|
|
| Bildiri Özetleri | |
| Ana Sayfaya Dönüş |
| ISD Ana Sayfası |
|
A.A. Jafarzadeh 1 , J.A. Zinck 2 1 University of Tabriz, Faculty of Agriculture, Soil Science Department, Tabriz, I.R. Iran 2 ITC, PO Box 6, 7500 AA Enschede, the Netherlands ABSTRACT A literature review revealed important knowledge gaps on the relation between mechanical resistance of gypsum crusts and root elongation. There is no known research on the penetration resistance of gypsum crusts. To discriminate between factors controlling the strength of gypseous surface crusts, a soil column experiment, with four gypsum-saturated water table heights, four textures and three chemical treatments, was conducted for 120 days. Sandy loam accumulated the highest amount of gypsum at any water table height, followed by medium-fine sand and silt loam. Acid-washed medium sand was comparatively less efficient in conducting and concentrating gypsum in the surface layer. Penetration resistance (PR) substantially increased upon drying of the crust. In fine materials, PR decreased as gypsum content increased because of granular crystallization and structure improvement. In coarse materials, gypsum caused cementation and reinforced the strength of the crust. In general, PR decreased with the addition of CaCO3 and increased with the addition of NaCl. PR values of the dry soil materials are significantly correlated with texture, sodium chloride and the strength of the moist soil materials, in this order. The factors contributing to the total variation of dry PR values are : texture (38%), NaCl (28%), water table height (15%), gypsum (9%) and CaCO3 (5%). Measurements of PR with a 1 mm probe are highly aleatory as they are controlled by short-distance variations of the soil microfabric. Seedling emergence and root elongation may be severely hindered by gypsum crusting. INTRODUCTION Gypsum crusts have been reported from many (semi-) arid regions. Their geographic distribution closely coincides with the areas receiving less than 250 mm rainfall per year (Watson, 1982). Low temperature limits their development and this is why they appear to be rare in the cold deserts (Watson, 1979). Extensive gypsum crusts have been described in Middle East countries, where they are a major limitation for crop production because water infiltration rate, seedling emergence and crop growth are largely controlled by the thickness and gypsum content of the crust (Chartres et al., 1985; Nafie, 1989). The severity of the mechanical hindrance that gypsum crusts oppose to crop development can be assessed by measuring penetration resistance. The effect of moisture and gypsum content on the penetration resistance of gypsiferous horizons has been studied by Poch and Verplancke (1997), but gypsum crusts developing on the soil surface have not been given yet the same attention. The objective of the present paper is to investigate the effect of selected soil properties, including texture, moisture, chemical additive and water table level, on the penetration resistance of gypsum crusts. MATERIALS and METHODS To examine the penetration resistance (PR) of gypsum crusts, column experiments with different water table heights, soil textures and chemical treatments, were conducted (Jafarzadeh, 1991). Four natural soil materials, originally free of gypsum, were used in the main experiment, including silt loam, sandy loam, medium-fine sand and acid-washed medium sand. Very fine and very coarse materials were avoided because of their negative effect on water movement. The materials represent four textural conditions sufficiently differentiated to significantly influence the ascending water movement from a constant, gypsum-saturated water table at variable depths. The textural differentiation is reinforced by differences in porosity, clearly separating the loamy materials (50% porosity) from the sandy materials (40% porosity). The soil materials were air-dried, sieved at 2 mm, and then packed into columns. Polyethylene tubes 6 cm in diameter and 7.5 cm, 12.5 cm, 25 cm, and 50 cm in height, were used in the main experiment on surface crust formation. Additionally, columns 1 m high and 9 cm in diameter were used with two textures (silt loam and medium-fine sand) to determine the maximum height of the water table ascent. All samples were placed in gypsum-saturated water, prepared from Paris plaster and distilled water, in saucers (samples with 7.5 cm height) and beakers (samples with 12.5 cm, 25 cm, 50 cm height). The core experiment was conducted on natural soils submitted to gypsum enrichment from the ascending water table, but without other chemical additives. In a parallel experiment, the natural soil materials were artificially enriched with calcium sulfate, calcium carbonate and sodium chloride, respectively, to study the effect of salts usually present in gypseous soils. The sun heat in desert was simulated during the experimental period (120 days) by means of continuously glowing 100-watt radiant lamps placed 20 cm above the soil surfaces. A constant water table level was maintained at the bottom of the soil columns until the experiments were terminated. After 120 days, samples were cut from the top 7.5 cm (0-7.5 cm) of the columns. The pipes were divided by heat except the columns of 7.5 cm height. Samples from the experiments were selected and their mechanical resistance measured using a laboratory penetrometer with a 1 mm probe. A modified version of the apparatus described by Gooderham (1973) was used to let the samples move upwards on the platform of the penetrometer at a speed of 100 mm/hour. All treatment samples of the four different experiments were tested with five replicates, randomly distributed over three samples according to a sequence of 2-1-2. Penetration resistance was measured before and after drying for samples with no chemical additives and only after drying for all other samples (table 1). Samples were exposed to drying for 45 days in an oven at 38-40°C.  RESULTS GYPSUM CONCENTRATION AND CRUST FORMATION During the conduction of the experiment lasting 120 days, gypsum moved upwards in the columns from the constant, gypsum-saturated water table at the bottom. Gypsum ascent was stimulated by the artificial heat applied on top of the columns. After oven-drying, gypsum accumulating in the upper part or on top of the columns formed a crust, with gypsum content varying between 0 and 18%. Sandy loam was the material with the highest amount of gypsum (11 to 18%) at any water table height, followed by medium-fine sand and silt loam. Acid-washed medium sand was comparatively less efficient in conducting and concentrating gypsum. These results show that the balanced particle size distribution of the sandy loam material, with 44% silt + clay and 39% fine and very fine sand, is more favorable to capillary rise than the other textures. High amount of fines in the silt loam creates more retention of the soil solution than upward conduction. A high amount of sand coarser than fine sand breaks the capillary continuity in the acid-washed medium sand. Fines are not necessary to secure the ascent of the soil solution, even up to 50 cm height, if there is enough fine and very fine sand (e.g. 22% in the medium-fine sand material). For all materials, with only a slight deviation in the case of silt loam, largest surface accumulation of gypsum was obtained when the depth to the water table was 12.5 cm. Smaller amounts of gypsum concentrated with water table heights of 7.5 cm and 25 cm. Surficial gypsum accumulation was very low or even nill in the case of silt loam and clean medium sand, when the water table was as deep as 50 cm. Maximum elevation of gypsum accumulation was 40-42 cm for acid-washed medium sand , 46-47 cm for silt loam and 60-70 cm for medium-fine sand (the last two in tubes 1 m high and 9 cm internal diameter). For all four soil materials, the surficial concentration of gypsum decreased from 12.5 cm to 25 cm to 50 cm water table height. This seems to reflect the efficiency of the capillary rise in each case. The lower accumulation of gypsum for the shortest capillary rise (7.5 cm) responds to a different mechanism, where concentration is retarded because of the shallow water table, close to the soil surface, causing pores to be permanently water-filled. The thickness of the crust formed by the surficial gypsum concentration varied between 1 and 3 cm according to the texture of the material. The thickest gypsum crust was found on sandy loam + 13% gypsum, the shallowest on acid-washed medium sand. In columns where 2% NaCl was added to medium-fine sand, upheaval of the material occurred and a crust 3-4 cm thick grew outside the tubes. EFFECT OF SURFACE DRYING ON PENETRATION RESISTANCE Water content has been reported as an important factor controlling penetration resistance of horizons with gypsum (Callebaut et al., 1985; Poch and Verplancke, 1997). But there is no information about the effect of moisture on the penetration resistance of gypsum crusts and the change in crust strength upon drying. As the strength of the surficial layer, enriched in gypsum, is supposed to increase upon drying, PR was measured before and after drying. To evaluate the magnitude of the change and its effect on the strength of the surface layer before and after drying, a simple index relating moist and dry PR values was established (index = [dry PR - moist PR ] / dry PR). In general, the absolute values of dry PR are higher than those of moist PR, with large variations from less than one time to almost eight times higher. Except one case, the index is higher for silt loam and medium-fine sand than for sandy loam and acid-washed medium sand in each column height, respectively. In general, the index values are higher in the 7.5 columns than in the others, but no strong trend is visible. In spite of five measurement replications, some data seem to be aleatory, even erratic. For instance, in three cases moist PR = 0, in three cases moist PR is higher than dry PR and in one case dry PR and moist PR = 0. In the cohesionless acid-washed medium sand, dry PR is lower than moist PR probably because of the effective stress caused by the pressure from the water films between particles. In general, data inconsistency can be explained by strong variations occurring in the soil microfabric at short distance. Small cracks, packing voids and sand grains make PR measurements aleatory when using an instrument of size (1 mm probe) similar to that of the microfabric features. EFFECT OF SOIL TEXTURE ON PENETRATION RESISTANCE As soil texture plays a significant role in the development and stability of soil structure, it can also be expected to influence the susceptibility of soils to crusting. In fine-textured gypseous soils, movement and evaporation of gypsum-saturated water are very slow and accumulation of gypsum at the surface takes a long time to produce a crust. In the experiments conducted, fine materials with low percentage of gypsum cause high penetration resistance. Silt loam texture, with 97% fine particles (<212 um), many well connected water-filled pores of fine calibre which slow down the water movement, and low organic matter content of 0.3 %, shows high penetration resistance in all experiments, even in samples with 50 cm water table height and no gypsum accumulation at the surface (table 1). Sandy loam texture, with 83% fine particles (<212 um), many well connected, medium sized water-filled pores and very few air-filled pores, and 1.2% organic matter, shows lower penetration resistance than silt loam samples in the majority of the experiments. In both cases, a high percentage of gypsum reduces the penetration resistance. In silt loam samples with 7.5 cm height and 7.7% gypsum and in silt loam samples with 50 cm height and no gypsum, penetration resistance is higher than in silt loam samples with 25 cm height and 12.6% gypsum (table 1). Also in sandy loam samples with 7.5 cm height and 14% gypsum, penetration resistance is higher than in sandy loam samples with 12.5 cm height and 18% gypsum and with 25 cm height and 16% gypsum. Thus, in the presence of fine particles (<212 um), penetration resistance decreases with increasing gypsum content, which causes flocculation of clay particles and improvement of the soil structure. In sandy materials, gypsum formed bridges between the skeleton grains, promoting cementation and strengthening the crust. Strongest cementation occurred in medium-fine sand with 22% fine particles, dirty grains carrying thick water films and few obstructing coarse air-filled pores. In acid-washed medium sand with only 7% fine particles, clean grains carrying thin water films and large-size pores which give many barrier air spaces, cementation was weaker. Sometimes, penetration resistance was affected by sand particles during the recording with the penetrometer, as has occurred in medium-fine sand samples with 25 cm height and acid-washed medium sand samples with 7.5 cm height. EFFECT OF GYPSUM CONCENTRATION ON PENETRATION RESISTANCE There is no significant correlation between the amount of gypsum concentrated in the upper part of the columns and PR values (R2 = 0.09, F = 1.41). Neither is there a clear relationship between gypsum percentage, on the one hand, and texture and water table height on the other. The amount of gypsum which has moved upwards from the constant, gypsum-saturated water table does not seem to be enough to create significant strength differences in the crust formed on top of the columns. To better isolate and assess the effect of gypsum on penetration resistance, sandy loam samples were mixed with 13% gypsum and submitted to the same additional treatments as the gypsum-free soil material (four water table heights, drying, + 10% CaCO3, + 2% NaCl). In this experiment, gypsum accounted for 94% of the regression in relation to sandy loam samples without supplementary gypsum. EFFECT OF CHEMICAL ADDITIVES ON PENETRATION RESISTANCE Calcium carbonate 10% and sodium chloride 2% were mixed with all four natural soil materials and with the sandy loam enriched with 13% gypsum to assess their effect on crust formation and penetration resistance, since these chemicals are frequently associated with gypsum in soils of (semi-) arid environments. In general, the addition of CaCO3 decreases the penetration resistance of the gypsum crust, but there is no clear trend in the relationship between PR values and the two supposedly controlling factors, texture and height of the water table. Calcium carbonate could have opposite effects on PR, improving the structure and decreasing the cementation in the loamy soils while favoring laminar crust formation in the sandy soils. Unlike with CaCO3, the PR values in general increase and some decrease only very slightly with the addition of NaCl. This trend is particularly visible in the case of the 7.5 cm high columns for all textures and in the case of the silt loam material in all columns. Sodium chloride, being more mobile than calcium sulfate and carbonate, largely controls the chemistry of the surface crust and enhances its strength in sandy materials, where NaCl crystals occupy the bulk of the crust. In loamy materials, sodium chloride might contribute to dispersion and cause PR values to raise. Additional gypsum, as in the crust of sandy loam + 13% gypsum, tends to decrease penetration resistance. DISCUSSION PR values of the dry soil materials are significantly correlated with texture, sodium chloride and the strength of the moist soil materials, in this order. Texture alone explains 38% of the variation of the PR values, which increases to 61% when texture and NaCl are considered together (table 2). Sodium chloride, for being more mobile than calcium sulfate, has a stronger influence on the strength of the crust than gypsum (R2 = 0.28 for NaCl versus R2 = 0.09 for CaSO4.2H2O). Regression of PR against gypsum improves after logarithmic transformation of the dry PR values (R2 = 0.19). Moist PR is only a loose predictor of dry PR because of the important physical changes which take place in the surficial soil layer upon drying, leading to crystallization and crust formation. The water table height does not significantly influence the dry PR values. PR values of samples enriched with CaCO3 are more (negatively) correlated with the height of the water table than samples enriched with NaCl or CaSO4.2H2O, because of the lower solubility and mobility of calcium carbonate.The strongest relation of gypsum is with texture, indicating that gypsum probably accumulates as granular material rather than cementing the receiving soil material, as already observed by Poch and Verplancke (1997). This causes short-distance irregularities in the crust structure and might explain the low correlation or lack of between some explaining factors and PR values. The factors contributing to the total variation of dry PR values are: texture (38%), NaCl (28%), water table height (15%), gypsum (9%) and CaCO3 (5%). In spite of five replications for each measurement, PR values seem inconsistent in relation to the factors contributing to the formation and strength of the crust, such as texture, gypsum, NaCl and CaCO3. The 1 mm probe used in the experiments is very sensitive to short-distance variations of the microfabric caused by small cracks, packing voids, coarse skeleton grains, crystals or local concentration of chemical precipitates. Thus measurements of the resistance encountered by the needle when penetrating the crusted surface layer are highly aleatory. This calls for a different approach, based on high-density grid survey of the crust and application of geostatistics to account for micro-spatial variations, using spatial interpolation via kriging.  A conical steel probe maybe a good simulator of a root in silicate soil material, but less appropriate in crystalline gypsum. Since gypsum crystals rate hardness 2 and steel about 6 on Moh's scale, a steel needle readily crushes or perforates a gypsum crystal and thus penetrates into the crust. A root in the same condition probably would not pierce a gypsum crystal and, therefore, root elongation may be severely hindered by a gypsum crust. CONCLUSION Crusting is a major problem for crop production in gypseous soils, in which seedling emergence is largely controlled by the thickness, gypsum content, strength, texture, moisture condition and pattern of crystallization of the crust. Drying up of the soil surface layer causes gypsum precipitation, changes crystallization patterns and reinforces the crust strength. In coarse-grained materials, gypsum accumulation leads to cementation which causes PR values to increase, while granular crystals form in the presence of fine particles, contributing to decrease penetration resistance. The total variation of the dry PR values is controlled by texture, sodium chloride, water table height, gypsum and calcium carbonate, in this order. REFERENCES Callebaut F., Gabriels D., Minjauw W., De Boodt M. (1985) Determination of soil surface strength with a needle-type penetrometer. Soil and Tillage Research 5: 227-245. Chartres C.J., Green R.S., Ford G.W., Rengasamy P. (1985) The effect of gypsum on macroporosity and crusting of two red duplex soils. Aust. J. Soil Res. 23: 467-679. Gooderham P.T. (1973) Soil physical conditions and plant growth. PhD Thesis, University of Reading, UK. Jafarzadeh A.A. (1991) Experimental studies of gypsum migration and deposition in soil profiles. PhD Thesis, Wye College, University of London, UK . Nafie F.A.A. (1989) The properties of highly gypsiferous soils and their significance for land management. PhD Thesis, Wye College, University of London, UK . Poch R.M., Verplancke H. (1997) Penetration resistance of gypsiferous horizons. European Journal of Soil Science 48: 535-543. Watson A. (1979) Gypsum crust in deserts. J. of Arid Environments 2: 3-20. Watson A. (1982) The origin, nature and distribution of gypsum crusts in deserts. PhD Thesis, University of Oxford, UK. Acknowledgment This paper is based on data produced by the first author in the framework of his doctoral research at the Wye College, University of London, UK, under the supervision of Prof. Dr. C.P. Burnham. |