the soil surface and of reference plates, that had been installed in the soil at 40, 80, 120 and 200 cm depth just after reclamation, as a function of time. Fig. 2F shows changes with time of the volume of soil cracks in mm (= liter/m2) in the same soil. In both figures, the lines were calculated by computer simulation of the ripening process, while x indicates a measured value. Only use calculated values, to be measured from the figures, in answering the questions. Assume that no changes took place below an original depth of 150 cm. a. Estimate the subsidence (in cm) that took place between 1968 and 1979 in the layers with the original depths 0-40 cm, 40-80 cm, 80-125 cm and 125-200 cm. b. How much water has been removed from the soil (in mm) between 0 and 150 cm depth over the period 1968-1979? Take crack volume into account. Assume that no water loss took place below the depth indicated by the 50 cm dotted line (150 cm below original soil surface). c. Explain (1) the stepwise character of the lines, and (2) the presence of "sills", small peaks, on the steps in Fig. 2F. 34 Figure 2H. Soil shrinkage characteristics at four depths in a recent IJsselmeerpolder. The values printed in the graphs refer to the pressure potential (indicate by short verticals at different values of moisture ratio), expressed in cm (equivalent height of water column). From Bronswijk, 1991. 35 Table 2J. Characteristics of the solid soil material of the Dronten profile. From Bronswijk, 1991. depth (cm) clay org. matter particle density (g/cm3) mass fraction (%) 0-22 37 9.9 2.66 22-42 46 8.1 2.66 42-78 35 6.6 2.63 78-120 16 5.8 2.59 2.6. Answers Question 2.1 A pressure potential of -10 cm equals a pF of 1; a potential of -100 cm a pF of 2. For sand, the hydraulic conductivity is less than 1 cm/day at a pressure potential of -50 (pF = 1.7) and for clay this value is -10 (pF = 1.0). In Figure 1.1, the water contents at these pressures are approximately 35% for sand, and 50% for clay. Question 2.2 The hydraulic conductivity decreases with decreasing pore size. At lower water contents, the larger pores are filled with air. Question 2.3 Coarse-textured soils have few fine pores and these are restricted to contacts between sand grains and therefore not continuous. Question 2.4 A 'complex pore system' refers to all three characteristics. Question 2.5 Bypass flow causes the strongest dispersion, because there is hardly any interaction with water in small pores. Question 2.6 Wetting front instability does not occur when a coarse layer overlays a fine one, because the finer pores in the underlying layer cause rapid transport into that layer. Question 2.7 ‘Ion exchange chromatography' refers to the fact that different cations in solution are retained to a different degree by the adsorption complex, causing differences in the speed by 36 which they move through the soil. The same principle is used in chemical analysis by chromatography. Question 2.8 If 20 cm3 of a volume of 100 cm3 consist of solid matter, this solid matter consists of 2 cm3 of organic matter and 18 cm3 of mineral matter. Together, this weighs 18*2.7 + 2*1 g = 50.6 g. The weight of 80 cm3 water is 80 g. The water/solid weight ratio is therefore about 1.58. The difference with the factor of 1.5 lies in the assumptions. The dry bulk density of such a soil equals 0.506. This is much lower than that of a ripened soil, because there is very much water between the particles, and shrinkage would increase the bulk density. For (dry) bulk density, the dry weight is divided by the wet volume. Question 2.9 Clayey under-water sediments have a very low hydraulic conductivity because all pores are inter-particle pores, which are very small. Problem 2.1 a/b. Situations B and C depict saturated flow (also macropores filled with water). These must refer to situations with water standing on the soil. A moist soil has open voids, through which water can percolate rapidly, so C matches with 2. Irrigation water on wet soil causes slower replacement of water already present in large pores, so figure B matches with situation 4. A short rainstorm causes some percolation along major channels, while a slow drizzle does not. Therefore, figure D belongs to situation 1 and figure E to situation 3. c/d. Rapid breakthrough of a tracer indicates bypass flow. The most rapid flow occurs in figure C, which should match situation ii; the next is figure B, which should match with situation i. For situations D and E, breakthrough occurs earlier in D (between the third and the fourth row, while breakthrough in E is clearly in the fourth row. Complete displacement will be faster in E. Therefore, figure D belongs to situation iv and figure E to situation iii. Problem 2.2 a. The total annual precipitation is the amount that passes through depth 0. b/c.The graph on the next page is drawn for the data from De Bilt. They illustrate the speed with which water is withdrawn from the soil and the effect of rainfall intensity on wetting depth. The amount of water passing through a specific layer increases with decreasing moisture-holding capacity of that layer. Of intense rainfall, a smaller percentage disappears through evapotranspiration. d. Leaching of the topsoil in combination with accumulation at some depth means that large amounts of water should be transported through the topsoil, and much smaller amounts through the deeper layers. This is the case at low rainfall intensity and high available soil moisture. Maximum removal of solutes occurs with high rainfall intensity and low available soil moisture. e. In a strongly pedal soil, most water at high rainfall intensity would be removed through major channels, without removing solutes from the bulk soil (bypass flow). In such 39 van Olphen and F.A. Mumpton (eds) Proc. Int. Clay Conf. Denver, The Clay Mineral Society, Bloomington, IN. Bouma, J., 1977. Soil survey and the study of water in unsaturated soil. Soil Survey Papers No 13, Soil Survey Inst., Wageningen, The Netherlands, 107 pp. Bronswijk, J.J.B., 1991. Magnitude, modeling and significance of swelling and shrinkage processes in clay soils. PhD Thesis, Wageningen Agricultural University, 145 pp. Dekker, L.W., and P.D.Jungerius, 1990. Water repellency in the dunes with special reference to the Netherlands. Catena supplement 18, p173-183. Catena Verlag, Cremlingen, Germany Driessen, P.M. and R. Dudal (eds.), 1991. The major soils of the World. Agricultural University, Wageningen, and Katholieke Universiteit Leuven, 310 pp. FAO, 1989. Soil Map of the World at scale 1:5,000,000. Legend. FAO, Rome. Feijtel, T.C.J. and E.L. Meyer, 1990. Simulation of soil forming processes 2nd ed., Dept of Soil Science and Geology, WAU, Wageningen, The Netherlands, 74 pp. Hillel, D., 1980 Applications of soil physics. Academic Press, 385 pp. Pons, L.J. and I.S. Zonneveld, 1965. Soil ripening and soil classification. ILRI publ. 13, Veenman, Wageningen, 128 pp. Reinierce, K. 1983. Een model voor de simulatie van het fysische rijpingsproces van gronden in de IJsselmeerpolders. (in Dutch). Van Zee tot Land no. 52, 156 pp. Ritsema, C.J., and L.W. Dekker, 1994. Soil moisture and dry bulk density patterns in bare dune sands. Journal of Hydrology 154:107-131. Smits, H., A.J. Zuur, D.A. van Schreven & W.A. Bosma. 1962. De fysische, chemische en microbiologische rijping der gronden in de IJsselmeerpolders. (in Dutch). Van Zee tot Land no. 32. Tjeenk Willink, Zwolle, 110 p. Tessier, D., 1984. Etude experimentale de l'organisation des materiaux argileux. Dr Science Thesis. Univ. de Paris INRA, Versailles Publ,. 360 pp. United States Soil Conservation Service. 1975. Soil Taxonomy; a basic system of soil classification for making and interpreting soil surveys. Agric.Handbook no. 436. USA Govt. Print Off., Washington. Wilding, L.P. and D. Tessier, 1988. Genesis of vertisols: shrink-swell phenomena. p.55- 81 in: L.P. Wilding and R. Puentes (eds) Vertisols: their distribution, properties, classification and management. Technical Monograph no 18, Texas A&M University Printing Center, College Station TX USA. General reading: Hillel, D., 1980. Fundamentals of soil physics. Academic Press, New York, 413 pp. Hillel, D., 1980. Applications of soil physics. Academic Press, New York, 385 pp. Koorevaar, P., G. Menelik and C. Dirksen, 1983. Elements of soil physics. Elsevier, Amsterdam 40 Plate D. Stone cover due to frost heaving, Iceland. Photograph P. Buurman. Plate E. Frost polygons with clear sorting of coarse material along the cracks, Iceland. Note also hummocky vegetation. Photograph P. Buurman