|
Auteurs: |
Souirji A. and Marcoen J.M. |
|
Institution: | Faculté Universitaire des Sciences Agronomiques de Gembloux. |
|
Adresse: | Unité de Géopédologie Faculté Universitaire des Sciences Agronomiques B-5030, Gembloux, (BELGIUM) |
|
e-mail: |
1. Introduction
Deserts cover a wide range of environmental conditions, but do share major characteristics which are an important precipitation deficit, extreme temperatures, strong wind activity and a high percentage of bare ground. These characteristics determine desert soils morphology, composition and behaviour.
It must be noted that the presence of palygorskite (and/or sepiolite) is connotative of a desert climate but it has not been reported in all desert. This is probably due to the fact that desert climate does not necessarily produce fibrous clays, but preserves them provided they exist in the parent material or in the dust that falls on the soil (Marcoen and al. 1993).
Given the shallow penetration of atmospheric precipitation, the processes of desert soil formation are particularly active in the topsoil, pedogenic development in deeper soil horizons of well-drained desert soils being generally a relict of past moister climates.
In order to explore the genesis of desert surface horizons, we shall consider below the pedogenic development paths that may be followed by a freshly deposited alluvial or aeolian sediment. Subsequently two important desert soil forming processes, namely physical soil dispersion and chemical weathering induced by clay surface acidity will be described.
2. Genesis of desert surface horizons
Development from a sediment to a typical desert topsoil can be divided in three phases: i) a stabilisation phase ; ii) a horizonation phase and iii) a maturation phase (Souirji, 1996)
(Fig.1: see Genesis diagram in http://www.fsagx.ac.be/departements/etnf/gp/desert-soil )
2.1 The stabilisation phase.
Most desert surface materials being bare, wind and rain can act with little hindrance on them. Before substantial soil development can take place, the initial sediment must first stabilise.
When there is rapid sedimentation from running water or aeolian deposition, the removal of soil material by wind deflation and rain wash is compensated or outweighed by deposition. The topsoil keeps its juvenile character and remains a C horizon because sedimentation outpaces soil development.
When there is no rapid sedimentation the development of the initial sediment will differ according to the geomorphic position, the depth to the watertable, the initial coarse fragments content and the vegetative cover.
2.1.1 In periodically ponded areas.
Desert surfaces where periodic ponding with runoff water takes place collect aeolian dust washed into them from surrounding higher terrain (Pye, 1987). This dust will infiltrate in the sediment at a rate which depends on its permeability. Two situations may occur:
2.1.2 In well-drained areas.
There are three possible situations:
2.2 Horizonation phase.
The horizonation phase consists in the formation of pedogenic horizons, marked by the onset of structural organisation in the initial sediment:
A desert pavement is present. The formation of a pavement is an essential part of the stabilisation phase and the first step in the horizonation and organisation of the sediment. In other words the pavement is itself a soil horizon which is the result of a pedogenic process (Bockheim, 1982; Parsons et al., 1986). The pavement acts as a trap for aeolian dust and sand, when present. Part of the deposited aeolian dust is washed inside the topsoil at a rate and depth which depend on the initial topsoil permeability to dry dust and to ‘slurry’ (a somewhat thick water suspension of dust) and to the amount of rainfall which drives them in. A thin seal eventually forms at the surface and thickens gradually due to repeated aeolian dust additions. If the original sediment is coarse, e.g.. sand, the incorporation of dust is initially more rapid and will eventually allow bridges to form between the skeleton grains, hence permitting the soil mass to contract and expand during drying-wetting cycles. Periodic soil dispersion can then start inside the topsoil and in the surface seal and allow the formation of a vesicular crust and a platy structure. Thin cracks can form, thus providing new inlets for aeolian dust and sand. The vesicular crust will thereafter play a major role in soil development as it reinforces the protective effect of the pavement and, most importantly, largely controls the water, solids and gases exchanges between the soil and the atmosphere.
If a source of aeolian sand is present in the landscape, the horizonation will proceed from a pavement (symbol D), to a pavement with a trapped aeolian sand veil (symbol S), D/S, to D/S/Av, with the formation of a vesicular crust (symbol Av), then to D/S/Av/A with the formation of a platy A horizon.
If no source of aeolian sand is present in the landscape, no sand veil will form and the horizonation will proceed from D to D/Av to D/Av/A. In some very cold deserts (e.g. in parts of Antarctica), coarse cryoclastic material dominates and there is little dust to be deflated and deposited by winds. In these conditions a vesicular crust and a platy structure will not form. A dense pavement with varnish, ventifacts and aeolian sand veil may directly overlie loose C horizons. We then have only a D/S surface horizon.
A desert pavement is not present. In vegetated areas, the horizonation process is similar to the case where there is a pavement and we have an evolution from Av to Av/A. It must be noted that horizons assemblages such as S/Av, then S/Av/A are possible if there is a thin layer of sand above the Av.
In closed depressions, the Az horizon may overlie a variety of Bz or Cz subsurface horizons, if groundwater is shallow. In the case of deeper groundwater, the compact takyr crust that forms gradually thickens and desiccation cracks appear upon wetting-drying cycles, hence facilitating the penetration of suspended material that comes with the onset of each flooding. The subcrustal layer usually evolves into a platy A horizon.
- Maturation phase.
This phase consists in the physical, chemical and mineralogical evolution of the surface horizons. The pavement evolves physically by reduction in size of the coarse fragments, under the influence of extreme temperatures, variations in air humidity and airborne salts, leading to a denser pavement protecting better the soil. The vesicular crust may thicken and the underlying platy A horizon incorporate more aeolian material. Desiccation cracks may widen and absorb more aeolian sand and dust.
Calcium carbonate, salts or gypsum may gradually accumulate in the A horizons, depending on the amount and nature of the dust and aerosols that affect the area, at a rate and a depth which depend again on the permeability and the amount of rainfall (and runoff). They may also accumulate as coatings below the coarse fragments of the pavement. Desert chemical weathering processes eventually induce the formation of desert varnish and the relative concentration of free oxides of manganese and iron in surface horizons (Souirji, 1996).
3. Desert soil forming Processes
Strong physical dispersion in the surface horizons of desert soils is responsible for the formation of vesicular crusts, for the widespread occurrence of platy structure in the topsoil and for the higher mobility of the clay fraction, and to a lesser extent of silt, which can be more easily leached (Souirji 1990, 1996).
The physical dispersion process is particularly well-known for surface crusts but, except for raindrop impact which does not affect deeper surface horizons, it is the same for all the periodically wetted surface horizons.
Desert vesicular crusts are structural crusts according to the classification of Valentin (1989). After Hillel (1960) structural crusts result from processes which cause particle sorting and rearrangement leading to the formation of a compact zone at the soil surface. Three mechanisms are involved in structural crusts formation: (i) Mechanical breakdown of soil aggregates due to externally applied pressure, mostly the kinetic energy of raindrop impact; (ii) Spontaneous slaking in the course of wetting-drying cycles, and (iii) Densification of the crust.
3.1.1 Mechanical breakdown
The beating action of raindrop impact breaks down fragile surface aggregates and reduces the average size of the pores of the uppermost layer of the soil which becomes more compact, hence creating a thin seal at the soil surface (Shainberg, 1992). A higher intensity of rainfall and a reduced soil surface cover, e.g. sparse vegetation, favour crust formation.
3.1.2 Spontaneous slaking.
A number of soil physicists, among which Panabokke and Quirk (1957) and Le Bissonnais (1989), have demonstrated experimentally that slaking of soil aggregates upon wetting may result from two different mechanisms which are the compression of entrapped air and differential swelling.
Compression of entrapped air. The above mentioned authors have found that loamy aggregates, i.e. those which contain less than 40% clay, remain stable when wetted under vacuum but slake when wetted under atmospheric pressure. Thus they demonstrated that it is the compression of air by the advancing moisture front which causes the slaking. Indeed, if air pressure builds-up to a point greater than the tensile strength of the soil, it escapes explosively, breaking-off the soil aggregates. This process can only take place if: i) air cannot escape or can escape too slowly to avoid pressure build-up. This means that pore walls must be relatively sealed and air circulation within them must be hampered. Gras (1974) has shown that unevenness of pore space, measured by the range of pore radii, increases air entrapment. In this regard loamy soils, which have a more uneven porosity than clayey ones, would experience more air entrapment. ii) the tensile strength of the soil is low enough. That is the case if clay and cementing agents (organic matter, exchangeable divalent cations such as calcium, and sesquioxides) contents are low. Silty soils in particular have a low strength when wet. However, the soil must be rigid enough, otherwise the gradual deformation of the soil material will not permit pressure build-up. This excludes relatively pure sands. iii) the rate of wetting is rapid. The rate of wetting is mostly controlled by the initial soil water content. In fact, the lower the initial water content the higher the potential gradient driving water into it (Panabokke and Quirk, 1957). In natural conditions the initial water content is controlled by texture and climate. In arid climates, where the relative air humidity is well below 50% for protracted periods, the surface soil is usually extremely dry. A lower organic matter content also favours rapid intake of water because the hydrophobic organic molecules slow down the water flow. Sullivan (1990) has shown that irregular distribution of organic matter in the pores induces the formation of encapsulated air which prevents rapid water flow.
Differential swelling. Panabokke and Quirk (1957) and Le Bissonnais (1989) have also found that clayey soils ( i.e. with >40% clay) do slake when wetted under vacuum. They attributed in this case the slaking to differential swelling which causes uneven strains leading to the distortion and the weakening of the aggregates internal structure.
After Bolt and Bruggenwert (1976), in case water is removed from the soil to the point that the diffuse double layer (D.D.L.) shrinks below its potential thickness, we are in the presence of a truncated diffuse double layer (T.D.D.L.). The greater the difference between the potential (maximum) extent and the actual extent of the D.D.L., the stronger the tendency of the system to absorb water and expand. Translated in terms of physical behaviour, the system is similar to an osmometer, and will develop, upon wetting, a very strong swelling pressure of the order of several tens of bar. The swelling pressure is higher if: i) the initial water content is low; ii) the adsorbed counter ions are monovalent; iii) the soil solution has a low electrolyte concentration, therefore a high salt content is unfavourable to differential swelling; or iv) higher specific area clays minerals (e.g., smectites) are present.
Panabokke and Quirk (1957) have found that slaking of clays is reduced if they are drier than pF 5.5 and attributed this phenomenon to the protective effect of entrapped air which prevent direct contact of the penetrating water with some of the colloidal interfaces. This phenomenon seems to be associated with pores smaller than 0.5µm. As a consequence, very clayey soils will not slake well in very dry conditions.
Le Bissonnais (1989) suggests that both air entrapment and differential swelling may contribute simultaneously to slaking in soils which have intermediate clay content ranges (30 to 50% clay).
3.1.3 Densification
Hillel (1960) found that the crusting zone coincides with the uppermost layer which gets saturated with water when the topsoil is wetted. As saturation is approached the original structural arrangement begins to collapse and the platy particles, in semi-suspension, tend to assume a horizontal and more nearly parallel orientation of greater density. This creates horizontal planes of weakness which are probably responsible for the occurence of frequently observed platy structure in desert surface horizons. Expanding air imprisoned in the vesicles may escape along these planes of weakness, hence creating horizontal fissures and a platy structure (Figueira and Stoops, 1983).
3.2 Chemical weathering and the formation of secondary Fe and Mn oxides.
Schwertmann and Taylor (1989) reported that, if ferrous iron is present in a silicate’s structure, the mineral will break down in the presence of protons according to the reaction:
This protolysis causes the instability and progressive destruction of the silicate minerals which will also release other metal ions such as Mn2+. The oxidation of Fe2+ would lead to the formation of a Fe(OH)3 precipitate which evolves into ferrihydrite which can be stabilised by adsorption of silicate or organic matter or evolve further into hematite (Schwertmann and Taylor, 1989). In hot oxidising conditions, Mn2+ would form Birnessite (McKenzie, 1989).
It is however necessary to explain how excess protons, i.e. strong acidity, required for the above weathering process to take place, can occur in desert soils. It is known that in a soil that becomes very dry, its clay minerals surfaces develop superacidic behaviour (Chaussidon and Pedro, 1979; McBride, 1989). Indeed, the hydrogen atoms of soil residual water molecules become about one million times more mobile than those of an ordinary soil solution, increasing their chemical activity, which may become equivalent to an H+ concentration of about 0.5 N, hence lowering the pH at the clay mineral’s surface by up to 4 units and creating superacidic clay surface conditions (Chaussidon and Pedro, 1979).
Desert climate, which is very dry, offers ideal conditions for the development of strong clay surface acidity. However, given the scarcity of precipitation, leaching opportunities are rare and concern mostly the topsoil. Hence the protolysis of the clay minerals will be slowed by the overall low rate of removal of the weathering products. Weathering can therefore be expected to be slow and superficial in well-drained soils in deserts, except in their margins where there is somewhat higher precipitation. Also, the permanent aeolian addition of material to the surface horizons masks the effect of weathering, which may appear weaker than it actually is.
The occurrence of desert varnish on pebbles and stones of desert pavements, and of iron (and manganese) coatings on sand particles, is probably caused by the weathering of adhering dust particles caused by residual water remaining from moisture films deposited by the dew or precipitation (see the review by Whalley, 1983).
3.3 Conclusion
The nature of surface horizons determine the faith of desert soils which are submitted to rapid erosion by wind and water unless protected by a pavement or vegetative cover. Surface vesicular crusts are particularly important for management because they are a serious obstacle to water infiltration. Albergel et al., cited by Casenave and Valentin (1988) have demonstrated through statistical analysis that the presence of vesicular pores in surface crusts is well correlated with unfavourable conditions for water infiltration. Vesicular crusts reduce water intake hence aggravating the moisture deficit caused by the scarcity of precipitation.
Surface horizons play a very important role in the genesis and management of desert soils, they must therefore be taken into account in their classification (Souirji, 1990, 1993, 1996; Souirji and Marcoen, 1998).
Keywords : clay surface acidity, desert pavement, desert soil, platy structure, soil dispersion, surface horizons, vesicular crust.
REFERENCES
Bockheim, J.G. 1982. Properties of a chronosequence of ultraxerous soils in the Trans-Antartic mountains. Geoderma 28:238-255.
Bolt, G.H. 1976. Surface interaction between the soil solid phase and the soil solution. p. 43-53. In G.H. Bolt and M.G.M. Bruggenwert (ed.) Soil chemistry. A. Basic Elements. Elsevier.
Casenave, A., and C. Valentin. 1988. Les états de surface: Une des clefs de l'hydrologie sahélienne. p. 61-72. In M. Demissie and G.E. Stout (ed.) Proceedings of the Sahel forum on the state-of-the-art of hydrology and hydrogeology in the arid and semi-arid areas of Africa, Ouagadougou, Burkina Faso, November 7-12, 1988. UNESCO.
Chaussidon, J., and G. Pedro. 1979. Rôle de l'état hydrique du système poreux sur l'évolution du milieu, réalité de l'altération dans les systèmes à faible teneur en eau. Bull. Assoc. Francaise pour l'Etude du Sol, Numéro Double 2 et 3, p.223-237.
Figueira, H., and G. Stoops. 1983. Application of micromorphometric techniques to the experimental study of vesicular layer formation. Pedologie, XXXIII:77-89.
Gras, R. 1974. L’emprisonnement d’air lors de l’humectation des corps poreux. Science du Sol 1:49-59.
Hillel, D. 1960. Crust Formation in Loessial Soils. p. 330-338. Proceedings of the 7th Intern. Congress of Soil Science, Madison, Wisc.
Le Bissonnais, Y. 1989. Analyse des processus de microfissuration des agregats à l'humectation. Science du Sol 27:187-199.
Marcoen, J.M., Naud, J. and Souirji, A. (1993). Aeolian processes and the occurrence of acicular clays in desert soils from south-western Saudi Arabia. p. 425-431. In Proceedings of the International Workshop on the Classification and Management of Arid-Desert Soils. Edited by the Inst. of Soil Science and Xinjiang Inst. of Biology and Desert Res., Academia Sinica. China Science and Technology Press. Beijing.
McBride, M.B. 1989. Surface chemistry of soil minerals. p. 35-88. In J.B. Dixon and S.B. Weed (ed.) Minerals in soil environments. Soil.Sci.Soc.Am. Book Series 1, second edition.
McKenzie, R.M. (1989). Manganese oxides and hydroxydes. p. 439-465. In J.B. Dixon and S.B. Weed (ed.) Minerals in soil environments. Soil.Sci.Soc.Am. Book Series 1, second edition.
Panabokke, C.R., and J.P. Quirk. 1957. Effect of initial water content on stability of soil aggregates in water. Soil Sci., 83:185-195.
Parsons, R.B., R.C. Herriman and T.D. Cook. 1986. Geomorphic surfaces and soils, Colorado river area, Arizona and California. Technical Monograph, S.C.S. U.S.D.A.
Pye, K. 1987. Aeolian dust and dust deposits. Academic Press.
Schwertman, U., and R.M. Taylor. 1989. Iron oxides. p. 379-438. In J.B. Dixon and S.B. Weed (ed.) Minerals in soil environments. S.S.S.Am. Book Series 1, second edition.
Shainberg, I. 1992. Chemical and mineralogical components of crusting. p. 33-53. In M.E. Sumner and B.A. Stewart (ed.) Soil crusting, chemical and physical processes. Advances in Soil Science, Lewis Publishers, Boca Raton.
Souirji, A. 1990. Classification of aridic soils, past and present: Proposal of a diagnostic desert epipedon. p. 175-184. In J.M. Kimble (ed.) Proceedings of the sixth international soil correlation meeting (VIth ISCOM)- Characterization, classification, and utilization of cold Aridisols and Vertisols. 1989. USDA, Soil Conservation Service, National Soil Survey Center, Lincoln, NE.
Souirji, A. 1993. Definition of desert soil taxa for the FAO-UNESCO-ISRIC legend of the soil map of the world. p. 418-424. In Proceedings of the International Workshop on the Classification and Management of Arid-Desert Soils. Edited by the Inst. of Soil Science and Xinjiang Inst. of Biology and Desert Res., Academia Sinica. China Science and Technology Press. Beijing.
Souirji, A., 1996. Pedogenesis and classification of desert soils: an example from the Arabian peninsula. Doctoral thesis, Faculté Universitaire des Sciences Agronomiques de Gembloux, Belgium.
Souirji, A. and Marcoen, J.M. 1998. Using topsoil characteristics to classify desert soils. Poster presentation – 16th World Congress of Soil Science, Montpellier.
Sullivan, L.A. 1990. Soil organic matter, air encapsulation and water-stable aggregation. J. of Soil Sci. 41:529-534.
Valentin, C. 1989. Les états de surface des savanes de l'ouest africain: Relations avec les sols et incidence sur l'économie en eau. In Les sols tropicaux: Bien les connaître pour mieux les utiliser. SOLTROP, Premier Séminaire Franco-Africain de Pédologie Tropicale, Lomé.
Whalley, W.B. 1983. Desert varnish. p. 197-226. In A.S. Goudie and K. Pye (ed.) Chemical sediments and geomorphology: Precipitates and residua in the near-surface environment. Academic Press.