EVALUATION OF SOME OXISOLS, ULTISOLS AND INCEPTISOLS WITH THEIR PRACTICAL SIGNIFICANCE IN SIERRA LEONE BY Patrick Magagie Sutton A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Crop and Soil Sciences 1978 ABSTRACT EVALUATION OF SOME OXISOLS, ULTISOLS AND INCEPTISOLS WITH THEIR PRACTICAL SIGNIFICANCE IN SIERRA LEONE By Patrick Magagie Sutton Eleven representative profiles from four physiographic groups were studied. The soils of the Upland Surfaces of Highly Weathered Material are Oxisols. Segbwema and Timbo series of the Steep Hills and Slopes were Ultisols and Inceptisols, respectively. In the soils of the Colluvial Footlepes and Upper River Tributary Terraces, Pendembu series belongs to Oxisols and Masuba series to Ultisols. The soils of the Alluvial Terraces and Floodplains are all Inceptisols. Following are some of the other findings and relationships. Percent clay values, when free Fe oxide is removed before particle size analysis, are higher than when free Fe oxide is not removed and are also commonly higher than clay estimated by the factor 2.5 x 15 bar moisture content. 3.0 x 15 bar H20 seems to give a better estimate of the percent of clay in these soils. The criterion: sum of exchangeable cation + exch. Al x 100 < % Clay (Fe removed) - 10 me/lOOg does not separate the oxic from non-oxic horizons of the soils studied; all pedons studied meet this criterion. A critical value Patrick Magagie Sutton of 0.06 for the ratio * % Fe.g d - % Fe29_ox % clay (Fe removed) gives a better separation of the soils, but is also unsatisfactory. Ironstone nodules adsorbed a relatively smaller percent of P than did the fine earth fractions of representative gravelly soils. P adsorbed by the surface and subsurface horizons of the fine earth fractions is correlated with percent organic C, Al 0 ox, Fe 0 d and 2 3 2 3 Fe203ox. The ironstone nodules have a diluting effect on the total amount of P fixed by the gravelly soils. The dominant clay mineral in the total and fine clay is kaolinite, which is less ordered in the fine clay. Clay mineralogy of the ironstone nodules is similar to that of the fine earth fractions of two representative profiles. Mica flakes in the medium and fine sand fractions of Segbwema and Timbo series are mainly inter- layered illite-chlorite and kaolinite. Thin section studies of B horizons of whole soils showed that most of the soils lack argillic horizons. Thin sections of ironstone nodules showed two main types: one derived from rock fragments and the other from plinthite. Argillans lined old channels of some nodules, indicating past genetic processes. The term petroplinthite is suggested in place of skeletal at the family level when the >2mm fraction is dominantly ironstone nodules. *Fe203 and A1203 total, and sodium bicarbonate citrate dithionite, ammonium oxalate, or sodium pyrophosphate extractable components are referred to as Fe203t and Fe203d, Fe203ox or Fe203pp and A1203d, A1203ox or A1203pp, respectively. ACKNOWLEDGEMENTS The author is very appreciative of the guidance of Dr. E. P. Whiteside, Professor, CrOp and Soil Science, Michigan State University, under whose direction this work was done. He is also grateful to Dr. B. G. Ellis, Dr. Del Mokma, Dr. M. M. Mortland, Dr. R. T. Odell and Dr. H. Winters, all members of his guidance committee, for their useful suggestions. Sincere thanks are due to the African American Institute for providing financial support for the study, Dr. Ray Laurin for helping with the photographs, and Mr. James Cawray, of Njala University College,for helping to collect the soil samples. Finally, the author is very grateful to his wife for her encouragement. ii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . IJST OF FIGURES. . . . . . . . . . . . . . . LIST OF PLATES . . . . . . . . . . . . . . . . DHRODUCTION . . . . . . . . . . . . . . . . . . . LITERATUREREVIEW................ Laterite and Lateritic Soils. . . . . . . . . . Definitions (Laterite, Lateritic Soils, Plinthite and Ironstone) . . . . . . . . . Genesis and Mode of Formation of 'Laterite' (Plinthite and Ironstone). . . . . . . . . . Classification of Laterite (Plinthite and Ironstone) . . . . . . . . . . . . . . . . Classification of Lateritic Soils (Oxisols, Ultisols and Inceptisols). . . . . . . . . Genesis of Oxic, Argillic and Cambic Horizons. Criteria and Limits of Oxic, Argillic and Cambic Horizons in Soil Taxonomy (1975) . Oxides and Hydroxides of Fe and Al . Soil Phosphorus (Adsorption) . . . . . . . . . Geology and Soils of Sierra Leone . . . . . . . . . . Physiography, scales" and Paront Motorial . . Time . . . . . . . . . . . . . . . . . . . . . Climate. . . . . . . . . . . . . . Organisms. . . . . . . . . . . . . . . . . . . Topography . . . . . . . . . . . . . . . . . . Taxonomy of the Major Kinds of Soils in Sierra Leone. Present Land Use and Management in Sierra Leone MATERIALS AND METHODS. . . . . . . . . . Field Investigations. . . . . . . . . Physical Analyses . . . . . . . . . . . . . . . . . . Particle Size Analyses . . . . . . . . . . . . Particle Size Analyses, After Fe Oxide Removal Chemical Analyses . . . . . . . . . . . . . . . . . . Previously Reported by Odell et a1. (1974) . . Iron and Aluminum Extractions. . . . . . Phosphorus Adsorption. . . . . . . . . . . . . Mineralogical Analysis. . . . . . . . . . . . . . . . Micromorphology . . . . . . . . . . . . . . . . . . . iii Page vi vii ix 10 12 15 17 2O 21 21 23 24 25 27 27 29 33 33 36 36 36 36 36 37 38 39 4O RESULTSANDDISCUSSION.................... General Grouping of the Soils . . . . . . . . . . . . Physical Analyses . . . . . . . . . . . . . . . . . . . Total Clay . . . . . . . . . . . . . . . . . . . Fine Clay (ils are presently classified as Oxisols because they do not meet 1:}1e requirement of at least 3% of weatherable minerals within the Ebznofile. Most of the soils so classified are found on the lower 1:6zrraces of the major rivers or they are alluvial in nature with DniJiimal profile development. The source of their parent material is hnelieved to be the soils of the old erosional surfaces, which are Pznesently classified as Ultisols. Also, by virtue of their physio- graphic positions, they are considered to be the youngest soil larudscapes. Should these soils be classified as Oxisols derived from Ultisols? Soils presently classified as Ultisols mostly occupy the oldest Posit:ions on the landscape. Many have hardened plinthite gravels (ironstone gravels) that extend from the surface to depths greater than 14.55 meter within the profiles. Generally, the ironstone gravels 2 éaccount for 20—70% of the total soil weight. These ironstones are Inelieved to be the result of active soil formation processes within tzhe profile. They are believed to represent the irreversibly hardened iform of plinthite. Plinthite (which represents an earlier stage in t:he formation of ironstone gravel) is important in the separation of (Exxisols from Inceptisols and Oxisols from Ultisols in Soil Taxonomy. fi<3wever, ironstone gravel does not appear to be useful in making the Eilaove—mentioned separations in the Taxonomy. Also, the argillic horizons that are present in some soils of t:}1e upland erosion surfaces (based on the B/A clay ratio) could be E1 result of secondary development in areas dominated by termite nn<>umds high in water-dispersible clays. Termites, through their Eicztivity in the soil, bring fine textured material from the subsoil t:<> the surface horizons where they are used for building mounds. “PINE mounds may eventually be eroded, and fine fractions can move 1:}1rough the surrounding soils by mechanical means. Eleven profiles representing soils from three of the six soil Plfiovinces where soils have been surveyed are studied in this research. TPuese soils are selected from four cited physiographic positions on the landscape. The objectives are: 1. To determine additional mineralogical, physical, and Chemical characteristics of these soils. 2. To study the micromorphology of the profiles. 3. To evaluate some mineralogical, physical, chemical and miCromorphological characteristics of the ironstone gravels. 4. To evaluate the present criteria and/or limits used in sseaparating the Oxisols, Ultisols and Inceptisols in the light of 2 and 3 above and to make suggestions for possible modifications. 1 r 5. To present a hypothesis concerning the genesis of the soils 5 tudied. 6. Finally, to relate the characteristics of these soils to t:})eair potential uses and management for agricultural purposes. Ln‘ufi‘ ’- .— LITERATURE REVIEW Laterite and Lateritic Soils Definitions (Laterite, Lateritic Soils, Pl inthite and Ironstone) The term laterite has been used for many years to describe Sasquioxide-rich material found in soils of tropical and subtropical regions. This material has been studied since the 19th century by many researchers who have defined it in different ways. Buchanan in 1807 (cited in Humbert) was the first to define the material that was called laterite. His definition was "an iron- oRide-rich, indurated quarryable slag-like or pisolitic illuvial horizon developed in the soil profile." This definition is restrictive and the material so defined occurs to only a limited extent in the trOpics. Evans, in 1910, referred to laterite as material that con- tains some Al oxide and Fermor, in 1911, considered use of the term only for soft material (that contains Fe) that can be cut into bricks. Prescott (1931) defined laterite to include hard ferruginous Surface formations and A1 rich material. Walther (1889, 1915 and 1916) (cited in Sivarajasingham et al., 1962), thought that laterite signified red colors and proposed that the term be used for all red-colored alluvial material. Later, as the need for a standard definition of laterite became i . mpc’rtant, chemical analyses of the material were conducted. The — t“ .. I _-v r” )"' 5 5:102/A1203 and Si02/R203 ratios were adOpted as the basis for calling a soil laterite, lateritic and non—lateritic (Martin and Doyne, 1927 and 1930) . Baldwin et a1. (1938) and Thorp and Smith (1949) used the terms laterite and lateritic soils for zonal great soil groups, found in hmnid tropical and subtrOpical areas. In 1946, Pendleton and Sharasuvana defined laterite soils as " one in which a laterite horizon is found in the profile." du Prez (Cited in Sivarajasingham et al., 1962) supported Pendleton and Srharasuvana definitions of laterite soil and lateritic soil, but did not include the presence of A1 as an important criterion. Mohr and van Baren (1954) supported the use of laterite and lateritic for soils composed of similar weathering products that Produce soil as well as material that hardens. In 1949, Kellogg used the term laterite to describe four kinds of material that are hard or that harden upon exposure. These materials include (a) soft mottled clays that change irreversibly to hEnfdpans or crust when exposed; (b) cellular and mottled hardpan and crusts; (c) concretions or nodules in a matrix of unconsolidated material; and (d) consolidated masses of such material, i.e., concre- tions or nodules. Alexander and Cady (1962) gave a concise definition of laterite as a highly weathered material rich in secondary oxides of Fe, A1, or both. It is nearly void of bases and primary silicates, but it may contain large amounts of quartz and kaolinite. In the USDA Soil Survey Staff (1960) Soil Classification System, a new term was introduced, plinthite, with the intention to avoid 6 confusion arising from the use of the term laterite. Plinthite as cieafined in this system referred to the soft laterite material. It is a sesquioxide rich,humus poor, highly weathered mixture of clay with quartz and other diluents, which commonly occur as red mottles usually in platy, polygonal, or reticulate pattern. Plinthite changes irreversibly to hardpans or irregular aggregates, on repeated wetting arid drying. It is a form of the material which has been called 1 aterite. This definition is also used by the FAO System (FAQ-UNESCO, 1974) for Fe-rich clay which can be cut with a spade. The hardened form of this material is called ironstone in both S)r'stems. In this dissertation laterite is considered to include both plinthite and ironstone. Se‘nesis and Mode of Formation of 'Laterite' (Plinthite and Ironstone) Several theories on the genesis and formation of laterite have been prOposed, dating back to mid to late 19th century. Some of the t1'leories proposed, included residual weathering in place and volcanic Origin with subsequent weathering. Holland (1903) supported the idea Of weathering of material in place to give laterite, but concluded that a simple chemical weathering cannot explain the abrupt transi— tion from laterite to weathered rock. Humbert (1948) conducted Studies on laterite found in New Guinea. He concluded that under COnditions of abundant moisture and high temperature in humid equa- tOrial regions, a rigorous weathering and transformation of parent material occurs. At the final stages of these processes, laterization o ' u u a . c‘-‘urs in which oxides of Fe are concentrated, which become indurated V Wad-'- . 7 upon exposure. He also advanced the idea that three main stages are involved in the genesis of laterites: (a) advanced decomposition of the mineral constituents of the parent rock, release and removal of Si from the surface horizons and separation of the sesquioxides; (b) formation of free Fe oxides by decomposition of ferromagnesium material; and (c) dehydration resulting in color change. "As dehy- dration progresses the surface is reduced and eventually a compact con cretion is obtained. The Fe oxide is dehydrated and irreversibly fixed. This is the COncretion stage. As the hydrated oxides are precipitated in the Vicinity of the concretions, a concentration gradient is established, Fe moves with the gradient and the concretion grows in size as the SL1pply of Fe moves into position through voids and channels of the Weathering matrix. Growth of the concretion continues until an indurated crust is formed." D'Hoore (1955) suggested two chemical processes that are involved in the formation of laterites: (a) concentration of sesquioxides by reI't‘toval of Si and bases; or (b) concentration of sesquioxides by accumulation from outside sources. Alexander and Cady (1962) suggested that materials within the Observed range of composition of laterite may be developed from rock in place by several possible courses of weathering and mineral trans- formation, all of which involve almost complete removal of bases and at least substantial losses of the combined Si of primary minerals. Hamilton (1964) and Schmidt (1964) observed fine droplets of Fe o"Yhydraltes in micromorphological studies they conducted on laterite. T he latter relates the development of laterite to a coalescence of 8 triese droplets. Later, micromorphological studies on laterite by some researchers, including Eswaran and Mohan (1973), did not tend to support Schmidt's theory. What is clear in these theories is that all of the researchers agreed that the removal of Si is necessary for the formation of laterite. Classification of Laterite (Plinthite and Ironstone) Several classifications of laterite have been proposed which are either based on genetic or morphological (including micromorphological) PrOperties. Genetic classifications have been suggested by Aubert (1963), Maignen (1959) and D'Hoore (1954). Pendleton and Sharasuwana (1946) classified laterites by their morphological properties. They suggested two forms, (1) vesicular and (2) pisolitic, with many intermediate types; vesicular laterite may be soft or of varying hardnesses. Alexander and Cady (1962) recog- 1'lized three main types of laterite: (a) residual laterite, which has eVidence of rock structure; (b) laterite with features that resemble those of soil; and (c) pisolitic laterite composed of pisolitic bOdies more or less closely packed together. Sivarajasingham et a1. (1962) used the term nodular laterite as material consisting of individual concretions, pisolites or other crudely round masses, usually the size of a pea but commonly larger or smaller. It is generally ferruginous and cementation of the nOdules gives rise to pisolitic or concrete-like laterite. Young (1976) proposed a morphological classification of laterite b . . asGd on that of Pullman (Cited in Young). He recognized five main 9 types of laterite: (a) Massive laterite: This material possesses a continuous hard fabric, its subdivisions are cellular laterite and vesicular laterite. (b) Nodular laterite: This consists of indi— xzidual, approximately rounded, concretions for which he suggested .fflour subdivisions, (i) cemented nodular laterite, (ii) partly c:e:mented nodular laterite; (iii) non—cemented nodular laterite, and (c) Recemented laterite. (d) Ferrugenized (.isv) iron concretions. Rock structure is still visible, but with substantial iso— zr<><3kz (e) Soft laterite: Mottled Fe—rich morphous replacement by Fe. C=Zl.aay which hardens irreversibly on exposure to air or to repeated We tting and drying . The nodular types of laterite have been called different names JJTI the literature. These include lateritic concretions, ironstone n(Dciules and glaebules (Brewer and Sleeman, 1964). Young's (1976) ferruginized rock is what has been referred to £353 residual laterite by Alexander and Cady (1962), and his definition (315 .soft laterite is envisaged in the definition of plinthite in the Soil Taxonomy, USDA (1975). Westerveld (1969) recognized two types of nodular laterite in 631 Iapland soil of Sierra Leone based on their external morphology. (MIG! type he called SLC (Smooth Laterite Concretion). These are renlnded and darker in color than the RLC (Rough Laterite Concretions), Which are more angular and lighter colored. Eswaran and Mohan (1973) ‘NDServed that the matrix of indurated laterite concretions which they Studied were predominantly clay which had been coated by Fe deposi- tion. Plinthite has been recently classified by Daniels et al. (in Press) into platy and nodular forms. 10 Classification of Lateritic Soils (Oxisols, Ultisols and Inceptisols) Lateritic soils have been classified in various classification systems, including the USDA, French, Belgian and FAQ—UNESCO (Soil Map (of the Wbrld). In Marbut's time, these soils were referred to as Ferruginous .Igeaterite soils. In 1938 Thorp and Baldwin considered Lateritic ssc>ils as zonal soils with great soil groups such as Reddish-Brown ILaasteritic soils, Yellowish-Brown Lateritic soils and Laterite soils. 1:11 the FAQ—UNESCO (1974) system, Ferralitic soils are classified in tlkiee highest category as Nitosols, Acrisols and Ferralsols. In the FIrench system they can be classified as $015 Ferrallitigues and $015 Hb’ciromorphes. Young (1976), using a modified form of the CCTA (Commission for 17€3<2hnical Co—operation in Africa) 1964 Soil Map of Africa, suggested t1‘lree major divisions of Latosols. (The term Latosol was first used 533’ .Kellogg in 1949 to include the zonal soils in tropical equatorial regions that have their dominant characteristics associated with low Si“sesquioxide ratios of clay fractions, low base exchange capacity, J‘DVV activities of the clay, low content of most primary minerals.) YC>'~-‘lng's divisions were: Ferrallitic soils, Ferruginous soils, and SC’ils derived from basic rocks (Basisols). In the USDA Soil Survey Staff (1960-67) Soil Classification and 531 Soil Taxonomy (1975), soils that were described as great groups of Lateritic soils in the 1938 classification were absorbed into four of the ten orders in the system. These orders include Oxisols, Ultisols, Alfisols and Inceptisols. The Ferrallitic soils of Young and the Ferrasols of the FAQ-UNESCO system are mainly Oxisols and 11 Ultisols in the USDA Soil Taxonomy. The Ferruginous soils of Young are the Ultisols and Alfisols and the Basisols can be put in the orders Oxisols, Ultisols, and Alfisols, depending on the degree of weathering. The Reddish-Brown Lateritic and Yellowish—Brown Lateritic soils of the 1938 classification fall mainly in the order Ultisols and a few in the Alfisols and Oxisols. A majority of the Laterite soils are Oxisols in Soil Taxonomy. The single most diagnostic properties of Oxisols, Ultisols and Inceptisols, respectively, in Soil Taxonomy (1975) are the presence of oxic, argillic and cambic horizons (the genesis of these three diagnostic horizons is discussed in the next section). A major criterion for separating the Oxisols from the Ultisols and Inceptisols is the presence of plinthite within 300m of the soil surface. Plinthite that forms a continuous phase or constitutes more than half the matrix of some sub-horizon within 1.25m of the surface is considered as diagnostic at the great group level, e.g., Plintha- qualf, Plinthaquepts, Plinthustalf, Plinthudult, etc. If plinthite occurs in the soil but does not constitute a continuous phase, a plinthic subgroup is used (Soil Survey Staff, 1960, and Soil Taxonomy, 1975). However, hardened laterite (irreversible) materials, which include ironstone (by definition),are not plinthite, and are there- fore excluded as diagnostic properties at the great group and sub- group levels discussed above. However, hard materials rich in secondary Al oxides are diagnostic at the great group level in the order Oxisols, e.g., Gibbsiaquox. 12 A diagnostic petroferric layer is recognized as separate from a lithic contact. The petroferric layer is more or less a laterite sheet (vesicular laterite). Most laterite materials are rich in Fe oxides. These are excluded from the limits of an oxic horizon that apply to <2mm material. In other classification systems, e.g., the French system, con— cretionary laterite and lateritic crust (sheets) are diagnostic at the subgroup level, and hardened (indurated) subgroups have been described within the Ferrallitic groups. Sys in 1968 proposed the inclusion of great groups and sub- groups in the Ultisols, Oxisols and Inceptisols with hardened lateri- tic materials as diagnostic features. He proposed the name petro- plinthic for Fe and/or Al oxide individualizations which have hardened irreversibly. When moist it cannot be cut with a spade. It appears.as hard concretions in a clayey matrix or as a hard crust or sheet. He suggested the use of petroplinthic horizon as a diag— nostic horizon for classification at the group level, and also the presence of petroplinthite in the profile as a diagnostic subgroup property. Genesis of Oxic, Argillic and Cambic Horizons Oxic horizon: The oxic horizon is the major requirement for the Oxisols as defined in the USDA Soil Taxonomy (1975). The current theory endorsed by the U. 8. Soil Conservation Service is that it occurs in soils of very old, stable geomorphic surfaces (Mid- Pleistocene or older rather than late or post-Pleistocene). The old age of the oxic horizon has allowed time for mixing by plant roots l3 and animals so that there is little or no evidence of original rock structure, with the exception that if Fe oxides or gibbsite coat and cement fragments of weathered rock, the original rock structure may be retained in the interior of the cemented parts. Weatherable minerals are absent or present only in traces, which makes this horizon low in bases except for those held in exchange complexes and plant tissues. Most of the soils with oxic horizons are found in tropical and subtropical climates. They usually occur on nearly level or gentle slopes. The geomorphic position is one in which weathered sediments could have been deposited and not one in which recent unweathered sediments could accumulate. Argillic horizon: The horizon of illuvial silicate clays in soils has been recognized in the USDA Classification Systems (Soil Survey Staff, 1960, and USDA Soil Taxonomy, 1975) as a diagnostic horizon at the highest level of classification. This horizon is called an argillic horizon and it is a major diagnostic property of Ultisols and Alfisols. Several theories concerning the dispersion, migration and accumu— lation of the silicate clay in this horizon have been proposed by various researchers (Jenny and Smith, 1934; Hallsworth, 1963). One such theory is that of Kubiena, as cited by Stephen (1960). He suggested that the presence of colloidal Si acts as an efficient peptizing agent to confer on kaolinite and halloysite, swelling capacity and plasticity as well as extraordinary hardness when dry. Kubiena's idea was recognized by other workers (Hallsworth, 1963), who also suggested that particles can be kept in a condition of l4 suspension under the protective influence of organic matter and Si. Studies on clay migration have also been conducted by Jenny and Smith (1934), Buol and Hole (1961) and Hallsworth (1963). How- ever, there seems to be no general agreement among researchers about the processes involved in clay movement from the A to the B horizons. The current theory recognized by the Soil Survey Staff (1975) is that of mechanical migration. The stages involved are: (a) The parent material contains very fine clays or weathering must produce them. The very fine clays carry negative charge, as does the soil matrix, and tend to disperse, unless flocculated by salts, including carbonates and free oxides. (b) Wetting of the dry soil leads to disruption of the fabric and to dispersion of clay. Once dispersed, the clay is believed to move with percolating water and to stOp where the percolating water steps, Water percolating in noncapillary voids commonly is stopped by capillary withdrawal into the soil fabric. During the withdrawal the clay is believed to be deposited on the walls of the noncapillary voids. Carbonates can also be effective in stOpping the moving clay. Mixing of horizons by animals, frost, shrinking and swelling must be slow or absent to permit formation of an argillic horizon. Cambic horizon: This is a major diagnostic horizon for the order Inceptisols. In Soil Taxonomy (Soil Survey Staff, 1975), a cambric horizon is considered to be an altered horizon in which the texture of the <2mm fraction is very fine sand or loamy very fine sand or finer. Two types of alterations in this horizon are physical and chemical. 15 Physical alteration is the result of (a) movement of soil particles by frost, roots or animals to a point at which most of the original rock structure is destroyed, including the fine stratifica— tion of silt, clay and very fine sand in alluvial or lacustrine deposits, and (b) aggregation of the particles into peds. Chemical alteration is the result of (a) hydrolysis of some of the primary minerals to form clays and liberation of sesquioxides, (b) solution and redistribution, and (0) reduction and segregation or removal of free Fe oxide along with biologic decomposition of inherited organic matter. Criteria and Limits of Oxic, Argillic and Cambic Horizons in Soil Taxonomy (1975) Oxic horizon: This horizon must be at least 30cm thick and have 10me or less of NH4OAC extractable bases plus Al extractable with 1.0N KCl per 1009 clay; C.E.C. of fine-earth fraction (NH4OAC) of 16 me or less per 1009 clay; only traces of weatherable minerals; and <5% by volume that shows rock structure. The soil texture is sandy loam or finer in the fine-earth fraction and has >15% clay and soil horizon boundaries are gradual or diffuse. Argillic horizon: This is an horizon that contains illuvial silicate clay. If an eluvial horizon remains and there is no litho- logic discontinuity, the argillic horizon contains more total clay, and more fine clay, than the overlying eluvial horizon as follows: If the eluvial horizon has <15% clay in the fine-earth fraction, the illuvial horizon should have 3% more total clay or the ratio of fine clay to total clay in the illuvial horizon should be one-third 16 greater than that in the overlying eluvial horizon or underlying horizon. If the total clay in the eluvial horizon is >15% but less than 40% in the fine-earth fraction, the ratio of illuvial/eluvial horizon clay should be 1-2 or greater. The ratio of fine clay to total clay in the illuvial horizon is normally one—third or more greater than in the eluvial horizon. If the total clay is >40% in the eluvial horizon, the illuvial .h<>rizon should have 8% more clay, or if the total clay is >60%, 8% more fine clay is required in the illuvial horizon. These clay J.-1’l<:reases are reached within 30cm or less vertically. The argillic h(Burrizon should be at least one-tenth as thick as the sum of the t1“Inchkness of all overlying horizons or it should be 15cm or more thick 14:? the eluvial and illuvial horizons together are >1.5m thick. If the argillic horizon is sand or loamy sand, it should be at least ls5<::m thick. If the argillic horizon is loamy or clayey, it should IDEE at least 7.5cm thick. Clay skins should be present on ped sur— féi<2es (vertical and horizontal) or thin sections should show oriented (31~Eiys in 1% or more of the cross section. If the horizon is clayey, ‘113 the clay is kaolinitic, and if the surface horizon has >40% clay, it: should have some clay skins on peds and in pores in the lower :péilrt of the horizon that has blocky or prismatic structure. If there is a lithologic discontinuity in the profile between (tiles eluvial horizon and the argillic horizon, or if only a plow 1ENYer overlies the argillic horizon, the argillic horizon needs to ‘hiive clay skins in only some part, either in some fine pores or, if PEds exist, on some vertical and horizontal ped surfaces. Either ‘- M or. l7 thin sections should show that some part of the horizon has about 1% or more of oriented clay bodies, or the ratio of fine clay to total clay should be greater than in the overlying or the underlying horizon. Cambic horizon: For a cambic horizon in soils such as those in this study, soil texture should be very fine sand or loamy very fine sand or finer in the fine-earth fraction; there should be soil Stzructure or absence of rock structure in at least half the volume; £5iugnificant amounts of weatherable minerals (enough amorphous or 2:1 iléafttice clays to give >3% weatherable minerals or >6% muscovite); €3\?Zidence of some alterations, e.g., gray colors, an aquic moisture 17€2€3ime or artificial drainage; and cation exchange capacity >16 me/ l-C)(Dg clay; regular decrease in amounts of organic C with depth; and 3- (zontent of <0.2% organic C at a depth of 1.25m below the surface ‘33? immediately above a sandy-skeletal substratum! that is at a depth 015 210cm. On Steep Slopes, the soils are shallow over parent rock and on Colluvial Footslopes the soils usually have a gravel-free colluvial layer between 60 and l20cm thick. On the alluvial terraces the soils are usually developed on deep alluvium. The soils of the Upland Surfaces are probably on former pene- plains (Stobbs, 1963) from the Tertiary Period. On the Alluvial Terraces and Floodplains, different levels of terraces may occur, i.e., upper, middle and lower terraces. The Upper Terraces can be associated with the soils of the Colluvial Footslopes. The different levels of terrace are more obvious in soils associated with the Rokel River series. The differences in terrace levels probably occurred during the Pleistocene Epoch. Taxonomy_of the Major Kinds of Soils in Sierra Leone The major diagnostic horizons recognized in Sierra Leone to date (Odell et al., 1974) are as follows: (a) surface horizons - ochric and umbric epipedons, and (b) subsurface horizons - argillic, oxic, IF. KN. Qt 28 spodic, cambic, and sulfuric horizons and also plinthite. The definitions of these horizons are given in Soil Taxonomy (Soil Survey Staff, 1975). Oxisols, Ultisols, Inceptisols, Entisols and Spodosols are the five soil orders that have been recognized to date. According to the present taxonomic classification of the soils (Odell et al., 1974), the majority of the soils in Sierra Leone belong to the order Ultisols. The classification of the eleven profiles that will be studied in this research are given below, as reported by R. T. Odell et al., Baoma series (144801A) Manowa series (Kpaubu I) Njala series (N109) Makeni series (P2) Segbwema series (145005) Timbo series (p19) Pendembu series (Kpuabu 2) Masuba series (p9) Moa series (Kpuabu 3) Gbesebu series (N125) Makundu series (P104) in 1974: Typic Paleudults (or Tropeptic Haplorthox), clayey over clayey-skeletal, kaolinitic Orthoxic Palehumults (or Typic Umbriorthox), clayey-skeletal, oxidic Orthoxic Palehumults, clayey-skeletal, oxidic Typic Paleudults, clayey-skeletal, oxidic Tropeptic Haplorthox (or Udoxic Dystropepts), fine-loamy, mixed Typic Umbriorthox (or Udoxic Dystropepts), fine-loamy, skeletal, mixed Typic Paleudults, fine-loamy, mixed "Plinthic" Udoxic Dystropepts, fine-loamy mixed Tropeptic Haplorthox (or Fluventic Udoxic DystrOpepts), clayey, kaolinitic Fluventic Udoxic Dystropepts, fine-clayey, kaolinitic Plinthic "Tropeptic" Umbriorthox, clayey, kaolinitic CE ed ( is 29 Most of these soils may have evidence of diagnostic properties of another soil order. The second classification in parentheses indi— cates the problem of placing the soils in Soil Taxonomy, a major concern in this dissertation. Present Land Use and Management in Sierra Leone The most common systems of cultivation in Sierra Leone are upland farming and swamp farming (Sivarajasingham, 1968; Waldock et al., 1951). These are done mainly by peasant farmers. Upland farming_is basically shifting cultivation or what is sometimes called a bush—fallow system. The system involves felling the forest or farm bush, farming the land for a few years before returning it to fallow, while newly cleared land is farmed. The period of cultivation of the cleared land varies from one to three years. The fallow period also varies from four to ten years, depend- ing on the population pressure on the land. In the Makeni area it is about four to five years, six to seven years in the Njala area, and up to ten years in the Kenema area. The main advantage of this system is that it helps to restore the natural fertility of the soils during the fallow period. The major operations involved in shifting cultivation are: (a) brushing (cutting down vegetation); (b) burning; (c) planting; (d) weeding; and (c) harvesting. The major crop is upland rice, but within the rice, a large number of subsidiary crops, such as pumpkin, tomatoes, cassava, maize and other vegetables, can be found. On the Njala and Makeni sample sites, the upland soils are used mainly for rice production. On the Kenema site (Figure 1), aside from upland rice cultivation, permanent tree crops, such as cocoa and coffee, are also grown (Plate 1 A,B,C). Some representative soil profiles and land uses. PLATE 1 . 30 Baoma soil under Baoma Profile cocoa plantation 31 ouoz .cocwaadm canoe: roan» map .Amv uflmv waflmoum chxmz Ape .mucmam moooo who moons .chos wuHEumu paw mofiwmm onsopcom :ue3 pwumfloommm ommumpcmq on Apmscaucooo H maf the soils in this study are from such areas. In the swamps, rice is grown during the rainy season and vege- tzables are often grown during the dry season. The same swampland jL:s cultivated for many years, unlike the Uplands. The original gyirassy vegetation is cut and then piled into mounds where it is ailllowed to rot. The seedbed is prepared by hoeing and spreading tilde rotted organic material. Rice is transplanted from a nearby rillrsery. After the rice is harvested, the stubble is buried again jsri mounds, which are then used for vegetable production during the CiITy season. The above procedures are frequently practicedin the NIjala and Makeni sample areas. The swamps are utilized less in the Re nema area. The two farming systems described are time-consuming and very zliibor-intensive. All the Operations are done by hand. Crop yields Eilfla usually very low and are commonly only sufficient for the farmer's rNeeds. Improved swamp-rice growing practices are now being intro— dLiced into the three sample areas. MATERIALS AND METHODS Field Investigations The field investigations were carried out in Sierra Leone, West Africa. The samples were collected from three governmental provinces in the country, viz., Eastern, Southern and Northern Provinces (Figure 1). They also represent the soils of three major soil provinces, of Sierra Leone (Odell and Dijkerman, 1967). These soil provinces are: (a) L-soils of the Upper Moa Basin; (b) G-soils of the Rokel River series under secondary bush; and (c) J-soils of the Escarpment Region from granite and acid gneiss under secondary bush and forest, respectively. Eleven pits were sampled. Five of these are from soil province L, two from soil province G, and four from soil province J (Odell et al., 1974; van Vuure and Miedema, 1973). The soil series studied from each province, their reported tentative classifications, and their moisture regimes are as follows: Soil Province L Subgroup Moisture Regime Baoma Series Typic Paleudults WD (4 months)* (or Tropeptic Haplorthox) Manowa Series Orthoxic Palehumults MWD-WD (4 months)* (or Typic Umbriorthox) Segbwema Series Tropeptic Haplorthox WD (3 months)* (or Udoxic Dystropepts) 33 314 "v N. Longitude 1d ON. Latitude 5 Figure 1. Asterisks indicate the locations from which soil samples were collected. Also shown are the soil province boundaries of A thru P. °N. Latitude 9 314. ‘f N. Longitude Asterisks indicate the locations from which soil Figure 1. Also shown are the soil province samples were collected. boundaries of A thru P. 35 Soil Province L Subgroup Moisture Regime Pendembu Series Typic Paleudults MWD (2 months)* Moa Series Tropeptic Haplorthox WD (1 month)* (or Fluventic Udoxic Dystropepts) Soil Province G Njala Series Orthoxic Palehumults WD (5 months)* Gbesebu Series Fluventic Udoxic WD (2 months)* Dystropepts Soil Province J Makeni Series Typic Paleudults WD (5 months)* Timbo Series Typic Umbriorthox WD (4 months)* (or Udoxic Dystropepts) Masuba Series Plinthic Udoxic MWD (3 months)* Dystropepts Makundu Series Plinthic "Tropeptic" WD (2 months)* Umbriorthox These profiles are from the same sites described in Odell et a1. (1974). New surfaces were cut from the walls of the pits, then the depths and descriptions were checked. Both clod and bulk samples were collected from the pits. The clod samples were collected mainly in the B horizons and transitional horizons to the B. Metal cores were used to collect undisturbed samples from the gravelly profiles. Peds from the subsoils were collected for examination of clay skins and other morphological features (see Appendix A for profile descrip- tions). The bulk samples were collected in plastic bags and prepared * Drought period 36 according to standard procedures described in Soil Survey Investiga- tion Report No. 1 (1972). Physical Analyses Particle Size Analyses Procedures for particle size analyses, by the sieve and pipette methods, and soil moisture retention, are those outlined in Methods of Soil Analysis, edited by C. A. Black (1965). Bulk density was done at Njala University College, Sierra Leone, by the clod method, outlined in Soil Survey Investigation Report No. 1 (1972). Particle Size Analyses, After Fe Oxide Removal Free iron oxides were removed by the dithionite-citrate— bicarbonate method of Mehra and Jackson (1960), after the removal of organic carbon with 30% hydrogen peroxide. The method of Kilmer and Alexander (1949) was employed for the size separations. Chemical Analyses Previously Reported by Odell et a1. (1974) Organic C was determined by the Walkley-Black W€t oxidation method. Cation exchange capacity was by ammonium saturation, with normal ammonium acetate buffered at pH 7. The exchangeable bases were removed by leaching with neutral ammonium acetate. Exchangeable Ca and Mg were measured by EDTA titration, and K and Na by flame photometry. Exchangeable Al was removed by leaching with 1.0N KCl solution and then measured by emission spectroscopy. 37 Iron and Aluminum Extractions Acid ammonium oxalate extraction for Al and Fe: The method of Tamm as modified by McKeague and Day (1966) was used. Four grams of fine earth (oven-dry basis) were shaken continuously in polyethylene bottles (in the dark) for four hours with 200ml of 0.2M ammonium oxalate solution adjusted to pH 3.0 with oxalic acid. Five drops of 0.4% superfloc was added to the mixture after the shaking to aid flocculation before centrifugation at 2,000 rpm (International type SB centrifuge). Determinations of the A1 and Fe in the extractant were by the Aluminon Colorimetry Method of Chenery (1948, cited in Soil Survey Report No. l, 1972) with overnight color development, and by K- thiocyanate colorimetry, by Jackson (1956), respectively. The oxalate in the extractant was destroyed by successive oxidation with 30% hydrogen peroxide and concentrated HNOB. The residue was taken up with 10ml of 1.0N HCl, and then made up to volume with distilled water. This treatment is necessary to prevent oxalate interfering with color development. The free Fe oxides were precipitated with 25% NaOH solution (M. L. Jackson, Advance Chemical Analysis, 1956) before Al was determined in the extractant. High Fe in solution will affect the accuracy of the Aluminon method (Jackson, 1956). Dithionite-citrate-bicarbonate extraction for Fe and Al: The method of Mehra and Jackson (1960), as outlined in Soil Survey Investigation Report No. l (1972) was followed. Fe and A1 in the extractant were determined by Kpthiocyanate colorimetry and the Aluminon method, respectively. The extractant 38 4, HNO3 mixture, and the residue taken up was digested with a H280 with 1.0N HCl. Fe was precipitated from the solution with 25% freshly prepared NaOH solution (Jackson, 1956) before the Al determination. Sodium pyrpphosphate extraction for Fe and Al: The method of Bascomb (1968), as outlined in Soil Survey Investigation Report No. 1, was adopted. Overnight shaking was for 8 hours. The determinations of Fe and Al in solution were by the methods described earlier under ammonium oxalate extraction. Phosphorus Adsorption Phosphorus adsorption studies on fine-earth: Phosphorus adsorption studies of the Langmuir type were made on the surface and subsurface horizons of four soil series, i.e., Makeni, Gbesebu, Masuba and Segbwema, and on the surface and subsurface ironstone gravels of the Njala and Makeni series. For adsorption studies with the <2mm fraction, 50ml of 0.001M CaCl2 solutions containing Ca(H2PO4)2-H20 at concentrations of 0—150ppm P were added to 2gm of fine-earth in 250ml polyethylene bottles. Two drops of toluene were added to each bottle to inhibit microbial growth. The bottles were then shaken in a reciprocating shaker for 24 hours. The room temperature was 26°Cil. The samples were then centrifuged and the concentration of P in the supernatant liquid was determined by the method of Dickman and Bray (1940) as modified by Kurtz (1942). The quantity of P adsorbed was taken to be that lost from the solution during shaking. 39 P adsorption studies on the gravels: Ten grams of ironstone gravel, selected to represent the size distribution present in each horizon, were washed with a mixture of alcohol and water to remove loose soil particles on the surface of the gravel. Fifty milliliters of 0.001M CaCl solution containing Ca(H P °H ' 2 O4)2 O at concentrations 2 2 of 0, 1, 2, 3, 4, and Sppm of P were added to the gravel in 125ml pyrex flasks. Two drops of toluene were added to inhibit microbial growth and the flasks were shaken for 24 to 48 hours in a rotating shaker at a maximum of 150 rpm at room temperature. A rotating shaker was used because initial studies showed that shaking in the reciprocal shaker tended to cause slaking of the ironstone, as a result of possible grinding action. P adsorbed was determined by the method described earlier for the fine-earth. Mineralogical Analysis The total clay fractions were separated by sedimentation (Black, 1965) after free Fe oxides had been removed. Fine clays were sepa— rated by centrifugation, using the method of Tanner and Jackson (1947). The International type SB centrifuge was used. The R/S of this centrifuge was calculated for use in the determination of the time of centrifugation, using the integrated form of Stokes' law as proposed by Svedberg and Nichols, and also given by Steele and Bradfield (1934). Corrections for temperature were also made. X-ray_ana1yses: Samples from the total (<2u) and fine clay (<0.2u) fractions were treated for x-ray diffraction using standard procedures (Grim, 1968). The Mg-glycerol saturated clays on ceramic 40 tiles were dried over CaClZ, and x—rayed using CuKm radiation. Scanning was between 2° and 29° 28 in most cases. Mineral grain counts: Grain counts were conducted on cleaned sand fractions of the 250—100u and 100-50u size range. No heavy mineral separation was made before the counting, which was done with the aid of a polarizing microscope (Black, 1965; Marshall and Jeffries, 1946). A total of 300 to 500 grains were counted per sample. Micromorphology Observations of peds by split and debris studies were conducted with the aid of a binocular micrOSCOpe at various magnifications ranging from 9 to 54X. The main purpose of this was to identify any clay skins on the surfaces of peds from clod samples. Thin sections of both soil and ironstone gravels were prepared by Gary Section Service, Tulsa, Oklahoma. The thin sections were studied under a Spencer polarizing microscope, at magnifications ranging from 20 to 400x, starting with the lowest magnification. The micromorphological characteristics were described and interpreted much as suggested by Brewer (1964). RESULTS AND DI SCUSS ION General Grouping‘of the Soils Table 1 contains the grouping of the soils studied according to physiography, parent material, moisture regime, vegetation and climate. Figure 2 shows the physiographic relationships of the soils in each of the three soil provinces. The soils of the Upland Surfaces may be very old and probably date to the late Tertiary (Odell et al., 1974). They are considered the oldest of the soils studied. The other three groups of soils are comparatively younger, and may be of Pleistocene or Holocene age. The alluvium on the river terraces is recent. These Alluvial soils may be derived, however, from highly weathered materials that have gone through several cycles of weather- ing and soil formation before they were deposited as parent materials for a new cycle of soil formation. In terms of years of development, the Alluvial soils are the youngest on the landscape, although their parent material may be very strongly weathered. Physical Analyses Table 2 contains the results of particle size analyses, with and without the removal of free Fe oxides, and also the percentage clay as estimated by the factor 2.5 x 15 bar moisture content. 41 42 muflcmum Eoum Edapflmmw nm>o HMHMOBME >EMOH mcflw ««>Ham>mum onEfiB amps >wmpcooom chCOE v Eooomlmbm we as b m0 avowanoo any mamumsfis rump CH amen can wuumsv CH 30H oufluoflpocmwo Maozooom swab >umpcooom mcucoe m Eomnmlmmm wmv as A Eoum Esflpflmom Ame mmmOHm pom madam mmoum .N HoewoumE poxuozon acorn: swan zappcooom mnucoe m EOOOmImeN wvHIoH 03 h >o>mao madman: «Hafiz swan snmesooom mcucos m Eomsmuomm we as O can endow dBOcoz swan wumpcooom mcucoe v Eomhmlmmm mm QBIOZZ A aa>aam>mno >uo> ADV coflumucwam wwmmoo ESHOm mEOmm \cmsn >unpcooom wrucoe v Eomnmummm waucoo Q3 A aa>aam>muo mmmq Adv Hmfiuwuoz monotone: Nacmflmlmo moomwusm panama .H mofiuom Heom coflumuooo> «pofluom Aflosccm macaw oocfl>oum Heeumums ucwwom . pronouo cmoev can aeom can >£mmuooflm>£m Hamwcwmm oomcflmuo :OHumuoom> can msemmn musumflOE .Hamwcfimu .Hmfluoumfi pseudo .>Ldmuooflm>cm >3 poflpsum modewoum co>mao ecu mo OCHQDOHU .Hmfifie 43 mausoa NIH uOm pommOAuoumz + Cowmom >cemw CA coflumusp canoe Ann Eoum umme uo>Au >A popooAm on woe mAHom A «a; mAm>mum OCOpmcowH as >ocoAOAwop monumAOE mo coflumHSO oSA I pofluom ucmsouo .1 AA o>mo spcsxmz cmsA >nmpcooom mnucoe m EOOOmImnm Icoo “oz 6 AO>OA >Awmoc saw xo>coo soomooo smog >umccooom mcucoe m Eomhmlomm “DB 0 mo: swan sumncoowm «.«eucoe A acmemummm Am>mA as u asA>sAAm swson Amt chmAmcooAm can woomuuoe AmA>sAA< .v snows: noon >nmpcooom mcucos m Eooomumem mmuo Q3: n meow Aw>mum :OAumucmAm moooo am 03: On ESA>DAAM\ESA> sAEoccom \cmsn >unpcooom mLucOE m EOmemImmm pommuomEH A (SAAOO >EMOA wcflm adv + moomnnoe >nmusnAuB uo>Am momma can mommwmuoom Amfl>sAAOO .m mofluom AAOm coAumuomo> «okoom Aaosccm OQoAm wocA>OHQ AwAuwumE ucmumm propose cmmEv tam AAom pcm >nmmumOAm>£m AAmwcflmm mmmCAmwo (I .II' 1.54711 Aposcflusoov A wanMB Segbwema Baoma Manowa L (a) Pendembu EWW Moa River Gbesebu G (b) Taja River Makeni Makundu Mabole Rwer J (c) Figure 2. Relationship among the soils studied in each soil province with respect to the topography. 1:5 me“: come kn.0 0e.o _.AA A._ uOm ~.. ~.mn o.eA ~.Am x.- sow Lc.: 3.0 v.9A 4.42 a... .m. m.om so 2~amieaa~_-ncsc vv.o Mn : 0n.0 v..o m.- 2.2 now m.~ a.0n c.n. 5.:m 0.4. now 45.: A _ c.12 . :1 \.\. ..~_ 4.04 e; s_.ziausc_-:s. mn.o A~.0 xmuo 0A.0 v.n~ ~.L ;.: m.A s.0~ ~.LL :.am 0..“ Om Lh.: m.s ~.:~ _.~A q._e 5.: L.:: rm m<.suca-sv. mv.o -.0 n~.c -.0 ~.Az ~._ sum u.x~ a.n. ~.e¢ 0.xw Om eo.0 m.n~ 5.x. n..v _.n a._: an mz<.s;~7-c.. mm 0 ms 0 h~ : m~.0 8.0 ~._ Son ~.:~ A... ~.:: m.v~ SO: 25.0 m.c~ 0.5 0.4a 4.8 3.:n he AAmwu um+>sAu A>s~o ON: OO\O eusu >s_o >sAO a..m scam om: assu Asuob >s~U adv :~.n:~ >sAo Ousm scan II~A weasom «now uAsm OA.m 04m sum .sen >usmcoo -xee .x\m o -a o sun A. nave mesa <\m Lagos o o a tan .ton AA xAzn .n:>:eam Ooz 00.x: or xm.~ .. - a mu mA >s~u a po>an¢ ofiaxo on unaucoo Ousumuos nun m. was .>u.zcoo xusn .m_n>Ascs swan subsets; una~0s030ua .3;~v>sa slow .~ oAnsP 46 .2235: do: A: .0... 5:.) .03... be 0.03:; 02.3330 0 A A .AO5OL tquaAzsAc . A50AO .32.: .40 yo :ovc 50.: .ucoIOsuu :00» Ass .zAsoc= acoumcoOA yo: 0 u Oi . 50AO :cxsuv .h.73: Uhfiztuucw c .OAaau OxmA 40.2.5039“ HO: UL u N>0~U .va’néflh UL d H>IAU .VBO.— ..—n0 90 dub?» EChu £9160 Qua: O 50.0 00.0 00.0 50.0 5.05 A.A O 0.A 0.00 0.0, 0.0 0.00 O --- 0.A --- 0-- 0.00 5.10 0.0 --- 550.0000A-0AAO 00.0 00.0 00.0 00.0 0.05 5.A O 0.A 0.A0 0.00 5.0 0.00 O --- A.A --- :-- A.A5 5.05 5.A --u A54AsO0AA-0AO 00.0 00.0 00.0 50.0 0.05 A.A O A.A 0.00 0.0A 0.0 0.00 O -0- 0.A -u- --- 0.05 0.05 A.0 --- 0.2O05un0. 00.0 A0.0 00.0 00.0 A.55 0.A 0.0 A.A 0.00 A.50 A.0 0.50 O --- 0.A --- --- 5.50 0.55 0.0 --- 00.00A0-00O 05.0 00.0 00.0 50.0 0.05 ..0 O 0.00 0.00 0.5A 0.55 O --- nu- -u- 5.00 5.05 0.0 ..u 5A<.suo0-0mO 05.0 00.0 50.0 00.0 A.05 0.0 O 0.00 A.50 0.0A 5.55 O -- -t- -n- A.00 0.05 0.5 -c- AA0AO.nnummMu .mummo o5: uu\0 0500 >0AO 50AO 0AA0 0:00 05: 0.00 A0005 >0AO 05v 05.-05 50AO 0AA0 0:00 115. 005500 AAom 0AAO 0AAO o z 0 z .000 >0A0000 uxoe <\0 0 In a 0 500 0A -005 ocAO <\0 A0000 0 0 0 .00 .000 0A 0A0: .00>oaoz 002 00000 on 00.5 0 mu 3 >040 a vo>oio¢ 00:0 0.0 I‘ll I!" a 1."... I. la.“ ." «MI-i"lri Iii-'1‘ 0.1 I a! 7-71 ‘40 I 4.l~..l.‘|s 1 V 1}..|. I. 1471‘; 1.--..‘13 I AvoacAucooO 5 oAnas 47 Data for the particle size analyses without the removal of free Fe oxide were obtained from Odell et a1. (1974). A few of the samples were analyzed without the removal of free Fe oxide in the present study, to compare with the values obtained by Odell et al. (1974). These are listed in parentheses in Table 2, in the column for percent clay without Fe oxide removal. The difference between the values is usually not more than 3% in the subsoil. The greatest variation was in the surface soil samples. In general, the percentage clay values, when free Fe oxide is removed before particle size analyses, are higher than those obtained when the free Fe oxide is not removed. The percentage clay values, obtained with free Fe oxide removed, are also commonly higher than the clay values estimated by the factor 2.5 x 15 bar moisture content. The exceptions to this latter relationship are the Baoma series, the subsoil of the Segbwema series and the A horizon of the Makundu series. In these situations the percentage clay estimated by the factor 2.5 x 15 bar moisture content is higher. The reason(s) for the deviation in the Baoma and Segbwema series is not fully known, and may need further investigation. One possibility is the nature of the Fe oxide present in these soils, and the ease of its removal by the dithionite-citrate-bicarbonate method. The high percentage organic matter in the surface horizon of the Makundu soil may be responsible for the difference there. The increase in percent clay of the pedons, as a result of the treatment, varies from -2.4 to +23.7%. There is no significant relationship between these values and the general groupings of the soils. However, there is significant correlation (0.89) at the 1% 48 level between percent clay (Fe removed) and organic C for surface horizons. (The correlation value is 0.26 with the percent clay for the whole soil correlated with organic matter.) This value is significant at the 20% level. In general, the values of percent sand and silt decrease in most of the soils when free Fe oxides are removed. However, the total loss in percent silt and sand is not reflected in the per- centage increase in the clay fraction in most cases. Some of the Fe oxides in the form of discrete particles may have been removed from the soil. Desphande et a1. (1968), in their studies on the changes in soil properties associated with the removal of Fe and Al oxides in soils, which included some laterite and lateritic soils, reported a decrease in the sand and silt fractions after the Fe oxide has been removed by the dithionite—citrate-bicarbonate method. They recorded not an increase in percent clay content, but a decrease. They suggested that the loss could be due to the iron oxides not present as cementing agents, but as discrete particles or that the amount of Fe initially present in the clay fraction and removed by the treatment was sufficiently large to offset any additions of clay resulting from removal of cements in the silt or sand fractions. In the present study, the increase in percent clay, as a result of the removal of the free Fe oxides, suggests that some of the Fe oxides are in the soil as cementing agents binding clay particles together. However, the lack of correlation between the decrease in percent sand and silt with that of the increase in percent clay may indicate that some of the Fe oxide is present in discrete clay size particles. 49 In some of the pedons slight changes in the soil textures are observed as a result of the removal of free Fe oxides, towards a clayey texture. Also, changes in the distribution of the percent clay are observed in some profiles after the removal of Fe oxide. In the soils of the Steep Hills and Slopes (Segbwema and Timbo series) (Table l), a decrease in percent clay from the surface downward is observed in the Segbwema series, and a zone of clay accumulation is more clearly observed in the Timbo series with the removal of free Fe oxide. In the soils of the Upland Surfaces of Highly Weathered Material (Baoma, Manowa, Njala and Makeni soils), there are gradual increases of clay with depth after the removal of Fe oxides. The Manowa profile also shows this gradual increase with depth, before removal of Fe oxides. In the soils of the Colluvial Footslopes and Upper River Tribu- tary Terraces, a zone of clay accumulation is more clearly observed with the removal of Fe oxide in the Masuba series, but in the Pendembu series only a gradual increase in clay with depth is evident after removal of Fe oxide. A gradual increase in clay with depth, both before and after the removal of Fe oxides, is evident in the soils of the Alluvial Terraces and Floodplains. Only incipient illuvial horizons are observed in the Makundu and Moa series after the removal of Fe oxides, as opposed to an increase in percent clay with depth before removal. In the Gbesebu series the incipient nature of the illuvial horizon is recognized both before and after the removal of Fe oxides. 50 From the above discussion of the soils studied, the removal of the free Fe oxides, before particle size analyses, seems to aid in under- standing the genetic processes that are taking place within these profiles. Total Clay* The lowest percent of total clay is observed (Table 2) in the soils of the Colluvial Footslopes and Upper River Tributary Terraces, i.e., the Pendembu and Masuba series. They are also lowest in total Fe oxides. They are closely followed by soils of the Steep Hills and Slopes, the Segbwema and Timbo series. The soils of both the Upland Surfaces of Highly Weathered Material (Baoma, Manowa, Njala and Makeni series) and of the Alluvial Terraces and Floodplains (Moa, Gbesebu and Makundu series) have greater than 50% total clay throughout their sola. The highest percent clay is in the B horizons of the Makundu and Gbesebu series. Clay accumulation in the subsurface horizon is a basic criterion for the identification of an argillic horizon, a major characteristic for Ultisols and Alfisols (Soil Taxonomy, 1975). The bulk of the accumulated clay in such soils is thought to have been eluviated from the overlying horizons. Recrystallization from solution and weathering of primary minerals in place account for the remainder of the accumulated clay in the subsurface horizons. An increase of 3% more clay in the illuvial horizon if the total clay is <15%, or B/A ratio of 1.2 if the total clay is >15% but <40%, or an increase of 8% clay in the illuvial horizon if the it Discussion is limited to clay after Fe has been removed. 51 total clay is >40%, constitute an argillic horizon. The main requirement is that there is evidence to show that the clay has moved. This is usually recognized by the presence of clay skins on the surfaces of natural peds. Clay skins are usually difficult to observe in the field in soils of the tropics high in free Fe oxides. Thin section studies are therefore commonly necessary to identify the oriented clays. Thin section studies on the eleven pedons in this study are discussed in the "Micromorphology" section. The B/C ratio for the identification of argillic horizons has been suggested by Smith and Wilding (1972) as a better criterion for identifying the argillic horizons for soils with a lithologic dis— continuity in the upper part of the solum. The B/A ratio for most of the soils studied here are given in Table 2. The B/C ratio for Segbwema series is given in parentheses instead. The B/A ratio for the Segbwema soils is less than 1 through- out the profile. This is unlike the Timbo soil in which only the B22 horizon has a B/A ratio of less than 1. Both of these soils are developed from residuum of the parent rock. The Segbwema soil is on a Steep SlOpe of 42% as compared to 6% for the Timbo series. This will affect the depth of weathering in the Segbwema soil, with the deepest horizons being the least weathered. The B/A ratio in this situation may not be a good index for the recognition of an argillic horizon. The B/C ratio for this profile indicates the presence of an argillic horizon. This has been supported by thin section studies (see section on "Micromorphology"). In the soils of Upland Surfaces of Highly Weathered Material, the B/A ratio may not be useful for the identification of an argillic 52 horizon, as they all have a total clay content of >40%. The require- ment for an increase of 8% or more clay is diagnostic for these soils, if it occurs within 30cm depth. Both the Njala and Baoma profiles meet this requirement. The Makeni and Manowa soils have values of 7.9 and 7.2, respectively, which are only slightly less than this requirement. In the soils of the Colluvial Footslopes and Upper River Tribu- tary Terraces, the B/A ratio for the Masuba soil is greater than 1.2 and thus meets the requirement of an argillic horizon. The ratio is less than 1.2 for the Pendembu series, within 30cm depth, but shows a continual increase with depth. The soils of the Alluvial Terraces and Floodplains do not meet any of the requirements for the argillic horizon discussed earlier. Fine Clay (<0.2u) Data for the fine clays are also included in Table 2. In general, the percent of fine clay is higher than that of coarse clay (Zn-0.2u) for most of the profiles. Exceptions are the Gbesebu series (Alluvial Terraces and Flood- plains) and the lowest horizons of the Njala, Makeni (soils of Upland Surfaces) and Segbwema series (soils of Steep Hills and Slopes). The higher values of the fine clays may be indicative of the intense weathering in these soils. The ratio of fine clay to total clay is also a useful diagnostic criterion for an argillic horizon (Soil Survey Staff, 1975). A ratio of fine clay to total clay in the argillic horizon greater by one- third or more than the overlying eluvial horizon, and/or the underlying 53 horizon, is required for soils with less than 15% total clay and 15-40% total clay. If the total clay is >40%, 8% more fine clay is required in the argillic horizon. The Segbwema soil has a ratio of fine clay to total clay in the B horizon that is greater than one—third of that for the over- lying horizon, which further confirms the presence of an argillic horizon. The Timbo series does not meet any of these requirements of an argillic horizon. The fine clay to total clay ratio for the soils of the Upland Surfaces of Highly Weathered Material show a clear decrease with depth, and meet none of the above criteria for an argillic horizon. In the soils of the Colluvial Footslopes and Upper River Terraces, the Pendembu series shows a general decrease with depth in the fine clay/total clay ratio. (Fine clay is not available for the Masuba profile.) This trend is similar to that of the soils of the Upland Surfaces of Highly Weathered Material. In the Gbesebu series the fine clay/total clay ratio is constant thrOughout the profile. A general increase with depth in the ratio is observed for the Moa series. Silt/Clay Ratio The values for the silt/clay ratios when Fe oxide is removed and not removed from the samples are also given in Table 2. The latter value is included only for comparison. Discussion will be limited to values obtained when the percent clay is that from which Fe oxide is removed. In general, the ratio increases with depth in the Segbwema and Timbo series of Steep Hills and Slopes. The ratio shows a general trend of slight decrease with depth in the Baoma, Manowa, Njala and 54 Makeni series, all of which are soils of the Upland Surface of Highly Weathered Material. The Masuba and Pendembu soils, of the Colluvial Footslopes and Upper River Terraces, also show a gradual decrease with depth in the silt/clay ratio. The silt/clay values for the soils of Alluvial Terraces and Floodplains are irregularly distributed in the Gbesebu and Makundu profiles, and show a slight decrease with depth in the Moa profile. van Wambeke (1962) suggests the use of silt/clay ratios for distinguishing tropical soils according to the age of their parent material, with certain limitations. These limitations include: (a) that the soil should be well drained, (b) that the clay fraction is dominated by kaolinite, and (c) that the textural B horizon should not be much richer in clay than the overlying and underlying horizons. He observed that tropical soil materials older than end- Tertiary have silt/clay ratios around 0.10, and that the critical value of 0.15 proved to correspond with a sharp distinction in their ages. Other researchers such as D'Hoore (1954) and Ashaye (1969) have also noted the usefulness of this ratio. Low silt/clay ratios have been suggested by these authors to indicate truly ferrallitic soils, or soils that have undergone ferrallitic pedogenesis. In the soils studied, the silt/clay ratio decreases with removal of free Fe oxides in mechanical analysis, and none of the soils have ratios below 0.15. The silt/clay ratios are the lowest through- out the profile in soils of the Upland Surfaces of Highly Weathered Material. This suggests a more advanced stage of ferrallitic pedo- genesis in these profiles (Ashaye, 1969). 55 Ratios for the other soils are comparatively higher with the exception of the Moa and the upper horizons of the Timbo series. The highest ratio is obtained for the Segbwema soils, followed by the Gbesebu, Makundu and Masuba series. The higher values are indicative of less ferrallitic weathering, suggesting younger soil profiles. This concept may not be true for the Alluvial Terrace and Floodplain soils, as they generally have high silt contents in their parent materials. On the basis of the above discussion, the Segbwema series is the youngest of the soils studied. Bulk Density_and 15 Bar Moisture The bulk density and 15 bar moisture content values are shown in Table 2. In general, the bulk density increases with depth in all the profiles and tends to be constant in the lower B horizons in most of the soils. The lack of bulk density data for the upper horizons of the Manowa and Njala profiles is due to difficulties in getting either clod or core samples from these very gravelly horizons. Bulk density values reported for the gravelly soils may be subject to error, as a result of the difficulties in making the measurement with available methods. The bulk density values range from 0.8 to 1.6 g/cc. The highest values are obtained for soils of the Upland Surfaces of Highly Weathered Material. The lowest values are obtained in soils of the Alluvial Terraces and Floodplains. The very low values in the uppermost horizons of these soils may be attributed to high organic matter content and high percent porosity. 56 The 15 bar moisture content in most cases tends to follow the same distribution pattern as the clay content.* Correlation analyses between percent clay and 15 bar moisture content gave a correlation coefficient of 0.93 (significant at the 1% level). The same value of 0.93 was obtained when total clay was taken as the sum of percent clay when Fe oxide has been removed plus percent Fe203d** (Soil Survey Staff, 1975). Values of the ratio of 15 bar H20/% clay and 15 bar H20/% clay + % Fe203d are given in Table 2. In some cases the values of the ratio are higher in the surface than in the subsoil. The averages of the ratios are 0.36 and 0.33 for 15 bar H20/% clay and 15 bar H20/% clay + % Fe203d, respectively. In Soil Taxonomy (1975) the ratio of 15 atmosphere moisture content of clay has been suggested as a possible index for estimating percent clay in soils that have problems of clay dispersion. A value of 0.4 or less has been reported as the more common in oxic horizons when the percent total clay is taken to be the sum of percent clay with Fe removed plus percent Fe 0 d. In the soils studied, the ratios 2 3 15 bar HZO/clay and 15 bar HZO/clay + Fe203d are less than 0.4 for most of the profiles, with the exception of the Segbwema series, which has values >0.4 for the former. Ratios of less than 0.4 are obtained for soils with known cambic horizons, e.g., the Gbesebu series, average 0.39 or 0.36 for 15 bar HZO/clay and 15 bar HZO/clay + Fe203d, respectively. * Clay, when Fe oxide has been removed. ** Fe 0 d = Fe oxide extracted by sodium bicarbonate-citrate- dithionite. 57 Since percent clay increases with the removal of free Fe oxide, and the values obtained are also usually higher (in 28 of 41 com- parisons) than those obtained by 2.50 x 15 bar H 0, it would seem 2 more logical to use 2.78 instead of 2.50 x 15 bar H20 to estimate the clay contents of these soils after iron removal. The value of the sum of percent clay after Fe removed + Fe O 2 3d would require a factor of 3.0 x 15 bar H20 to estimate clay content. Chemical Analyses The values of percent organic C, pH, C.E.C., percent base saturation, exchangeable cations and some ratios of these values to the clay contents (in Table 2) are given in Table 3. Organic Carbon In general, the percent organic C decreases with depth. The values range from l.25-5.06 in the surface horizons, with the Gbesebu and Makundu soils having the highest percentage in their surface horizons. Comparatively higher percentages of organic C are also found in the lower horizons of these profiles. The presence of high percent organic C in the subsoil of Makeni series may be the result of dead roots or animal activity. Within each of the four groups of profiles in Table l the per- cent organic C in the surface horizon increases with the moisture regime. The Masuba profile of the Colluvial Footslopes and Upper Terraces is the only exception. 6m.mavh.w6 acm.m «6®.m m.66 c.vm 66.6 60.0 mc.o 06.0 mo.o o.m 00.0 m.a O.v m.m 06.0 60 553 16.6666.66 ..6.6 .66.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 wmmm 'I' I'- ll' III 'II. III II- I" III .'I .I'I III- '|l| I.II' -‘- I.-- . m 16.6.6.66 6.6 6.6 6.66 6.6 66.6 66.6 66.6 66.6 66.6 6.66 66.66 6.6 6.6 6.6 66.6 66 mam3nmmm AO.mvm.m «.66.». c¢©.N «66.x sang. 6H.C $0.0 M0.0 $5.0 $0.0 m.mN wwom 0.0 ©.v Nom MH..—. NNm 16.6666.66 ..6.6 ..6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 666 16.6.6.6 6.66 6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.66 6.6 6.6 6.6 66.6 66 66666: 16.6666.66 ..6.6 ..6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 666 16.6616.66 ..6.6 ..6.6 6.6.66 ..6.66 66.6 66.6 66.6 66.6 6 6 66.6 6 6 6 6 6 6 66.6 666 16.6666.66 ..6.6 ..6.6 6.66 6.66 66.6 66.6 66.6 66.6 6 6 66.6 6 6 6 6 6 6 66.6 66 16.6666.66 6.66 6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6 6 66.66 6 6 6 6 6 6 66.6 66 66662 16.6666.66 ..6.6 .66.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 666 --- --- --- --- --- --- --- -u- -u- --- --- -u- --- u.. -u- -u- 666 16.6666.66 ..6.6 ..6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 66 16.6666.66 6.66 6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.66 6.6 6.6 6.6 66.6 66 630:6: III III III III III II! III III III III III III III III III III MMQHH 16.666.66 ..6.6 ..6.6 ..6.6 ..6.6 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 66666 16,666.66 ..6.6 6.6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 66666 16.6.6.66 6.6.6 ..6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 66 16.6666.66 6.6 6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.66 6.6 6.6 6.6 66.6 66 mEomm 66 + 66666 zzrwm66o y- 66666 swmwww . 6666p 66 62 x 6: 6o .66. 6666 :66 66x 06: .o.o 666666 1:66:666x6 6c 2:6 666 x 6< 666 x 66 666 x 666 x 6666 6666\62 6066 .6666 .66 .66 .6 6666 cc6 x 6< 00:6;6xu + mc36bou + mco6ucu .U.m.b .U.m.b mco6umu mfinmmgcmzoxm6 Imu=umw \wE Acc6umuzumm 6< do oocczoxm catacuxm mmmm .U.m.U we Ezm mo 55m 6 ......,‘s,..,.h...( ..u_.u; uuhnnuun.unu.....¢ ... ... .. ...H.. H.Qflaq“44H.““My“:fl‘nflahruillllhlqddthfl4§.§dtufltl".uaaV3“dhdflad."drg|.tli.c i 66:06666 66occqcmzsxv tam mm 6 ..u.m.u .zd ..u.o my U6663u6 66606 wzu mo mw6uuom06m 600656cu .6 66269 559 III III III III III III III III U .o.mmv~.mh ¢6N.m ¢ah.~ cam.m~ 66m.66 66.6 mc.o no.6 6N.o co.o 0.0 om.© H.H o.v H.m av.o mm 66 III III III III III III III III III III III III III III III III m 66.66.6.66 ..6.6 ..6.6 6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 666 66.6666.66 6.6 6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.66 6.6 6.6 6.6 66.6 66 do: 66.6666.66 ..6.6 ..6.6 ..6.66 ..6.6 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 6666 16.6666.66 ..6.6 ..6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 6666 16.66.6.66 6.6 6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 64 62:60: --- -u- -u- --- --- -u- --- --- --- --- u.. u.. -u- nu- In- In- 66 --- -u- --- --- --- --- --- -u- -n- -u- -u- --- .u. .u. I»- -u- 666 A6.VMVH.bm «60.? 6.60.? «6Q.- 6.0.0 NN.H 26.0 mO.C Mo.c 00.0 A.A. Om.m w.o m.m h.v Nm.o ”mm -u- --- -u- --- --- --- --- -s- --: -n- --- --- -u- 1-- uu- In- 66 16.6666.66 6.66 6.6 6.66 .6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 66 66.6666.66 6.66 6.66 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 64 6.8609696 16.666.66 ..6.6 ..6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 “666 16,666.66 ..6.6 ..6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 666 66.666.66 ..6.6 ..6.6 ..6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 64 66.66.6.66 ..6.6 ..6.6 6.66 ..6.66 66.6 66.6 66.6 66.6 66.6 6.6 66.6 6.6 6.6 6.6 66.6 66< 16.6666.66 6.6 6.6 6.66 6.66 66.6 66.6 66.6 66.6 66.6 6.66 66.6 6.6 6.6 6.6 66.6 666 06566 66 + 66666 1iI6>66o 6666b amammw Iwwmww 66 62 .6 6: 66 6666 6606 :66 6oz o6: .o.o 666666 unaccoxm 66 5:6 666 x 6< 666 x 66 666 x 666 x 6666 6666\65 :66» .6666 .66 .66 .6 6606 006 x 6d mmcozoxm + mco6wmu + 6:66660 .L.m.u .U.m.u 6206660 ¢HQmmUcw£uxmc IMMSumm \oE Accwuwuaucm 6< av 63:6:5xm wscmcoxm mmwm .U.m.U 63 65% mo Esm 6 6666666666. 6 66666 6C) .uczo6c 06 uw>ocu6z3 .O .U.m.U 006 x 6< mmcmcoxw N .vsma ..Hm um Hava EOHu :wxmu Mann 0 >n vmum6soamo co6umusumm 64 = 603 m6 x 0.6 no cm>oemu uo: .vm>oewu .couwuon 06x0 uOu Owumocvmwo O a u A . mOmwm u ~>m6u mONwm u 6>MHU C .0.6600.00 ..0.6 ..6.6 ..0.0 ..0.0 00.6 06.0 60.0 66.0 66.0 0.66 06.6 6.6 6.0 .0 00.0 660 66.000.60 ..6.6 ..0.6 ..6.66 ..0.66 60.0 00.0 00.0 06.0 66.0 0.0 00.0 0.6 6.0 .0 06.0 660 60.66.0.06 ..6.6 ..0.6 ..0.06 ..0.66 00.6 00.0 00.0 66.0 66.0 0.0 06.0 0.0 6.0 .0 00.6 60 66.6606.06 ..0.6 ..0.6 6.66 0.66 00.6 66.0 00.0 66.0 66.0 0.6 66.06 0.0 6.0 .0 60.6 00 60.0.6.60 ..0.6 ..6.0 6.66 0.06 06.6 06.0 06.0 60.0 60.0 6.0 00.06 6.0 6.0 .0 60.6 660 66.600.6 0.0 0.0 0.66 6.06 60.0 60.0 60.0 06.6 06.6 0.66 60.66 0.0 0.0 .0 00.0 664 sucsxuz 66.0606.66 ..6.6 ..0.6 ..0.66 ..0.66 06.6 00.0 60.0 00.0 06.0 6.6 00.0 6.6 6.6 .0 00.0 660 66.06.6.00 ..0.6 ..0.6 ..0.06 ..0.66 00.6 00.0 60.0 66.0 06.0 0.0 60.0 0.6 6.6 .0 60.0 0660 66.0606.00 ..6.6 ..6.6 ..0.66 ..0.66 06.6 00.: 60.0 06.0 00.0 6.6 06.0 0.6 0.6 .0 00.6 0660 60.0606.60 ..0.6 ..0.6 6.06 6.06 00.6 00.0 60.0 06.0 66.0 6.6 00.66 0.0 0.6 .0 60.6 60 .0.6606.60 0.0 6.0 0.06 0.66 00.6 00.0 60.0 66.0 06.0 6.6 60.06 0.0 0.6 .0 60.0 60 memmmm 6< + mwmmm ~>060 6>160 m>c 6> w m . . . . 60 060 60 02 x 2 no 600. 6600 000 60x 0 z u o 006600 mocmcoxm 00 5:0 .mmwumxflm .mm6 x 60 006 x 006 x 6600 0006\02 :06» .0006 .00 .00 .o 6600 006 x 6‘ omcmcoxm + :oflumu + mzo6umu .U.m.u .U.m.U mco6umu m6nmmccwnuxm. Imusumm \mE ”cowumuSumw 64 my cozmzuxm wocmzLxm wmmm .U.m.U 0o 2:0 60 5:0 0 .vmscfiucouv m manna 61 Hg The soils are generally low in pH, the range being 4.1 to 5.5 (1:1 H20). The values generally increase with depth in the profile. The ApH* values are positive, but low (16 but would reduce the Makundu 21 3 B1 to a value of <15.0. The soil that did not meet this requirement of an oxic horizon in its uppermost B horizon is the Segbwema series. Both the Gbesebu and Masuba series are known to have cambic horizons (Odell et al., 1974). Incipient argillans were also observed in the B horizon of the Masuba series in this study (see.discussion on "Micromorphology"). * Clayz- Fe not removed or 2.50 x 15 bar H greater. 2O, whichever is 63 The value of percent clay, as determined by 2-5 x 15 bar moisture contents, or percent clay when Fe oxide is not removed (whichever is greater), was used by Odell et a1. (1974) for the C.E.C./clay ratio. All the soils with the exception of the Moa and Segbwema series have this criterion of an oxic horizon (<16 me C.E.C./ lOOg clay) in their uppermost B horizons. van Wambeke (1967) pro— posed a critical value of 12 me/lOOg clay, instead of 16, for older soils of the tropics, as this value seems to form a class which is characteristic for old stable landscapes in the tropics. He observed correlations between this criterion and soil structure, color, weatherable mineral content, the presence or absence of pressure faces, and clay skins on ped surfaces. Could perhaps a group of soils studied here be better distin- guished if this critical value is used? Using the value 12 instead of 16 with these pedons, the upper B horizon of Baoma series will be non—oxic according to this criterion. Similarly, the upper B horizons of the soils of the Alluvial Terraces and Floodplains (Gbesebu, Makundu and Moa series) will be non-oxic according to this limit. The Gbesebu and Makundu are eliminated from Oxisols by lack of an oxic horizon in the control section (25-100cm). On balance, the 16 me/lOOg of clay seems to give a clearer split of the oxic and non-oxic soils in this study if other criteria can be used for other important separations needed. Percent Base Saturation The percent base saturation of the eleven profiles is generally low, all samples being less than 36%. In most of the soils studied, such as the Njala, Makeni, Segbwema, Timbo, Masuba, Moa and Makundu 64 series, the percent base saturation is higher at the surface, then decreases, followed by increases with depth. The other soils (Baoma, Manowa, Pendembu and Gbesebu series) all show increases in percent base saturation with depth. The general decrease with depth of percent base saturation, below the immediate surface, in the Segbwema and Timbo soils may be related in part to the topo— graphic position of these profiles (Figure 2) and the circulation of bases to the surface by organisms. On this topographic position runoff is the predominant form of water movement. Infiltration is minimal. The increase of percent base saturation with depth may be in part a result of leaching to lower horizons of the exchangeable cations or continued release of bases from the less weathered deeper material. The higher percent base saturation in the surface horizons of the Makundu, Makeni, Timbo and Masuba soils (all from soil province J) is the result of unusually high exchangeable Ca found in the surface horizon of these soils. The source is not certain, but may be from the organic additions to the land surface and a relatively long dry season with some dust additions. Finally, as the less leached parent material is approached with depth, the percent base saturation rises as the C.E.C. decreases, and with leaching to lower horizons of exchangeable cations or less leaching with depth and release of more bases from less leached subsoil layers. Exchangeable Cations Generally, the soils are low in exchangeable bases. The Makeni and Makundu soils have the highest amount of exchangeable Ca and Mg in their uppermost horizons. All the other profiles have low 65 concentrations of these nutrients. These soils also have compara- tively low exchangeable Al. The exchangeable Al is highest in the Baoma, Manowa, Njala, Pendembu, Moa and Gbesebu soils. The_percent A1 saturation is highest in the Manowa, Njala, Pendembu, Masuba and Gbesebu series. The lowest values are in the subsoil of the Baoma series and in the Makeni profile (range = 5.0 in Makeni surface to 93.9 in Pendembu surface horizon). Decreases of percent Al saturation with depth were observed in the Baoma, Manowa, Pendembu and Gbesebu series, while Njala, Makeni, Segbwema, Timbo, Masuba, Moa and Makundu soils showed zones of high percent A1 saturation in their profiles. The high percent Al saturation of these soils poses toxicity problems, which will affect crOp production. Lime application is therefore necessary in these soils to reduce percent Al saturation and increase pH. The ratio of the sum of the exchangeable cations plus exchangeable A1 to percent clay (clay as determined by pipette method or by 2.50 x 15 bar moisture, whichever is higher) is one of the criteria used in identifying an oxic horizon. A critical value of 10 me/lOOg clay or less is required. In the soils studied, this ratio, if determined using percent clay with Fe oxide removed (Table 3), is less than 10 me/lOOg soils in the upper and lower B horizons of all the soils. A few surface horizons (Table 3) have values >10 me/lOOg clay, particularly if cal- culated with percent clay without Fe removal or as :2,5 x.lS bar moisture content. This means, then, that all the soils have this characteristic of oxic horizons in their B horizons. The Segbwema 66 and Masuba or Gbesebu series have been shown to have argillic and cambic B horizons, respectively (see section on "Micromorphology"). On the basis of the criteria Sum of Exchangeable Cations + Al x 100 _ <10 and C.E.C. x 100 = E 6 % Clay % Clay 1 for oxic horizons, the latter eliminates the Segbwema from the Oxisols. Extractable Iron and Aluminum Oxides (Dithionite, Ammonium Oxalate and Na-Pyrophosphate, Fe and A1) Fe and Al oxides are among the major components in tropical soils. Both physical and chemical behavior of these soils are affected in some ways by these components. Aggregate stability, the fixation of P (Udo and Uzu, 1972) and adsorption surface prOperties are among some of the soil properties affected by these oxides. These oxides are known to exist as amorphous and crystalline inorganic forms (McKeague and Day, 1966) and as organic-oxide complexes (McKeague et al., 1971). The acid ammonium oxalate method of Tamm as modified by McKeague and Day (1966) is known to extract mainly amorphous Fe oxide. The dithionite-citrate-bicarbonate method of Mehra and Jackson (1960) is known to extract total free Fe oxides and also some organic complex forms. The Na-pyrophosphate extraction has been shown to be specific and extracts Fe and Al organic complexes (McKeague et al., 1971). Dithionite-Citrate-Bicarbonate Extractable Fe Oxide This method generally extracted more Fe oxides than did acid ammonium oxalate (Table 4). This compares favorably with observations €57 no.0 no.0 v0.0 no.0 no.0 no.0 6.0.0 6.0.0 Orin 00.0 Am.ovmm.o mm.v no.0 ~N.o :30 mo.o No.0 mo; HU 60.0 60.0 00.0 00.0 00.0 00.0 00.0 06.0 00.0 00.0 666.0006.0 06.6 60.0 06.0 00.0 00.0 06.6 06.6 0660 -u- 60.0 --- 00.0 -u- 00.0 --- 00.0 00.0 60.0 -uu -u- 60.0 00.0 60.6 00.0 60.6 60.6 0660 00.0 60.0 60.0 00.0 60.0 00.0 06.0 00.0 60.0 00.0 .00.0.66.0 00.0 06.0 00.0 00.6 60.0 00.6 06.6 60 usoznmmm 60.0 60.0 00.0 60.0 60.0 60.0 06.0 60.0 60.6 66.0 +l06.0.00.0 00.06 66.0 00.6 60.0 60.0 60.0 06.0 660 60.0 60.0 60.0 00.0 60.0 60.0 00.0 60.0 60.6 66.0 .66.6.00.6 60.06 66.6 60.6 06.0 66.6 66.0 00.0 660 60.0 60.0 00.0 00.0 60.0 60.0 06.0 00.0 06.6 66.0 +.00.0.00.0 60.0 00.6 00.6 00.0 00.6 66.0 06.6 60 60060: 60.0 60.0 60.0 00.0 60.0 60.0 00.0 00.0 00.6 60.0 +.66.6.60.6 00.0 60.0 66.0 66.0 06.0 06.0 06.0 66m 00.0 00.0 60.0 00.0 60.0 60.0 60.0 60.0 06.6 06.0 +.00.6.00.6 00.0 06.6 00.0 00.0 06.6 00.0 00.0 66m 00.0 60.0 60.0 00.0 60.0 60.0 60.0 00.0 06.6 06.0 +.60.6.06.6 00.0 66.6 00.0 06.0 06.6 00.0 00.6 64 60.0 60.0 00.0 00.0 60.0 60.0 60.0 00.0 00.6 06.0 +.60.0.00.0 00.0 66.6 66.0 06.0 60.6 00.0 00.6 64 «mmflm 60.0 60.0 06.0 00.0 000.0 000.0 06.0 06.0 00.0 00.0 00.6 06.66 --- 06.6 60.6 06.6 06.0 00.0 660 00.0 60.0 00.0 00.0 60.0 60.0 66.0 00.0 06.0 60.0 --- 00.0 --- 66.6 00.6 00.6 06.0 00.6 660 00.0 00.0 00.0 00.0 60.0 60.0 06.0 00.0 00.0 00.0 60.6 -u- --- 06.6 00.6 00.6 06.0 00.6 66 00.0 00.0 06.0 00.0 60.0 60.0 66.0 00.0 60.0 66.0 00.0 06.0 -I- 06.6 60.6 60.6 60.0 00.6 6: USCG: --- 60.0 -u- 00.0 I)- 60.0 1-- 66.0 00.0 66.0 -1- -0: .u. 66.0 60.6 06.0 06.6 60.0 66066 60.0 60.0 00.0 60.0 60.0 60.0 06.0 06.0 60.0 06.0 06.6 00.06 nu- 66.0 00.6 66.0 06.6 00.0 66066 60.0 60.0 00.0 00.0 60.0 60.0 66.0 60.0 00.0 06.0 60.0 66.06 -u- 00.0 06.6 00.0 06.6 00.0 66066 60.0 60.0 00.0 00.0 60.0 60.0 66.0 66.0 00.0 06.0 60.6 66.66 --- 06.6 00.6 60.6 60.6 00.0 660 00.0 00.0 06.0 00.0 00.0 00.0 06.0 66.0 00.0 06.0 00.6 00.06 -u- 00.6 66.6 00.6 00.6 00.0 60 «loan 16M060 60060 60060 6>u6u m>mwm 60060 ~>0~u 6>mao uno~6< vnomwm .U.O momma ouozdmozm mama uumco Ouonmmonn 0906 ouuco noduom 060600 06mwwm .mm6m6mm mm6m6mm mmpm 00 xo6o6mm m6m6mm m60 on mummmma mm6m6mu Immmmmmmmm 60006 .0660.02 -000 -60000 .0660.02 -000 -60060 6600 0060600 nono~om nouxanaz|om co no-< noNoh U oficnvho 0:0 >060 ucwouwm saw) mo6uau 660:» can moofixo 6‘ can on uHAGuuuuuxo acuuuum .v canah 6&3 -- 60.0 -- 00.0 -- 60.0 -u- 06.0 06.0 66.0 -- -- -- 00.0 60.0 -- 00H0 00H0 60 600.0 60.0 00.0 00.0 60.0 60.0 66.0 00.0 66.0 66.0 -- 00.0 -- 06.0 60.0 -- 00.0 06.0 660 -- 60.0 -- 00.0 -- 60.0 -- 00.0 06.0 -06.0 -- -- -- 66 0 60.0 -- 60.0 00.0 660 60.0 60.0 00.0 60.0 60.0 60.0 00.0 00.0 00.0 06.0 -- 06.0 -- 66 0 06.0 -- 00.0 60.0 60 60.0 60.0 00.0 60.0 60.0 60.0 06.0 00.0 60.0 06.0 -- 06.6 -- 60 0 60 0 u- 60 0 06 0 «mm o o u c o a "I. u a “N 600 0 600 0 60.0 60.0 60.0 60.0 60.0 60.0 06.0 06 0 -- 06 6 u- 00 0 60 0 06.0 60.0 0660 60.0 600.0 60.0 60.0 60.0 60.0 60.0 60.0 06.0 06.0 -- 66.6 -- 00.0 66 0 -- MM.” MM.” 06“ 60.0 60.0 60.0 6 . . 60.0 60.0 60.0 60.0 06.0 -- 60.6 -- 66.0 06.0 -- 0 0 60 0 02:00: -- 60.0 -- 00.0 -- 600.0 -- 00.0 06.0 00.0 -- -- -- 00.0 06.0 -- 66H0 60H6 6M0 -- 60.0 -- 00.0 -- 000.0 -- 00.0 06.0 60.0 -- -- -- 66.0 00 0 -- 06.0 60.6 660 60.0 60.0 00.0 00.0 60.0 000.0 00.0 00.0 60.0 00.0 -- 06.6 -- 66.0 00 0 -- 06.0 66.6 660 -- 60.0 -- 00.0 -- 60.0 -- 00.0 60.0 66.0 -- -- -- 06.0 60.0 -- 06.0 00.6 60 60.0 60.0 00.0 00.0 60.0 60.0 00.0 00.0 00.0 06.0 -- 00.6 -- 06.0 06.0 -- 06 0 MM M 6“ 60.0 60.0 . . . . . . 6. 06.0 -- 60.6 -- 60.0 00.0 -- 06.0 . 00 0 00 0 60 0 60 0 00 0 00 0 0 6 00:00:00 60.0 60.0 66.0 66.0 60.0 60.0 66.0 66.0 00.6 00.0 66.0 00.06 -- 06.0 60.0 06.0 66.0 6on0 ”mm 00.0 60.0 66.0 66.0 60.0 60.0 06.0 66.0 06.6 60.0 60.0 00.06 -- 60.0 00.6 60.0 66.0 60 0 0 00.0 60.0 06.0 00.0 60.0 60.0 66.0 00.0 06.0 66.0 00.6 00.66 -- 06.6 66.6 66.6 00.0 60.6 6“: 00.0 60.0 66.0 00.0 60.0 60.0 66.0 00.0 66.6 06.0 66.6 00.66 -- 00.6 66.6 00.6 00.0 60 6 660 00.0 00.0 06.0 00.0 60.0 60.0 66.0 00.0 66.6 66.0 00.0 06.06 -- 00.6 06.6 06.6 00.0 00 6 < 00569 6.6060 [606.60 6.60.60 6.66.60 6.6.060 C630 6.6060 6.6060 vno~6< 6000660... .06 memo-.6 32300.3 0»: 36:0 ouocamozn 0636 056:0 006665 0606mm 06m6mm xo6m6mm munmmmm umdmwmm mm0m6mm .mnmmmw nonomom 0060660 xo6onmu muonmmonm 6ouoe -ou60-0z -0xo -6;u60 -o~>m-oz -0xo -6nu60 6600 -06o6om -06o600 -ou6a-0z-00 .0 6 6 6 6 O 6‘ O 0.... AvOSCMucou. v 06660.6. 6&3 .U.o o\6< muuznuozmou>monz I + .vo>oeou no: om I ~>¢HU .vo>0§ou um I 60060 .0606 ..6u as 66000 Iowa cacao 60.0 60.0 00.0 00.0 60.0 60.0 60.0 00.0 66.0 66.0 +.0.0.06.0 00.06 06.0 60.0 06.0 06.0 60.0 60.0 660 60.0 60.0 60.0 00.0 60.0 60.0 60.0 00.0 00.0 00.0 +.06.0.06.0 06.66 66.0 60.0 00.0 06.0 00.0 00.0 660 60.0 60.0 00.0 00.0 60.0 60.0 00.0 00.0 66.6 06.0 +.06.0.06.0 00.66 06.0 66.6 60.6 06.0 00.6 00.0 60 60.0 60.0 60.0 00.0 60.0 60.0 00.0 60.0 06.6 66.0 +600.0066.0 06.66 66.6 60.6 66.6 00.6 00.0 06.0 0< 60.0 60.0 60.0 00.0 60.0 60.0 00.0 60.0 06.6 66.0 +.00.0.60.0 00.06 00.6 00.6 66.6 66.6 06.6 60.0 66< 00.0 60.0 60.0 00.0 60.0 60.0 00.0 60.0 06.0 66.0 +.06.0.66.0 06.06 00.6 00.0 00.6 00.6 60.0 06.6 66¢ aucsxdx 60.0 000.0 00.0 60.0 60.0 60.0 60.0 00.0 00.6 06.0 +666.0.00.0 06.6 06.0 00.0 06.0 66.0 00.6 06.0 660 60.0 000.0 00.0 00.0 60.0 60.0 00.0 60.0 06.6 00.0 +.06.0.00.6 00.6 06.0 00.6 06.0 60.0 00.6 06.0 0660 60.0 60.0 00.0 00.0 60.0 60.0 00.0 00.0 06.6 00.0 +.00.0.00.6 60.6 00.0 06.6 60.0 00.6 00.6 60.0 660 60.0 60.0 00.0 00.0 00.0 60.0 00.0 60.0 00.6 60.0 0.06.0000.0 06.0 00.0 06.6 00.0 60.6 66.6 60.0 60 60.0 60.0 00.0 00.0 60.0 60.0 00.0 00.0 00.6 66.0 +606.0.66.0 00.0 00.0 00.6 00.0 66.6 66.6 06.0 60 anouonu m>a6u 6>06U «>060 60060 ~>06u ~>06u m>06u 6>m6u cmO~6< cmOqu .u.o MONUL wumcmmonm vuo6 0u6co cuozmmozm ouu6 ou6co ammuom m0m6mm mnm6ma xo606mm mm6m6mm mmdm6mm mm6mmmm m6m6mm mmnm6mm mm6m6ma mm6mmmm Imwmmmmmmm 600o6 -ou60-0z -0xo -6zu60 -ou>a-az -axo -6:060 6600 IUMONQL IUMONOE IOhxnnmIMZth Ct MONH‘ MONOh Inn. .l it..i.§'.f. I. ”u '1” .Irh'.’ Ill .- 1.“ Eli-l. . I g .0oac6ucou0 0 «6000 70 of other workers (Juo et al., 1974; McKeague and Day, 1966). The dithionite extracted Fe also exceeds the Na—pyrophosphate extracted Fe in these soils. Dithionite extractable Fe oxides increase with depth in the soils of the Upland Surfaces of Highly Weathered Material, i.e., Baoma, Manowa, Njala and Makeni series (Figure 3a). These soils are on more stable surfaces (Figure 2), interfluves and seepage slopes. Here, infiltration rather than runoff is the main process. Juo et a1. (1974) observed a similar trend on six well- drained upland soils in Nigeria. Rhodes and Sutton (1978), in earlier studies on some Sierra Leone soils (some of which are included in this study), observed a similar trend of increase in percent Fe oxide with depth, a trend similar to that of total Fe. In the soils of the Steep Hills and Slopes, there are marked decreases with depth of dithionite Fe oxide in the Segbwema series (Figure 3b). However, a zone of accumulation of Fe oxide is observed in the upper B horizon of the Timbo series. The decrease of Fe oxide with depth in the Segbwema profile is in sharp contrast to the dis- tribution of Fe oxides in the soils of the Upland Surfaces of Highly ~Weathered Material (see discussion above). The Segbwema profile being located on a 42% slope, where runoff is the predominant process rather than infiltration, may account for this trend. This effect of slope 'is not evident in the Timbo* series, which is on a 6% slope. The distribution pattern of total Fe shows a zone of maximum in the Segbwema series, and increases with depth in the Timbo series. * Timbo series generally occurs on slopes steeper than 6%. This profile represents the less sloping members. Figure 3a. Profile distribution of Fe203 and A1203 extracted by dithionite, oxalate and perphosphate solutions in Makeni series. Depth in cm and horizon % Fe203 and Al203 0 1 00 1 g 33 , 1 § £5 .1. L 25' - - 321T l 501 f 822- * Dithionite Fe203 O Oxalate F6203 A Pyrophosphate F9203 * Dithionite Al203 160‘ O Oxalate Al203 1) .1. "WWW” ’29 .-._ E‘. /0. 73 Figure 3b. Profile distribution of Fe203 and A1203 extracted by dithionite, oxalate and pyrophosphate solutions in Segbwema series. Death in cm and horizon 965:0an and A|203 1 2 a 1 § 3 °) AL 32- W 821t- 70 WP-‘L 82% 150 * Dithionite F9203 O Oxalate F6203 O Pyrophosphate F8203 * Dithionite Al203 C1 0 Oxalate Al203 Z5 $ ‘*i * Figure ?b 75 Figure 3c. Profile distribution of Fe203 and A1203 extracted by dithionite and oxalate solutions in Masuba series. Figure 3d. Profile distribution of Fe203 and A1203 extracted by dithionite, oxalate and pyrophosphate solutions in Gbesebu series. Depth ln cm and horizon Depth In cm and horizon ‘0 O\ % F0203 and N203 . i 1 O .- .a M .00 0Q 'V 18 82“ 55-1 u u- Dlthlonlte F6203 22 0 Oxalate Fe203 B * Dithionite AI203 “ O Oxalate AI203 168 ll r it: 5‘ 3 2 T 2' re c 0‘0 1 I § 1 § J J A .- As: 18‘ 821‘ 48-4 822‘ 63‘ ‘ 323.- ” * Dithionlte F9203 . Oxalate F9203 . Pyrophosphate F0203 * Dlthlonlte Al203 0 Oxalate Al203 158-- fi (t 77 Pendembu (Figure 3c) and Masuba soils of Colluvial Footslopes and Upper River Terraces have zones of Fe oxide accumulation within their profiles. The Pendembu profile has two zones of accumulation. Both these soils are on higher terraces than the soils of Alluvial Terraces and Floodplains (Figure 2). The unstable topographic posi- tions of these soils may be responsible for the comparatively low amount of dithionite extractable Fe oxide, and the double maxima of oxides observed in the Pendembu series. The Timbo series resembles these soils more than the Segbwema or the preceding group. Total Fe increases with depth in both Pendembu and Masuba series. The distribution of the dithionite extractable Fe oxide in the soils of Alluvial Terraces and Floodplains shows a zone of maximum within the profiles. The depth of the maximum varies slightly according to the profile, and it is the deepest and least pronounced in the Moa series. The presence of zones of maxima in the profiles suggests the movement of Fe within the profile or stratification of the materials. The maximum amounts of this form of Fe203 are similar in these soils to the amounts in the soils on the Upland Surfaces. The distribution pattern for total Fe203 is the same as the dithionite for Gbesebu and Makundu soils, but increases with depth in the Moa series. Ammonium Oxalate Extractable Fe Oxides The distribution pattern of Fe oxides extracted by this method, in the soils of the Upland Surfaces of Highly Weathered Material (Baoma, Manowa, Njala and Makeni series), show a gradual decrease with depth in the profile. Similar trends of decrease with depth of the oxalate extractable Fe oxide are observed for soils of the 78 Steep Hills and Slopes (Segbwema and Timbo series). In the Segbwema profile, the distribution pattern is the same as in the case of the dithionite extractable Fe (Figure 3b). The possible reason for this trend has been discussed previously. The main difference between the distribution patterns for the two groups of soils is that the decrease in depth is less marked for soils of the Upland Surfaces. In the soils of the Colluvial Footslopes and Upper River Tribu- tary Terraces, decrease of oxalate extractable Fe with depth is also observed in the Pendembu series. The distribution is fairly constant in the Masuba profile. The comparatively high surface values of the oxalate extracted Fe oxide may be explained by the extraction of some organic Fe complexes by the ammonium oxalate solution. McKeague and Day (1966) reported that ammonium oxalate solution extracts Fe and Al from amorphous inorganic substances as well as from horizons of Fe-Al organic matter complexes. The Alluvial Terrace and Floodplain soils (Gbesebu, Moa and Makundu series) show zones of maxima, a trend similar to that of the dithionite extractable Fe oxide. The zone of maximum is less dis- tinct in the Moa series. Sodium Pyrgphosphate (0.1M) Extractable Fe The distribution patterns of the Fe oxide extracted by this method for eight representative profiles are discussed. These profiles are selected from three* of the four groups of soils studied. * The soils of the Colluvial Footslopes and Upper River Tribu- tary Terraces (Pendembu, Masuba) are low in Fe oxide. 79 In the soils of the Upland Surfaces of Highly Weathered Material, Na-perphosphate extractable Fe oxide decreased with depth in the Baoma series, but showed zones of maxima in the second and third horizons of the Manowe,Njala and Makeni (Figure 3a) series. In the soils of the Steep Hills and Slopes, Na-pyrophosphate extractable Fe oxide decreased with depth in the Segbwema series (Figure 3b). A zone of maximum is observed in the Timbo series more like the Manowa, Njala and Makeni series. The distribution pattern of the Na-pyrophosphate extractable Fe oxide in the soils of the Alluvial Terraces and Floodplains (Gbesebu and Makundu series) also shows zones of maxima in their second and third horizons, the highest values being in the second horizon (Figures 3b and 3d). Decrease with depth of Na-pyrophoshate extractable Fe has been reported by Juo et a1. (1974) in five out of six Nigerian soils studied (depth of study llO-lSO+cm). The soil with an accumulation in the second horizon is the Alagba/Benin series, located in an area of high rainfall. Bascomb (1968) also reported zones of higher K—perphosphate extractable Fe oxide in the B horizon of some soils. One possible reason for the trend in the distribution of the Na-pyrophosphate Fe in the soils studied is that in the surface horizons the organic matter is in the less colloidal state, which is necessary for complexing the Fe compounds. The fact that the highest value of Na-pyrophosphate extractable Fe is obtained on the surface horizon of the Segbwema profile could be explained by its topographic location. The organic matter in the surface horizon of the Baoma series may be in a more decomposed form. 80 The correlation between percent organic C and Na-pyrophosphate extractable Fe is low. Figure 4a is a plot of the ratio Fe203 Na- pyrophosphate/% organic C (as determined by Walkley and Black method, cited in Odell et al., 1974) with depth. A zone of maximum is evi- dent for all of the eight representative soils studied, with the exception of the Segbwema series (Steep Hills and Slopes), which shows a gradual increase in the ratio from the B to the C horizon. The higher values of the ratio in the B horizons indicate a small relatively mobile colloidal organic matter that can carry a maximum of Fe oxide. In the surface horizons the organic matter, which is in a less colloidal state, is immobile. The small value of the ratio therefore indicates a large amount of immobile organic matter associated with a small amount of Fe oxide. Similar trends were observed by Bascomb (1968) (Figure 4b), who worked with various kinds of soils, including a Ferritic Brown Earth. The values for organic C and Fe used in his study were those extracted by K—pyrophosphate. However, the graphs obtained in this study show a striking resemblance to those obtained by Bascomb (1968) (Figures 4a and 4b). Total Fe Oxides Total Fe oxides, as reported by Odell et al. (1974), is the highest in the soils of the Upland Surfaces of Highly Weathered Material, followed by the soils of the Alluvial Terraces and Flood- plains. It is lowest in the soils of Colluvial Footslopes and Upper River Terraces. In soils of the Steep Hills and Slopes, total Fe is comparatively lower in the Segbwema series. The values for the Timbo series are within the same range as for soils of the Upland Depth ln cm 33 }) Fe203 (pyrophosphate) QC ? b0 -& r0! C MAKENI, Typic Umbriorthox O SEGBWEMA, Orthoxic Tropudult D GBESEBU, Fluventic Dystropept F1 . (wa1me Ezra 4. Ratio of pYrophosphate Fe203 to organic C. 3- lack) thh tepth in Makeni, Segbwema and Gbesebu series. r3 9.: Fe §Fe (pyrophosphate) % ( " ) O .1 .2 .3 .4 .5 .6 .7 .8 9 1 AHUmUS 1 1 1 1 1 1 J L l J A Upper a A Lower- B Upper- B Lower- C .- O Ferritic Brown Earth 0 Sols Bruns Acides o Humus - Iron Podzol Figure 4b. Data from C. L. Bascomb(1968x showing ratio of Fe(Pyroph0§phate) L" for com arison with F1 IO Organic C. (Pyrophosphate) p gure 4‘. 83 Surfaces of Highly Weathered Material. The pattern of distribution is similar to that of the dithionite Fe, with the exception being the Segbwema series, which has a zone of maximum in the B horizon. Amounts of Fe Oxides Removed by the Three Extractants The amounts of free Fe oxides extracted by the three methods are highest, with dithionite-citrate-bicarbonate solution as the extractant. This extractant has been shown to extract both crystalline and amorphous oxides by some workers (Blume and Schwertmann, 1969; McKeague and Day, 1966; Mehra and Jackson, 1960). The sum of the dithionite and oxalate extractable Fe oxide is the lowest in the soils of the Colluvial Footslopes (which are also lowest in total Fe); the highest values are obtained in the soils of the Alluvial Terraces and Floodplains. In the soils of the Upland Surfaces of Highly Weathered Materials (the Manowa, Njala and Makeni series), Na-pyrophosphate extractable Fe oxide is usually higher than that of the acid ammonium oxalate extracted Fe oxides. The Baoma series, however, has higher values of Na-pyrophosphate Fe in its surface horizon than oxalate Fe. The trend reverses in the lower part of the profile, with a sharp decrease in the Na-pyrophosphate Fe. This change is above the con- trol section for this soil. In the soils of the Steep Hills and Slopes, oxalate extractable Fe oxide is consistently higher than Na—pyrophosphate Fe in the Segbwema series, as opposed to the Timbo series, in which Na- pyrophosphate extractable Fe oxide is usually higher than that of the acid ammonium oxalate extracted Fe oxide. 84 Both the Gbesebu and Makundu soils of the Alluvial Terraces and Floodplains have comparatively high values of oxalate extractable Fe oxide. This form of Fe oxide is consistently higher than Na- pyrophosphate extractable Fe in the Gbesebu series. Higher values of Na-pyrophosphate Fe are in the top three horizons of the Makundu series. The trend reverses in the lower part of the profile, with a sharp decrease in the Na—pyrophosphate Fe. The higher values of Na-pyrophosphate Fe obtained in the Manowa, Njala, Makeni and Timbo series is in disagreement with results obtained by McKeague (1967) working with temperate soils, and Juo et al. (1974), who worked with some Nigerian soils. McKeague (1967) observed higher amounts of oxalate extracted Fe oxide than that extracted by Na-pyrophosphate. Juo et al. (1974) observed a similar trend for five of the six upland soils that they studied. In the Nkpologu/Nsukka series, the amount of Na-pyrophosphate Fe oxide was slightly higher than that of the oxalate Fe oxide throughout the profile. They suggested the presence of a moderate amount of gel- form (Fe-hydrous oxide) as the reason for the result which they obtained. This idea was supported by the studies of Bascomb (1968). While this may be true in the case of the upland soils they studied, it is also likely that some other factors are involved. High temperatures and a prolonged dry period have been shown to dehydrate amorphous Fe and Al and subsequently shift it to a system of greater crystallinity (Sherman et al., 1953). It is believed that this is a contributing factor to the low amounts of oxalate Fe (amorphous Fe) extracted from the above- mentioned soils, as compared to the Na-pyrophosphate Fe. 85 The Manowa, Njala and Timbo soils,vflfixx1experience a prolonged dry season of about 6* months, have been cultivated intermittently, and thus have had less continuous vegetative cover. The Baoma series, on the other hand (Plate lb),1unsbeen under prolonged cultivation of cocoa and coffee trees fgrown under shade) and thus have a much more moist surface over a longer period than the other profiles mentioned. It is also less gravelly. This local variation in moisture regime may explain the unusually high amorphous Fe oxide recorded for this soil. (However, all these soils have higher Na— pyrophosphate Fe than oxalate Fe contents in the A and upper B horizons.) A moisture study conducted on the Njala, Makeni and Timbo soils for a period of about two years (van Vuure and Miediema, 1972) showed these soils to have less than 10% total moisture between January and early April. These soils also have a drought period of 4 to 5 months (Table l), which tends to support the above explanation of the lower oxalate Fe than dithionite Fe. The comparatively higher values of the oxalate Fe in the Baoma, Segbwema, Gbesebu and Makundu series are a possible reflection of both moisture regimes and degrees of development. The ratio of oxalate Fe/dithionite Fe, the active Fe ratio, decreases with depth in the soils of the Upland Surfaces of the Highly Weathered Material (Baoma, Manowa, Njala and Makeni series). In the soils of the Steep Slopes and Hills, the ratio is fairly constant in the Segbwema series, but decreases with depth in the Timbo series. A decrease with depth of the ratio is observed in the * Duration of dry season. Soil is usually moist 1-2 months after the beginning of the dry season. 86 Pendembu series, while the ratio remains fairly constant in the Masuba series, both of which are soils of the Colluvial Footlepes and Upper River Tributary Terraces. In the soils of the Alluvial Terraces and Floodplains, the ratio shows zones of maxima in the Gbesebu and Makundu series. A decrease of the ratio with depth is observed in the Moa series. The decrease in the ratio with increasing depth in the profiles clearly shows that higher prOportions of crystalline Fe oxides are present in the lower horizons of the profiles. In the Segbwema, Masuba, Gbesebu and Makundu profiles, the highest proportions of crystalline Fe oxides are in the upper parts of the profile. Some workers (Alexander, 1974; McKeague and Day, 1966; McKeague et al., 1971; Juo et al., 1974) have used the active Fe ratio to indicate the degree of profile development or age trends in land— scapes. The latter authors worked with tropical soils and the other three with temperate soils. McKeague and Day (1966) and McKeague et a1. (1971) reported active Fe ratio of 30-60% for the soils they worked with, and Juo et a1. (1974) reported values of about 8% for upland Nigerian soils. Alexander (1974) obtained a much higher ratio for lower terrace soils than for the older terrace soils which he studied. The range of active Fe ratios in the soils studied here is 0.07-0.42 (7-42%) for the upper B horizon. The lowest values, 0.07-O.l7, are obtained for Manowa, Njala, Makeni (soil of Upland Surfaces of Highly Weathered Material) and the Timbo series of the Steep Hills and Slopes. The Baoma series, also of the soils of Upland Surfaces of Highly Weathered Material, has a 87 ratio within the range of 0.21—0.27, which also includes the Pendembu and Masuba series (both of which are soils of the Colluvial Footslopes and Upper River Tributary Terraces, with low free and total Fe203). From the foregoing discussion, the distribution pattern of the dithionite extractable Fe oxide appears to be useful in distin- guishing the soils studied in relation to geomorphic surfaces. The ratio Fe O ox/FeZO 2 3 d also appears to be useful in determining 3 the degree of profile development. Amounts of Dithionite, Oxalate and Na-Pyrophosphate Al Data for aluminum oxides extracted by all three methods are given in Table 4. Figure 3 includes a plot of A1 with depth for selected profiles. There is a general slight decrease with depth of Al oxide extracted by dithionite in ten of the eleven profiles studied. The Moa series, however, shows an increase with depth of dithionite Al oxide. Both decrease with depth and zones of Al maxima are observed in the profiles when ammonium oxalate is the extractant. The Makeni, Timbo and Makundu soils have the most prominent zones of Al oxide maxima. Na-pyrophosphate soluble Al oxide was determined for five profiles. Both the Njala and Makeni profiles (soils of Upland Surfaces of Highly Weathered Material) have zones of maxima of Na- perphosphate A1, a trend similar to that of Na-pyrophosphate extractable Fe oxide. The Segbwema (soils of the Steep Hills and Slopes), Gbesebu and Makundu (soils of Alluvial Terraces and 'Floodplains) profiles show decreases with depth of Na-perphosphate 88 Al. The values of Na-pyrophosphate A1 for the Segbwema profile are very low. The similarity in the distribution patterns of Na- pyrophosphate Al/% OC and Fe/% 0C in the five profiles suggests that the Al and Fe are moved by the same organic matter colloidal fraction in these profiles. The amounts of Al extracted by each of the three extractants show no consistent trend among the extractants. In the Njala, Makeni and Gbesebu series, ammonium oxalate extracted more A1 than the dithionite solution throughout the profile. In the Baoma, Manowa, Segbwema, Moa and Makundu series, dithionite extractable Al is consistently higher than that extracted by ammonium oxalate. In the remaining three profiles the trend is less clear. Similar trends have been reported by other workers (Juo et al., 1974; McKeague and Day, 1966). In the soils of Upland Surfaces of Highly Weathered Material, Njala and Makeni, Na—pyrophosphate Al is the predominant form of Al oxide in the A and upper B horizons. In the soils of the Alluvial Terraces and Floodplains, Al oxide forms are more varied. The lack of consistency in the trend of the amounts of A1 extracted by the different extractants makes interpretation of the results difficult, eSpecially in distinguishing the forms of the Al oxides. This is particularly true for the oxalate and dithionite extractants. Significance of the Study on Fe and Al Oxides The oxides of Fe and A1 are two of the major components of soils in the humid tropics. Their roles in the chemical and physical 89 behavior of these soils include fixation of plant nutrients, influ- ence on adsorption surface behavior, and influence on the aggregate stability (Kellerman and Isyurupa, 1966). It was, therefore, thought that a study of these oxides could be important in understanding the genesis and properties of these soils and in their classification. In temperate regions some researchers have used the oxides and hydroxides of Fe and Al to identify diagnostic horizons, e.g., the spodic horizon; to separate some great soils groups (Franzmeier et al., 1965; McKeague and Day, 1966); and to estimate soil age (degree of development) along a toposequence (Alexander, 1974). Karmanova (1975) showed that some ratios of different forms of Fe oxides together with their distri- bution in the profile can serve as diagnostic criteria of the group and subgroup characteristics of some soils in the USSR. In this study the distribution of various forms of extractable Fe oxides have been discussed. The distribution of the dithionite extractable Fe oxides has been shown to be more important than that of the other forms of extractable Fe, although the oxalate extract- able Fe oxide seems to show Fe-movement within the profile. This is especially so in soils of the Alluvial Terrace and Floodplains, as well as those of the Colluvial Footslopes and Upper River Terraces. The distribution pattern of the dithionite extracted Fe oxide sepa- rates the young soils from the older soils (increase with depth in the older soils vs. concentration zones in the younger soils), but does not appear to separate an Oxisol from an Inceptisol or Ultisol. 90 The active Fe ratio Fe203ox/Fe203d* tends to indicate the degree of soil development, and may have a potential for separating the Oxisols from the younger soils, Inceptisols and Ultisols. How- ever, its use may become limited by the fact that high temperatures and prolonged drying causes the amorphous oxides to dehydrate and shift to a system of greater crystallinity (Sherman and Alexander, 1959). This will have a reducing effect on the ratio and hence on any limits that may be set on this ratio. Prolonged dry periods and high temperatures are common in Sierra Leone, where the dry season is about six months, and soil tempera— ture during this period is over 21°C. Under the same environmental conditions a soil under prolonged vegetative cover will have less dehydrating effect on the oxides than a soil that is under continu- ous cultivation or subject to short fallow periods. Several ratios are tested here. Table 4 contains some of them. The remainder are included in Appendix C. Among those studied are the Fe . values (silicate Fe) of Karmanova (1975). He considered 811 the FeSil value to be equal to the difference between total Fe oxides and dithionite extractable Fe (Fe Fe - Fe ). The results 511: t d obtained in this study show no significant trends that can be used in separating these soils. The ratios of different extractable Fe oxides and percent clay are also examined for their usefulness in separating the soils studied (Table 4). The most useful values among those examined are the Fe203d/clay, the FeZO d—Fe O ox/clay, and Fe 3 2 3 O3ox/clay. 2 * ox = oxalate; d = dithionite, t = total. 91 These ratios are fairly constant with each profile, suggesting a possible co—migration of clay and the Fe oxides. This would indi- cate a mechanical migration of small mineral particles from the A to the B horizons of the soils, a process known as lessivage. The most common values of the ratio of dithionite extractable Fe to percent clay in the literature (Soil Survey Staff, 1975) is 0.10 ornmumzfor highly weathered stable upland soils. This value is obtained when percent clay is that from which free Fe oxides have not been removed. Values of 0.10 or more are obtained for the Baoma, Manowa, and Timbo series when Fe oxide was not removed from the clay fraction. These are soils that have been shown to have oxic horizons, by thin section (see section on "Micromorphology") and chemical analyses dis- cussed earlier. The remainder of the profiles have values of less than 0.10. Three of these previously have been shown to have argillic or cambic horizons within the control section. When the ratio is calculated, with percent clay equal to that when free Fe oxide has been removed, the ratio decreases to 0.09 and the division is not so clearcut. Even Segbwema would then qualify. The ratio of the difference between dithionite extractable and oxalate extractable Fe oxide (Table 4) relative to that of per- cent clay (with Fe removed) seems to have some promise in separating the soils studied. The mean ratio for the eleven profiles is 0.059 (0.06). If a critical value of 0.06 or more in the B horizon is used as one of the diagnostic properties of an oxic horizon in the soils studied, then Baoma, Manowa, Njala, Makeni, Timbo and Moa series will have this criterion on an oxic horizon, in their upper 92 B horizons, and the Segbwema, Pendembu, Masuba, Gbesebu and Makundu series would have non-oxic upper B horizons by this criterion. This division seems to support previous observations of argillic or cambic horizons in the Segbwema, Masuba and Gbesebu series, and oxic horizons for the Baoma, Manowa, Makeni, and Timbo series. This separation appears to be superior to that of the ratio of the sum of the exchangeable bases plus exchangeable A1, relative to Vpercent clay for the soils studied. The difference between dithionite extractable Fe and that of ammonium oxalate represents the crystalline form of Fe oxide (Karmanova, 1975; McKeague and Day, 1966). This is a more stable component in the older soils, where the distribution of oxalate extractable Fe is fairly constant throughout the profiles (except as affected by vegetative cover). The ratio shows a general gradual increase with depth for soils with values of 0.06 or more and an irregular or a definite decrease with depth in soils with values less than 0.06, the Makundu soil representing a borderline situation. Ratios of dithionite Al to that of percent clay are also included in Table 4. The ratio is constant'for the Baoma, Manowa, Makeni and Timbo series, suggesting also a co-migration of Al and clay (Juo et al., 1974). They have ratios of 0.03 or more in the upper B horizon (except Makeni). In the soils of the Alluvial Terraces and Floodplains and the Colluvial FootSIOpes and Tributary Terraces (except Moa) there is a definite decrease with depth in the ratio and all are 0.02 or below. The Segbwema soil shows increases and decreases with depth, with and without the removal of Fe oxides from the clay. 93 Phosphorus Adsorption Justification for these studies is: (a) to determine P adsorp— tion isotherms for these soils since such data were not available; (b) to relate the amount of P adsorbed to other soil parameters already studied and evaluate their practical significance; (c) to study P adsorption on ironstone gravels (a major constituent of some of the soils studied) and to evaluate their practical significance. Adsorption studies were conducted on the A and upper B horizons of the Makeni, Segbwema, Masuba and Gbesebu series, representing the four physiographic units (Table l and Figure 2) and on the ironstone gravels of the surface and subsurface horizons of the Njala and Makeni series. The latter two profiles represent the gravelly soils and also the major types of ironstone gravels recog- nized in this study (see section on "Micromorphology"). Phosphorus Adsorption on Ironstone Gravels Figure 5 shows plots of Langmuir adsorption isotherms for the surface and subsurface gravels of the Njala and Makeni series. Two main shaking times were employed for the gravels, 24 hrs and 48 hrs. The shaking was done on a rotating shaker at a maximum speed of 150 rpm. Higher rotation speeds caused the ironstone to slake, and so also did shaking with a reciprocating shaker. The percentage losses in weight of the ironstone gravels after shaking in distilled water for 6 hrs, with a reciprocating shaker, are given in Table 5. After 24 hours shaking the adsorption maxima for the subsurface gravels of the two profiles are higher than those of the surface 94 .mcfixmom 6: v 9066 .m w>mum m m cm 6cwxm mfinmcu0m6 co6umnom m C I m .Mm musmflm no waswwm 0-06 x 60666\006os o h. 0. m. V. m. N. _. O — _ _ H _ 6 6 Eu Ann-03 e .00. no: 63402 . Egon-mm; 3N novflzwxdz .\ panOSPe d 3 ‘ I N) m 3 1 ¢ 3001/d 8m 13311/8310m 96 .mcflxmcm up we “maw>mum mammz Ucm Howxmz .mEumnu0m6 COHDQHOmom m .00 060060 500-06 0.9.02 0 .5006 -0 0.9.22 0 w .5 00 -06 2&2 . .5 06-0 222 o 00 060060 -06 x 60066\006oe 60 m U x panospe d 80011.1 8m 13311/8310m 98 Figure 5c. A plot of initial P concentration vs. P adsorbed by Makeni gravels, after second 24 hr shaking and 30 days equilibration. P adsorbed PPm l l l 2 3 4 Initial P in solution ppm Figure 5c MAKENI 24mm 0 - 25 0 ° 25 - 50 C I 100 Figure 5d. P adsorption isotherms of Makeni gravels; second 24 hr shaking, followed by 30 days equilibration of the surface gravels. mg P/IOOg c P adsorbed 2 moles/liter m 1"l I46 13- ° / IZ- /. / II- I, . '0'- I, 9’ / I 8*- I, 7’ 7] . 6L / 5" [I l . 4b / 3’ o MakeniUOdIyflo-Zfacm 2 0 Mouni(2"°24hduoui-ug)o-25cm .1- Momi(2“‘24h «moms-50cm 11111111 |2345678 C moles/liter X 10'6 Figure 56 102 Table 5. Loss in weight of gravels with 6 hours shaking in dis- tilled water Weight after 6 Soil Series Initial Weight hours shaking (gravels) gms gms % Weight Loss Njala O-35cm 10.00 9.89 1.1 Njala 35—55cm 10.00 9.91 0.9 Makeni O-25cm 10.00 9.81 1.9 Makeni 25-50cm 10.00 9.89 1.1 horizons (Table 6 and Figure 5). The adsorption maxima increased for both surface and subsurface horizons after 48 hours shaking. In general, the subsurface gravels adsorbed slightly more P than the surface ones. This is particularly true for the 24 hr shaking and the 48 hr shaking of the Njala series (Table 6A). The subsurface gravels are less rounded (Westerveld, 1969) and have more loose Fe oxide coatings on the surface than the surface gravels (see section on Micromorphology). The more rounded nature of the surface gravels is possibly due to their solution* rather than transportation, as has been suggested by other authors (Odell et al., 1974; Westerveld, 1969). The Njala gravels have higher adsorption maxima than those of the Makeni profile (Table 6A). Thin sections of the Njala gravels (Plate 6) reveal a concentration of Fe oxide as distinct bands on * Detailed discussion under Micromorphology. 103 mx\moEokoH ox\m65mmm mx\mmemmv mxxmmEnom mx\mm2nama snmmmno mx\mmavmv madmmz mx\mmeHso mamsnmmm mx\mmsnsoa acmxmz :owwuon m momma coufiuon a mmflumm Aflom mcflxmsw mu: vN Houwm mmflumm aflom v mo mcowfluon m momma cam m mo guummleHm How wEmeE :oHumHomcd .mo manna mx\mmEnm mx\mwsm mx\mmsam ox\mmsmm mx\mmsna mx\mmEvH ficmxmz nu- In- nu- mxxmmsom mxxmmavm mx\mwsvm mx\mmemm mamflz wommusmnsm womwusm momwMSmndm momqumndm oommusm mommusmndm mommusm mmwnwm mama om maaxmcm up an 6cm mmaxmgm u: we mchmnm u: an umH aflom mam>mum mamflz cam acmxmz How MEmeE cofiumuomvd .<® magma 104 Hw>mH mom um pcmoflmacmflm ++ Hm>ma woa um unmoflMflcmHm+ Hm>ma wm um unmoflMflcmHm «« Ho>ma ma um unmoflMflGmHm k mUMMHSm Insm + H .o H . . .on . . . . m . m . mom Ham ++oa o n «m o +womo o vmma ++mmom o +mmmm ++mmmm o ++omom 0 wow «« mmh m momm ++hmam.o Hmoa.o wmmmo.o nHHH.OI Hmvv.o vvmb. Nam©.o ++moom.o ++vam. mahm. lusmnum vmao.o mmmm.o ++vmmm.o nmmma.on mam©.o ++oaam. mom~.on movm.o mmmm. «mamm. mommusm xomommm mmm0mH< x0m0mH< smomaa ammomom xoMomaa xomomwm cmommm Ha .o.o a coufiuom . up 0 mm +xo 0 mm .noxm mmoflxo Hm H0\Ucm mm wanmuomuuxm no .Hd wanmmmcmsoxm .0 vacmmuo usmoumm can mm magma cw mcofiuomum nuummnmCHm mo mfiwxms nodumuOmpm m :wm3umn mucmfloflmmmoo Gawumamuuoo Hmmcwq .00 manna 105 the surfaces of these gravels, a pattern that is not observed in the Makeni gravels. The Makeni gravels that were shaken for 24 hrs were washed several times with distilled water to remove water soluble P, and also to remove loose soil particles on the surface which may have been responsible for the adsorption of P. The washed samples were then shaken with P solutions containing l-Sppm P for another 24 hrs (second 24 hrs). The P adsorbed after this treatment was determined, and the gravels were then allowed to equilibrate in the solutions for a further 30 days. Figures Sc and 5d are plots of P adsorbed in relation to the initial P concentration in solution, and the Langmuir adsorption isotherms of Makeni gravels, respectively, with the above treatments. The P adsorption maxima after the 24 hour treatment are 8mgP/kg and 9.0mgP/kg for the surface and subsurface horizon gravels, respectively (Table 6A). These values may represent the amount of water soluble P that was originally adsorbed by these gravels. After the 30 days equilibrium the amount of P adsorbed per gram of gravel (Figure 5c) is much higher than that of the 24 hour shaking The maximum adsorption of the surface horizon increased to 27mgP/kg. This value may not represent only adsorption on the surface of the gravels, but possibly also P that diffused into the gravels. The mechanism of adsorption on the gravel is not clear. The sizes< -- N N Ha «H a v NNm m o m Ha 4H - m HNm -- m - 4m HH - - Ha flfimv—MZ NH OH N mm m N H monum>a 0H s m on m N m NNm NH m H mm a m - HNm . NH NH N Ha m - - ma 4H HH H mm o H N Ha mHmflz -- m mm mm 4N o N mmmum>< -- m 4 Ho mH m - NNm -- m m mm ON 6 m m4 -- a m as «m m - Ha 030:0: -- m m Ha m o m momum>< -- m m N» 0H s - NmmHH -- a m mm m o - HmmHH -- m m on m o 4 Na n- HH m co m 6 6 Ha mEomm w w w w w w w mCONHHOE flaw UmeWHanHumHQUCH wUHHOHSU QHHHHH wuvHCHHOMVm GUflmQQHO muflflumou NU.HMDQ mwflhmm HHOW thma ..Hm um Hamoov pwflosum moHHwoum co>oam mcu mo mcofluomuw wmao kuou wo coHuHmomEoo .h magma 117 m m m N» OH - o momum>a m N m Ha NH - a uNm N N v up m - m uHm a v m Hm oH - o 0a mmmmmm -- m m vs mH - H momsm>< -- a m «s 6H - H Nm -- m m mp mH - - ma -- a a mm «H - - H4 DQEwmvanm H m m Nm Hm m H mommm>< -- - - Ho mN HH n nNNm -- - m mm mN oH - uHNm -- v m 64 mm m - ma -- w H mm Hm a - NH< v v m Ne OH 0H m HH< onsHe N H v Hm v v v 0mm00>< -n- - 4 Na m a N Ho uNNm -- N v Nm N v 6 can uHNm m N a om v N m Ha mam3nmmm W cm w w w M. mo mGONHHOZ 0C0 GQHMHumuumumucH wufluoch OUHHHH ouHcHHomx wuHmQDHU wuflsuwow NUMMSO mmflumm aflow AvaCflucoov h mHQMB 118 m m O OO HH O momum>< N - m an OH O NNm N N O O» NH O HNm m H m ms HH O HO O O O OO HH m ma N m N ON NH O NHa O O a ma OH - HHm SQCDMMZ O O m Os OH O Ommum>< O O N On NH O ONm O O N ma OH O nNNm O O N On OH - nHNm O O m Ha NH O m4 -- O O ON OH - H4 530 w 090 n- O O OO HH m momum>< -- O O O» NH O mm -- O O as HH O HNm -- O O Os OH O .Imm 002 w w w w w w mCONHHom can OmHuHumuumumucH muHuoHno OOHHHH OOHOHHOOM muHmhnHo muchmoo Nuumso mmHumm HHom AOOOOHOOOOO a OHOOO 119 Figure 7a. X—ray diffractograms for total and fine clays of the A and upper B horizons of three representative profiles — Segbwema, Makeni and Gbesebu. (1) Total clay (2) Fine clay 121 Figure 7b. X-ray diffractograms for total clay fraction of Njala and Makeni gravels, A and B horizons. 50 ‘0 3. 36‘ Cl. 3. 'MA 60 50 122 4 1M 4. Jon C e ' JALA +0 AI 2 30 ‘0 Figure AY OF GRAVE LS 7b 2 60 60 ‘0 30 2O 50 40 123 Figure 7c. x-ray diffractograms for mica flakes in medium and fine sand fraction of the A, B and C horizons of the Segbwema series and B21 horizon of Timbo series. (1) (2) (3) (4) (5) (6) (7) (8) (9) A1, fine sand, Segbwema series A1, medium sand, Segbwema series A1, fine sand, K—saturated for one week and heat to 110°C, Segbwema series B21, fine sand, Segbwema'series B21, fine sand, K-saturated for one week, Segbwema series 821, fine sand, boiled with sodium citrate and K—saturated overnight, Segbwema series B22, fine sand, Segbwema series C, fine sand, Segbwema series B21, fine sand, Timbo series 124 .. 'MICA' SAND FRACTION : «- Figure 7c 125 Soils of the Steep Hills and Slopes: The predominant clay mineral in the Segbwema and Timbo series is kaolinite. Total clay analysis (Table 7) shows that the percentage of kaolinite is higher in the Segbwema than in the Timbo series, i.e., 81% compared to 52%, respectively. Gibbsite content is much higher in the Timbo (31%) than in the Segbwema series (4%). Both of these soils are derived from granite and acid gneiss. Intrusions of basic Kambui Schist are more common in soil provice L (Table 1), to which the Segbwema series belongs. Total analysis also shows traces of illite, chlorite and interstratified minerals in the profile. Goethite is also present in small amounts. X-ray diffractograms (Figure 7a) for total clay of the A and upper B horizons of the Segbwema series show mainly kaolinite with traces of gibbsite and quartz in the A and B horizon, respectively. Both of these samples have many mica-like flakes in their silt fraction. The very low content of illite in their clay fractions is therefore surprising. X-ray diffraction studies were conducted on the mica flakes of the sand fractions to aid in the explanation of the results obtained in the clay fraction. The findings are discussed below, under x-ray analyses of mica in sand fractions. Soils of the Colluvial Footslopes and Upper River Terraces: In these soils, kaolinite is also the predominant clay mineral. Both Pendembu and Masuba series have greater than 70% kaolinite and 10-15% gibbsite. Illite accounts for 5% of the clay fraction. Chlorite, interstratified minerals and quartz are present in very small amounts. 126 Soils of the Alluvial Floodplains and Lower River Terraces: The predominant clay mineral is kaolinite, which accounts for 74-75% of the clay fractions of the three soils in this group. The soils have appreciable gibbsite content, 15-11%, which is higher than for some of the soils of the Upland Surfaces of Highly Weathered Materials (e.g., Baoma and Njala). The soils have very low illite content. Chlorite and interstratified minerals occur as traces. The low content of illite in the Gbesebu profile is interesting, as many mica—like flakes are present in the silt and fine sand fractions. X-ray diffractograms of clay from the A and B horizon of the Gbesebu series are shown in Figure 7a. The mineralogy of the two horizons is strikingly similar. Also, mica flakes from the sand fractions of the B21 horizon were x-rayed (see discussion on x-ray analyses of sand fractions). X-Ray Analyses of Fine Clay Fractions X-ray diffractograms of fine clays from the A and B horizons of the Makeni, Segbwema and Gbesebu series are included in Figure 7a. The mineralogy of both the fine clays and total clays of these soils is similar, with kaolinite being the dominant material. How— ever, the amount of kaolinite is much smaller in the fine clay fraction. Peak intensity ratios between the fine and total clays varies between 0.34 and 0.56. The lowest is in the Makeni series and the highest is in the Gbesebu series. The main difference between the diffractograms of the two clay fractions is that the peaks of the fine clays are broader and flatter, indicating the lower degree of crystallinity or more amorphous nature of the minerals. 127 In the A and B horizons of the Makeni series, the 3.38 A°(101) quartz peak is more intense in the fine clay, and the gibbsite peaks are much reduced. In the Gbesebu series the gibbsite and quartz peaks are noticeably reduced in the fine clay fractions. X-Ray Analyses of Total Clay Fraction in Ironstone Gravels of Ejala and Makeni Series Figure 7b shows x-ray diffractograms of the total clay fractions of the A and B horizon gravels of the Njala and Makeni series. In the Njala gravels, kaolinite and goethite are the main components of the clay fraction of the A and B horizons. Unlike the clay fraction of the fine earth fractions, gibbsite is absent. The Njala gravels are predominantly composed of Type la ironstone nodules (see discussion on Micromorphology), which are mainly frag- ments of mudstones and shales that have been impregnated by Fe oxide (Plate 6A, B and C). The intensities of the kaolinite peaks are much lower than those of the corresponding fine-earth clay fractions, possibly due to a lower degree of crystallinity or Fe oxide coatings. The mineralogy of the clay fraction of the Makeni gravels is similar to that of the fine-earth components. Kaolinite is the dominant mineral and gibbsite is the second most abundant mineral. Quartz is also present as indicated by the 3.36 A°(101) and 4.30 A°(100) spacings. The quartz is assumed to be from the sample, as these peaks are absent for other samples, e.g., Segbwema A, for which the same mounting material was used. Most of the gravels of the Makeni series belong to Type 2a (see Micromorphology). The simi- larity between the clay mineralogy of the clay from the gravels and 128 fine-earth in this soil seems to support the idea that these gravels were formed in place. X-Ray Analyses of Mica Flakes in Sand Fractions of Segbwema and Timbo Series The decision to conduct x—ray studies on the sand fraction was prompted by the occurrence of very small amounts of illite in their clay fraction, even though the sand fractions contain many micaceous flakes. Figure 7c shows x-ray diffractograms of the mica flakes, from the medium and fine sand fractions, of the Segbwema and Timbo series. Interpretations of the air dried diffractograms show the presence of interstratified minerals, with peaks at about 11.8 A°, and 23.2 A°, kaolinite 7.18 A° and 3.57 A° peaks, and reflections of gibbsite 4.39 A° and quartz 3.34-3.42 A°(101) spacing, which account for a smaller percentage of the minerals present in the flakes. The lack of 9.9-10.l A° peaks indicates that the flakes are not true mica, but weathered material. This tends to explain why there is very little or no illite content in the clay fractions of these soils. X-ray analysis of the mica flakes from the 821 horizon of Gbesebu series (diffractogram not shown) gave a 9.8 A° peak of illite, together with the peaks for kaolinite, but no inter- layer mineral occurred. This suggests differences between the mica- like flakes that are found in the Segbwema and Timbo series and those in the Gbesebu series (i.e., biotite in the Segbwema and Timbo series and muscovite in the Gbesebu series parent materials). When mica-like flakes from the sand fractions of the A, B and C horizons of the Segbwema series and the B horizons of the Timbo 129 series were compared (Figure 7c), the main difference that occurs between the A horizon and the B and C horizons is the presence of more highly ordered peaks of the interlayered material 22—24 A° in the B and C horizons. There is also some resemblance between the clay minerals present in the B horizon of the Timbo series and that of the Segbwema (Figure 7c). To determine the interstratified material, the flakes were saturated with KCl, after which they were heated to 110°C. The time of K-saturation was varied, i.e., 8 hours, one week and two weeks. X-ray analyses after these periods showed no collapse of the peaks to the 10 A° of illite (mica) (Figure 7c, 3 and 5). The mica-like flakes were then boiled for five hours with Na-citrate, pH 7.3, before further K-saturation (Brown, 1961). This treatment is necessary to remove A1(OH); occurring in the interlayer position. Tamura (1958) observed that the presence of Al(OH): in interlayers position of interstratified Illite-Vermiculite will prevent the collapse of the peak to the 10 A°. Other workers (Rich and Obenshain, 1956; Brown, 1961) made similar observations. After the above treatment, the samples were x-rayed. Figure 7c, 6, includes the x-ray diffractogram for the treatment for the B horizon of the Segbwema series. The peaks for the interstratified mineral show a shift to the 11.3 A° for the 321 mica-like flakes. The higher order peaks disappeared with K-saturation. This suggests that the inter- stratified mineral is illite-chlorite. After one week of K-saturation and heating to 110°C, the 11.3 A° shifted to 10.6 A° (not shown). The lack of shift in the peaks of the interstratified minerals to that of 10.1 A° after K-saturation and prior to the Na-citrate 130 treatment indicates that K-fixation is not a serious problem in these soils. Significance of the X-ray Analyses The mineralogy of the clay fraction of soils of the Upland Surfaces of Highly Weathered Material and that of the Alluvial Terraces and Floodplains is strikingly similar (Table 7), indicating the common origin of the materials of the alluvium and the uplands. The main difference between the two groups of soils is the higher percentage of kaolinite in the alluvial floodplain soils. Also, in x—ray diffractograms of the Gbesebu and Makeni series, the kaolin- ite peaks are more intense in the Gbesebu than in the Makeni series, suggesting a higher degree of crystallinity of the minerals in Gbesebu. The kaolinite peaks of the Segbwema series are of the lowest intensity and also broadest of the three representative profiles (Figure 7a). This tends to indicate a lower degree of crystallinity. The lack of gibbsite and goethite peaks in the diffractogram also points to a lesser degree of weathering in the Segbwema series. The minerals in the fine clay fractions are poorly crystallized compared to those of the total clay components. This is supported by their broader, flatter, and lower intensity peaks (Figure 7a). Grain Counts of the Fine Sand and Very Fine Sand Fractions The fine and very fine sand fractions (Table 8A) of the five profiles (Timbo, Pendembu, Masuba, Moa and Makundu) are predominantly quartz (Table 8B). Mica (mostly muscovite) is the second most abundant mineral in these fractions. The highest amounts of mica 131 O.ON O.O O.O H.OH O.NH O.ON N.O uNNm h.ON O.O O.O O.OH O.HH N.ON O.N uHNm -- O.O N.O O.OH O.HH O.ON H.N ma Mgwwz O.ON O.O O.O O.OH O.O N.OH O.O mNm O.ON O.O N.O H.OH O.O O.OH O.O NNm H.Om O.O O.O H.HN O.O O.OH O.O HNm N.Om O.O H.O N.NN O.O H.OH 0.0 mm -- N.O O.O N.ON O.O H.OH O.O H< Sgwmvcwm N.ON O.m O.O O.OH O.O O.NH O.OH nNNm O.ON O.O O.O O.OH O.O 0.0H O.NH nHNm O.ON O.N N.O O.OH O.O N.OH N.NH ma -- N.H O.O 0.0H O.a 0.0 O.OH NHO -- O.O O.O O.NN 0.0 O.O 0.0 HHO oneHa H.N k.m O.O O.N H.m 0.0 NNm O.N O.O N.O O.O O.N O.O HNm O.H 0.0 0.0H N.O O.N N.O Ha Hcmxmz N eeNO.O-OO.O eeOO.O-H.O ssH.O-ON.O eeON.O-O.O . EEO.O-O.H EEO.H-O.N OOHHOO HHom .Hm.oo-mi OHHO mmumoo mi> mm m: mo Oo> Hmuou mo ucwoumm cw owmmmumxm meHwoum c H©m>oamu mm3 mm kummv HHOm w>wm ca coflusnauumfib uHHm wmumoo Ucm wwwm vcmm .fim OHQMB 132 O.O O.m N.H O.O O.O O.O O.O NNm O.O N.O O.H O.H 0.0 0.0 N.O HNm O.» O.O O.H O.H O.O O.O H.O Hm O.O O.O N.N O.H O.O O.O N.O m< -u- O.O H.N O.N O.O R.O N.O NH< -- O.O O.N O.N O.O H.H 0.0 HH< socsxmz H.N OH.O OH.O NO.O NO.O OO.O ONm k.N OO.O OO.O HO.O HO.O O0.0 ONNm N.O HO.O HO.O HO.O HO.O OO.O OHNm N.O ON.O OH.O H0.0 OO.O OO.O ma N.O OO.O ON.O OO.O O0.0 O0.0 Hm Sflwmwflu O.N O.O O.OH O.O O.O O.O o R.OH O.N O.O O.OH O.O O.O O.N mm O.OH O.N O.O O.HH O.O O.O O.H NNm O.ON O.N 0.0 O.HH 0.0 0.0 O.H HNm -- H.O O.O O.NH 0.0 0.0 O.H -Mmm MOS H eeNO.O-OO.O eeOO.O-H.O eeH.O-ON.O esON.O-O.O eam.OnO.H EEO.H-O.N mmHumm HHom .Hm.0O-mi OHHO «mumoo mi> mm m: mu mo> HOOOOHucoov am anma Table 8A (continued) FS—COoSio Coarse Silt MS FS VFS CS VCS 0.05-0.02mm 0.1-0.05mm 0.25-0.1mm 0.5-0.25mm 1.0-0.5mm 2.0-1.0mm Soil Series MO a A1 12.6 3.9 1.3 20.6 19.4 11.7 3.9 3.6 6.6 B21 B22 11.0 3.6 18.7 10.3 10.7 Gbesebu A1 132 0.54 0.25 0.01 0.05 0.14 0.20 0.10 0.01 0.03 0.10 0.03 0.01 0.01 0.04 0.07 0.01 0.01 0.02 0.00 0.00 0.00 0.00 B21b Bzzb B23 0.01 0.02 0.00 Makundu A11 A12 AB 2.4 0.9 0.6 1.1 0.5 0.2 0.2 1.8 7.9 0.4 0.2 B21 0.7 B22 133 m.NH m.HH ®.HH v.m ©.mm ©.mm v.¢a m.ma ¢.©H 0.50 H.h® ®.h© H.OH mm NNm HNm Ha mo: #Nmm #HNm -Illnmm magmas mm ONm NNm HNm me Ha snfiwncwm QNNQ Damn md NH< HHd oneHs «*mHOHOCHE «mavmmo ummmpamm . “OH: . uuumso «*mHOMOCHE «msvmmo ummmcamm pcmm wcfim-Num> Ocmm OOHO mmfluom HflOm mwaflmoum w>wm mo mcofluomum ocmm wcflm muw> cam vcmm mean How undoo :wmuo Hmumcaz .mm wanna 134 .mHmHmcaE ucmumflmmu uwsuo i ¥ .uwcmmfi mo own 0:» >3 cmEhflmcoo mm wuwuwcmmfi wacwmz k. O.O O.O N.H N.O O.ON O.NO k.O 0.0 O.N O.ON O.NO NNm O.O O.O O.O O.O N.NN O.OO O.O O.O O.O O.NN O.kO HNm O.O 0.0 O.O O.N O.NN 0.00 O.O N.H O.N O.NH 0.0n Hm O.O O.O O.O O.O O.NN O.OO O.O 0.0 H.N N.OH H.OO m< -- O.O O.O O.O O.OH O.OO O.> N.H H.H N.OH O.OO NH< -- O.N O.H N.H O.OH N.OO O.O H.O O.H O.NH O.OO HH< sccsxmz .xoummm *«mHmHOCHE «movOQO Hmmmoamm .woflz. uuumso ««mHmHmcHE «wsvmmo ummmUme .moflz. uuumso mmfiumm aOON 00:00 00:00 HHom 10m w pcmm mcflm mum> Ucmm mcam Awwscflucoov mm magma 135 are found in the subsoil horizons of Timbo and Makundu series. The mica in the B horizons of Timbo series is predominantly weathered biotite. X-ray analyses of this material (see section on x-ray analyses) show that it is a mixture of interstratified illite- chlorite and kaolinite. However, the 'mica' flakes in Pendembu, Masuba, Moa and Makundu series are predominantly muscovite. The highest amounts of opaque minerals (composed mainly of magnetite) are in the Pendembu and Moa series, both of which are from soil province L (Table 1). High magnetite content is also recognized in the other soils studied in soil province L (see Table 1). The Pendembu series appears to be the most intensely weathered of the five representative profiles. Most of the five profiles have >6% weatherable minerals (mainly muscovite) in 20-200u fractions of their upper B horizon and there; fore do not meet this requirment of an oxic horizon (Soil Survey Staff, 1975). But the Pendembu series meets this requirement of an oxic horizon. However, it does not meet the free oxides/clay ratios expected of Oxisols. Micromorphology Soil morphology has been (restrictedly) defined as: (a) the physical constitution, particularly the structural properties of a soil profile as exhibited by the kinds, thickness, and arrangement of the horizons in the profile, and by the texture, structure, con- sistence and porosity of each horizon and (b) the structural char- acteristics of the soil or any of its parts (Brewer, 1964). 136 Soil micromorphology, then, is the study of soil morphology in the range where optical instruments (microscopes) are needed to aid the naked eye (Buol et al., 1973). It involves examination of the soil at a lower level than that required for field investigations. The emphasis is on the spatial arrangement of simple discrete grains and associated voids as well as the composition of the soil matrix. There are two basic subdivisions of micromorphological studies: split and debris examination of the samples or profiles, after Kubiena (1938),and thin section observations, after Brewer (1964). The former were conducted here with aid of a binocular microscope, and the latter with a petrographic microscope. The terms used in describing features observed in thin sections are adopted from Brewer (1964). Also, the description of the soils is divided into two groups: gravelly soils and non-gravelly soils. Each of the series descriptions will be divided into the whole soil and the gravels. Split and Debris Analyses Examination of samples from the horizons of the Segbwema series with the binocular microscope revealed the presence of clay skins on the surfaces of peds in the B horizon. This supported similar observations in the field study. Samples from the B horizons of Pendembu, Masuba, Moa, Gbesebu and Makundu series were also examined. However, the studies on these soils revealed no clay skins on the surfaces of the peds. 137 Descriptions of Thin Section of Gravelly? Soils Plus Their Ironstone Nodules (Baoma, Manowa, Njala, Makeni and Timbo) Baoma series (Plate 2), 15-20cm, B2: The skeletal grains are randomly distributed, the predominant sizes being medium and fine sand. There are a few coarse sand grains. The fabric is intertextic to agglomeroplasmic and the plasma is composed mainly of a mixture of clay and Fe oxide. The clay is coated with Fe oxide; hence the undulic nature of the plasmic fabric at 200x. The horizon is porous. There are a number of channels and planes. The latter are mainly due to cracks in the soil matrix, possibly due to shrinking, evidence of craze planes. The channels are single and dendritic. Some are coated with sesquans, mainly Fe oxide, as confirmed with reflected light. A patch of ferri-argillan was observed between the soil matrix and the nodule, possibly due to pressure of the nodule against the soil matrix. Lithorelics are mainly small pieces of mica (muscovite) that weathered in place to clay. The clay shows preferred orientation. Pedological features are in the form of Fe oxide nodules; both Type 2a and 2b are recognized.(Plates 2a,b and 15) (see definition of the different types in the discussion). The shape of the nodules is subrounded to subangular. The percent of nodules in the horizon is low. * Ironstone gravels 138 PLATE 2. Thin sections, Baoma series (whole soil). (a) 15—20cm (B2), plain light, 40X. Soil matrix and a Type 2a nodule (round features). Note the similarity between the soil matrix and that of the nodule. (b) 15-20cm (B2), plain light, 40X. Soil matrix and a Type 2b nodule (left half of photo). Note a small Type 2a nodule near the larger quartz grain in right half of photo. 139 Plate 2b 140 PLATE 2 (continued) (c) 30-38cm (B2), plain light, 40X. Type 2b nodule (lighter color) and soil matrix. Note the Fe oxide on the surface of the nodule. Also, note the cracks in the soil matrix. (d) 30-38cm (B2), plain light, 40x. Oriented clay within a Type 2 nodule along an old channel. Plate 2d 142 PLATE 2 (continued) (e) 30-38cm (82), X nicol. Oriented clay within a Type 2 nodule along an old channel. (f) 43-50cm (B2), plain light. Soil matrix and both Types 2a (north- east part of photo) and 2b (southwest part) nodules. Note the cracks in the soil matrix and the Fe oxide concentration on the surface. Plate 2e Plate 2f 144 30-38cm, B2: The skeletal grains are randomly distributed and they are composed mainly of medium and fine sand grains. The fabric is intertextic-agglomeroplasmic and the plasma is composed mainly of clay coated with sesquioxide, mainly Fe oxide. The plasmic fabric is inundulic to undulic at 200x. The horizon is porous. There are manycracks (Plate 2c). The channels are mainly single and dendritic. The vughs are mainly irregular orthovughs. Pedological features occur in the form of nodules. Two types of nodules are recognized: Types 2a and 2b (Plate 2c), the plasma of which is composed of clay covered with Fe oxides, enclosing a few sand grains (mainly quartz grains). In both types, the perimeter (contact with the soil matrix) is coated with Fe oxides (Plate 2c). Some oriented clay minerals are evident within the matrix of the nodule (Plate 2d and e). Exfoliated, moderately weathered mica pieces (possibly muscovite) are also observed. The shape is sub- rounded to subangular. There is a slightly higher percentage of nodules than above. 43—50cm, B2: The skeletal grains are randomly distributed, as above. The fabric is agglomeroplasmic and the plasma is composed mainly of clay coated with Fe oxides. The plasmic fabric is inundulic-undulic at 200x and undulic at 400x. The horizon is very porous. There are numerous cracks and a craze plane is evident (Plate 2f). The plasma contains lithorelics, mainly small pieces of oriented clay, possibly derived from the weathering of mica in situ. Some of the channels are lined with sesquans, predominantly composed of Fe oxides. The vughs are mainly irregular orthovughs. 145 Orthic-pedological features include Mn concretions and both Types 2a and 2b nodules, which are elongated and mainly subangular in shape. The plasma of the nodules is similar to those described above. 60-68cm, IIB31(1): This horizon contains more nodules than the overlying one. The skeletal grains are randomly distributed and the size is predominantly medium and fine sand. The fabric is agglomeroplasmic and the plasma is composed mainly of clay and silt size material, with coatings of sesquioxides, mainly Fe. The plasmic fabric is undulic at 80X. The horizon is porous; a craze plane is evident in the soil matrix between the nodules. The channels are single and dendritic. Two kinds of orthic-pedological features are observed: Mn nodules and Fe oxide nodules. The two types of Fe oxide nodules described previously are also recognized. Some of the clays in the nodules show preferred orientation. In some cases the oriented clay appears to have been developed on the surface of the matrix of the nodules (i.e., plasma concentration of clay within the nodules). The present—day surface is coated with Fe oxides. Type 2b nodules appear to have been developed, in this case, around a nucleus of plasma concentration of clay possibly derived from the weathering of primary minerals, feldspar or mica. The nodules are subangular to angular. Some of the nodules are less than 2mm in size. 60-68cm, IIB31(2): The soil matrix is similar to that of IIB31(1). There are a few elongated metavughs, with vugh sesquans. 146 Also, some of the nodules are 2mm or less in size. These appear as a plasma of clay that has been coated with Fe oxide (Type 2c). Some of the nodules appear as compound pedological features having both host and included sesquioxide nodules. Channel ferri-argillans are present within the concretions. 110-115cm, IIB32: The skeletal grains in the soil matrix are randomly distributed. The sizes of grains are predominantly medium and fine sand. A few coarse sand grains are also present. The fabric is agglomeroplasmic and the plasmic fabric of the soil matrix is undulic-inundulic at 200x. The horizon is porous. Cracks are common, though fewer than those of the above-described horizon. A craze plane is evident. Orthic-pedological features are of two kinds: Mn nodules and sesquioxide nodules. The perimeter of the nodules is coated with sesquans. Both Types 2a and 2b sesquioxidic nodules are recognized. There are a few old channels in the sesquioxidic nodules that are filled with Fe oxide and lined with strongly oriented ferri-argillans. Weathered grains of mica are present within the nodules; so also is chalcedony. Baoma gravels (Plate 3), 0-15cm A1: Subrounded in shape, Type 2a. The matrix contains randomly distributed skeletal sand grains similar to those of the soil matrix, and include pedological features such as Mn nodules. There are a few old channels that are filled with sesquioxides. The plasma is predominantly clay coated with Fe oxide. The fabric is agglomeroplasmic. The plasmic fabric is undulic-inundulic at 200x. 147 PLATE 3. Thin sections, Baoma series, gravels. (a) Type 2a nodules: (a) 0-15cm (A1), plain light, 80X; (b) 160-180cm (B3), plain light, 80X. Note the similarity between the matrix; the lighter area in (a) is possibly a vugh that was filled with soil and organic material. Note the thick Fe oxide coatings on the edge of the vugh. Also, note fragments of weathered mica in both concretions. (b) ‘_-_-_——4- — 148 Plate 3a Plate 3b 149 The nodule is porous; some orthovughs are present. There are also a few metavughs. Some of the vughs have been filled by soil material and have their walls lined with Fe oxides (Plate 3a). Weathered grains of mica are present. The present-day surface is coated with Fe oxides. 15-30cm, B2: The skeletal grains are fewer in this nodule, which is similar to the Type 2a sesquioxide nodule described earlier. The shape is subrounded to subangular. The plasma is predominantly composed of clay coated with Fe oxides. The fabric is agglomeroplasmic- porphyroskelic and the plasmic fabric is undulic at 100x. There are old channels filled with ferri—argillans. The perimeter is coated in patches with sesquans. 30—60cm, B2: This nodule is similar to Type 2a described earlier. The skeletal grains are randomly distributed, and the size ranges between medium and fine sand. There are a few large sand grains. The matrix of this nodule is similar to that of the soil matrix. The nodule is porous. There are irregular orthovughs. The fabric is agglomeroplasmic. The composition is mainly clay, coated with Fe oxides. The plasma is undulic-inundulic at 200x. The outer surface of the concretion is coated with sesquans (Fe oxides). Plasma concentration of clay (partially coated with Fe oxides) with preferred orientation is present, possibly due to the weathering of primary minerals in place. 'In situ' weathering of mica flakes is observable. Shape subangular. 60-110cm, IIB31: This appears as a compound pedorelic feature with the host pedorelic feature similar to that of the soil matrix. 150 The skeletal grains are randomly distributed, and the size of the grains is predominantly medium and fine sand. The included pedo- relic is basically a concentration of clay coated with Fe oxide. The fabric of the host pedorelic feature is agglomeroplasmic and the plasmic fabric is undulic to inundulic at 200x. The concretion is porous. There are a few irregular metavughs lined with ferri-argillans. The shape is subangular to angular. There are also a few channels that are coated with sesquioxides (Fe oxides). The perimeter is irregular. Examination at 400x depicts the rough nature of the surface, suggesting weathering (possibly by solution). 160—180cm, B33: The skeletal grains are randomly distributed, predominantly medium size sand grains. The fabric is agglomero- plasmic. The plasmic fabric is inundulic-insepic at 200x. Included pedological features are plasma concentrations of Fe oxides, or Mn nodules and clay. The clay has preferred orientation, due possibly to the weathering of feldspars and other primary minerals 'in situ.‘ The vughs are mainly irregular orthovughs. Some vughs appear to be coated with a birefringent clay-like material. There are a few channels which are Open to the outside which contain sesquans. Weathered mica grains are present. The shape is angular. The perimeter is rugged (Plate 3b). Manowa series, 75-80cm, B21: The skeletal grains are randomly distributed, mostly of medium and fine sand size. The fabric is intertextic to agglomeroplasmic. The plasmic fabric is asepic to undulic at 80X. There are many cracks in the soil matrix, with some 151 evidence of a craze plane. Channels are single. The plasma of the soil matrix is a mixture of clay and Fe oxides. Orthic pedological features are sesquioxide nodules, mainly Type 2a. In some of the nodules a center pore is observed. There is a great similarity between the matrix of the nodules and that of the soil. The perimeters of the nodules are coated with Fe oxides. 125-130cm, B22: The skeletal grains are randomly distributed and are of predominantly medium and fine sand size, with a few coarse sand grains. The fabric is intertextic to agglomerOplasmic and the plasmic fabric is asepic to undulic at 80X. The plasma is composed of clay coated with some Fe oxides. Channels are mainly single and a craze plane is also evident. Orthic pedological fea- tures are mainly sesquioxide nodules. Mainly Type 2a and 2b nodules are observed. Old channels and vughs are evident in some of the nodules. Some of them have patches of oriented clay coated with Fe oxides. In some of the nodules a single grain is observed in the center, which acts as a nucleus for the deposition of Fe oxides. Manowa gravels (Plate 4), 0-25cm, Al: Type 2a nodules. The skeletal grains are randomly distri- buted; there are some large sand grains. The fabric is intertextic. The plasma fabric is undulic at l60X. The plasma is a concentration of clay with preferred orientation and high birefringence. There are a few orthovughs which are coated with Fe oxides. There are also some channels that contain moderately to strongly oriented clay. The shape is subangular to subrounded. 152 PLATE 4. Thin section, Manowa series gravel. 25-53cm (A3), X nicol, 50X. Weathered mica pieces and other primary minerals that have been protected within a Type 2a nodule. 153 25-53cm, A3: The skeletal grains are randomly distributed and there are common sand grains. The fabric is intertextic and the plasmic fabric is masepic to insepic at l60X. There are plasma concentrations of oriented clay. Some irregular orthovughs have Fe oxide coatings and a few have patches of oriented clay. Weathered pieces of mica (possibly muscovite) are observable (Plate 4). Two zones are recognized in this nodule: a zone with very few skeletal grains in which the fabric is agglomeroplasmic to porphyroskelic, at left, and a zone with intertextic fabric. Similar kinds of nodules have been described for the Makeni series. It resembles a Type lb gravel on which a Type 2b has been formed. Edge weathering is observed. Shape is angular to subangular. 53-88cm, B2: This is similar to that described in the A3. A few weathered primary minerals that have been protected by the nodule can be seen. The shape is angular and possible edge weather- ing is observed. 88-125cm, B22: This is similar to that described for 53-88cm. The main difference is that the Fe oxide concentration is higher. Weathered and weatherable materials, that have been protected within the nodule, are also present. The shape is angular. Possible edge weathering is observed. It is a Type 2a nodule. 125—175cm, B22: It is similar to the nodule described above. Weathered and weathering primary minerals are observed. It is sub- rounded to subangular in shape. A Type 2a nodule. 154 Njala series (Plate 5), 123-128cm, B22 (Plate 5a, b, c): The skeletal grains are randomly distributed and the grain size is mainly medium and fine sand. The fabric is agglomeroplasmic. The plasma fabric is asepic to undulic at 80X and undulic at 200x. The horizon is very porous. There are a few irregular orthovughs. The plasma is mainly composed of clay coated with Fe. Orthic pedologic features are mainly sesquioxide concentrations. Two types of iron— stone gravels are recognized, i.e., Type la and Type 2a (see later discussion on thin section of gravelly soils, and Plate 15). The perimeter of some of the nodules is coated with ferri- argillan. The shape of the nodules is subangular to angular. The channels in some of the nodules contain argillans (Plate 5a and b). Also present are lithorelics which are mainly concentrations of oriented clay, possibly derived from the weathering of mica or feldspar in situ. l40—l43cm, B22: This section has fewer nodules than that of 123-128cm. The skeletal grains are randomly distributed, the pre- dominant sizes being medium and fine sand grains. There are also a few very coarse sand grains present. The fabric is agglomero— plasmic and the plasmic fabric is insepic at 80X. The horizon is porous. There are also many cracks and craze planes evident. The channels are single and dendritic and the vughs are mainly irregular orthovughs. Some vugh argillans (ferri-argillans) are observed in the soil matrix. Some channels within the soil matrix have sesqui- oxide coatings. Both types 1a and 2a nodules were recognized. Their shapes are subangular to angular. 155 PLATE 5. Thin sections, B2 horizon, Njala series 156 PLATE 5 (continued) (b) 123-128cm (B22). (a) plain light, showing soil matrix (lighter portion) and Fe oxide nodule (dark portion). Note clay entry through pore in concretion; (b) X nicols, 50X. Note the flaky appearance of the oriented clay. Also, note patches of oriented clay as a result of weathering of mica within the concretion. 157 PLATE 5 (continued) 123-128cm (B22). (c) plain light, 50X. Soil matrix (lighter portion) between two ironstone nodules of Type la concretions. The white materials in the soil matrix and nodules are quartz grains. Some of the dark areas in the soil matrix are Fe and Mn concentrations. 158 Njala gravels (Plate 6), 0—35cm, A1: This is a Type la nodule which is basically a rock fragment that has been impregnated and coated with Fe oxides (Plate 6a). The rock fragment in this case is shale or mudstone, which has been impregnated with bands of Fe oxide. It contains very fine sand grains. The fabric is porphyro- skelic and the plasmic fabric is inundulic at 200x. The nodule is subrounded in shape. The perimeter is rugged. Possible edge weathering is evident at 400x. 35-53cm, A3: This is a section through a residual quartz grain. Thin bands of Fe oxide were observed between the joint planes. 53-95cm, B21(l): This resembles a Type 2b nodule with a higher percentage of skeletal grains. It is, however, possible that it is a Type la nodule derived from the weathering of sandstone (note both sandstone and shale are the common parent rocks found in this area). The fabric is intertextic and the plasmic fabric is undulic. The composition of the plasma is mainly clay coated with Fe oxide. The shape is subrounded to angular. A few metavughs are present which are lined with oriented clay. The edge is rugged in some places, and edge weathering is possible. Some oriented clay minerals were observed. 53—95cm, 821(2): The nodule is similar to that described in thin section of the A1 horizon (Plate 6a). It is a Type la nodule. The skeletal grains are randomly distributed and are of very fine sand grains. The fabric is porphyroskelic and the plasmic fabric is undulic at 200x. The plasma has a higher concentration of sesqui- oxide than that described in the A1. The Fe oxide bands are thick 159 PLATE 6a, b, c. Thin sections, Njala series gravels, Type la nodules. These are the predominant form of Njala gravels. (a) 0—35cm (Al), plain light, 100x. 160 PLATE 6a, b, c (continued) (b) 95-123cm (B21), plain light, 20X. This is basically shale or mudstone impregnated by Fe oxides. The Fe oxide bands are con— centrated on the edge of the gravel. 161 PLATE 6a, b, c (continued) (c) 95-123cm (821), plain light, 100x. _Note the similarity between this and (a). The dark areas or bands represent Fe oxide con- centration, as do the dark red specks. 162 at the edges. A few pores lined with oriented clay were observed. The shape is angular; edges are rugged, indicating possible edge weathering. 95-123cm, B21: This nodule is similar to that described above, except that it has a lower percentage of Fe oxide concentration (Plate 6b, c). The edges have the highest percentage of the con- centration. Within the matrix the Fe oxide concentration has a flecked appearance. The fabric is porphyroskelic and the plasmic fabric is undulic to inundulic at 100X. Possible edge weathering is taking place in this nodule also. 123-155cm, B22: This nodule is similar to that in the 95—123cm, B21 layer. There are fewer bands of Fe concentration and also a less flecked appearance of the Fe oxide coating of the soil plasma. The fabric is porphyroskelic and the plasmic fabric is inundulic to insepic at lOOX. There is also evidence of possible edge weathering. Makeni series (Plate 7), 35-38cm, B21: The skeletal grains in the soil matrix are randomly distributed, mostly medium to fine sand size. The fabric is agglomeroplasmic. The plasma is made of clay coated with Fe oxide. The soil matrix between the ironstone nodules shows high birefringence, possibly due to preferred orien- tation of the clay caused by the pressure of the nodules. The horizon is porous. The channels are single. There are also a few orthovughs. Orthic-pedological features are mainly sesquioxide concentrations in nodules. They vary in size. The distribution of the sesquioxide nodules is similar to that of the soil matrix. Old channels are recognized in some of the nodules. PLATE 7. Thin sections, Makeni series, B2 horizons. (a) 75-78cm (822), plain light, 40X. Soil matrix between two Type 2a ironstone nodules. The white areas around the nodules are channels. Note the similarity between the skeletal grain dis- tribution in the soil matrix and nodule. PLATE 7 (continued) (b) 75-78cm (822), plain light, lOOX. Type 2a ironstone nodule. The black area is Fe oxide. The large grains are quartz grains. Note the weathered mica pieces that have been protected. Also, note the oriented clay coated with Fe on the southwest corner near an old channel. 165 PLATE 7 (continued) (c) 115O118cm (822), plain light, 40X. Soil matrix, lighter color, and ironstone nodule. Type 2b in the center. Note channels connecting nodule and matrix. 166 75-78cm, 822: The skeletal grains of the S-matrix are randomly distributed, predominantly medium to fine sand. The fabric is agglomeroplasmic to intertextic. The plasma is made up of clay coated with Fe oxides. The plasmic fabric is insepic to inundulic at lOOX. The horizon is porous. The peds are easily recognized in the soil matrix, basically angular to subangular blocky. There are numerous cracks and craze planes. Channels are few and are mainly single channels, with a few bifurcating. Some channels between peds are coated with Fe oxides. Orthic-pedological features include Mn nodules and sesquioxidic nodules. These nodules vary in size. The distribution of the skeletal grains in the sesquioxide nodules is similar to that of the soil matrix (Plate 7a). No oriented clay is observed on the perimeter of the nodule; it is coated with sesquans. The \shape of the nodule is subrounded to subangular. Some of the nodules have a few pores with a layer of clay and Fe oxides. Weathered pieces of mica and possibly feldspar are also observable in the nodules (Plate 7b). A few Type 2b nodules are observed, which are predominantly composed of Fe oxides. 115-118cm, 822: The skeletal grains are randomly distributed in the soil matrix. They are fewer than those in the thin section of 75-78cm. There are also fewer cracks in this horizon (although a craze plane is also evident) than in the thin section of 75-78cm. The fabric is agglomeroplasmic. The plasmic fabric is inundulic to insepic at 200x. The horizon is porous. The channels follow the pattern of cracks. They are mainly single and dendritic. Orthic- pedological features are predominantly sesquioxide nodules. The 167 plasma of the nodule is predominantly Fe oxide (Plate 7c). The Fe oxide concentration within the nodules is higher than that of the above layer. The shape of the nodules is subangular to angular and the perimeter is coated with Fe oxides. The plasmic fabric of the nodule is undulic. Makeni gravels, (Plate 8) O-25cm, Al: Two areas are recognized. Area 1 consists of randomly distributed skeletal grains. The fabric is intertextic and the plasmic fabric undulic at 200x. The plasma composition is mainly Fe oxides. Area 2 consists of basically alternate bands of Fe oxide and clay, with low birefringence. The skeletal grains are very few. Area 2 is probably sedimentary in nature. The shape of the nodule is subangular to subrounded. 25-55cm, 821: The skeletal grains are randomly distributed and the fabric is intertextic. They are predominantly medium and fine sand grains, with a few large grains. The plasmic fabric is insepic at 80X. It is porous. Included as pedological features are Mn nodules. Weathered and weathering grains include mica and feldspar. Chalcedony is also present. Some of the minerals have weathered to give clay minerals which show preferred orientation in the nodule. The shape is subrounded to subangular and the perimeter is coated with Fe oxides. The nodule appears to have a rock structure. This resembles a Type 1b nodule. 55-93cm, 822: Two areas are recognized. Area 1 is similar to the nodule described in the 25—55cm, 821, in terms of the distribution of skeletal grains. The fabric is intertextic. In area 2, there are 168 PLATE 8. Thin section, Makeni series gravel. 93—115cm (822), plain light, Box. Type 2a nodule, showing edge weathering and weathered mica pieces protected within the nodule. Most of the large irregular white areas are vughs. 169 fewer skeletal grains, which are mainly fine sand. The fabric is agglomeroplasmic to porphyroskelic. The plasma composition is mainly Fe oxide coated clay. It is undulic at 80X, as compared to area 1, which is insepic to inundulic at 80X. Area 1 is more porous than area 2. Weathered mica pieces (possibly biotite) are included in the nodule. This resembles a Type lb nodule on which a Type 2b nodule has been formed. The shape is subangular to angular, with rugged edges. 93-115cm, 822: The skeletal grains are randomly distributed with the predominant size being fine sand. This concretion is Type 2b, and is similar to area 2 of 55-93cm. It contains a few channels that are single, some of which have Fe coatings, and metavughs coated with vugh sesquans. The fabric is agglomeroplasmic to por- phyroskelic and the plasmic fabric is undulic at 200x. Weathered and weathering mica pieces (Plate 8) are evident and patches of oriented clay are seen within the matrix. The shape is angular. Edge weathering is possible. 115-168cm, 822: This is similar to the section described above in both skeletal grain distribution and the nature of the matrix within the nodule. However, there are more channels in this nodule. Vughs are irregular metavughs lined with Fe oxides. Some of the channels are filled with Fe oxides. The plasmic fabric is insepic to inundulic at 100x. A few weathered mica pieces are present. The shape is angular. Timbo series, 35-65cm, A81: The skeletal grains are randomly distributed. The size is predominantly medium sand. There are a 170 few large sand grains. The fabric is intertextic. The plasmic fabric is silasepic at 200x. The plasma composition is mainly clay and Fe oxides, plus organic material. The horizon is porous. The channels are few and are single and dendritic. Cracks are numerous in this horizon. Lithorelics are mainly weathered pieces of mica and feldspar. Also observable are a few mica flakes and chalcedony. Orthic pedological features include Mn and Fe oxide nodules. The Fe oxide nodules, Type 1b, appear to have a rock structure, i.e., a piece of rock that was weathered in place with Fe oxide (Plate 9a, b). 95-100cm, 821(1): The skeletal grains are randomly distributed. There are many medium and fine sand grains, and a few coarse ones. The fabric is intertextic. The plasmic fabric is silasepic at 80x. The plasma is composed of clay and some Fe oxide. The voids are mostly of the simple packing type. The channels are single. This horizon appears dense and shows some evidence of rock structure. Lithorelics are mainly weathered pieces of mica and feldspar. Orthic pedological features are mainly Fe oxide nodules of Type lb. l70-173cm, 822: This section is similar to that of the 821 horizon. The sesquioxide nodules are fewer and are of Type 1b. Also present are plasma concentrations of Fe oxides. Rock structure is also evident. Many weathering mica pieces are present, plus some weathered feldspar. Timbo gravels (Plate 9), 0-30cm, A1: The skeletal grains are randomly distributed. The fabric is intertextic and undulic at 40X. The nodule is very porous and may be divided into two areas, one of 171 PLATE 9. Thin sections, Type 1b nodules, Timbo series. (a) 0-15cm (A1), plain light, 40x. Note rock structure in nodule. The dark brown areas represent Fe oxides. The white areas are quartz grains. 172 PLATE 9 (continued) (b) 70—115cm (821), plain light, 40X. Dark areas are Fe oxide mater— ial. White areas are quartz grains. Also, note the weathered mica flakes (pseudomorphs of mica). 173 which resembles a rock structure and contains weathered pieces of mica and also chalcedony. The other has a higher concentration of Fe oxide, making the fabric of the matrix predominantly agglomero- plasmic rather than intertextic. In some places, the perimeter is coated with Fe oxide. The edge is rugged and there is evidence of weathering at the edge. There are a few vughs that are coated with clay and Fe oxides. Most of the clay appears to have been weathered in place. The shape is subrounded to subangular. 30-48cm, A12: It is basically a rock fragment that has been weathered in place, and impregnated by Fe oxides, Type 1b. It is subangular to angular in shape. There are a few areas with plasma concentrations of Fe oxides. 48-70cm, AB: It is predominantly a rock fragment that weathered in place and has been impregnated by Fe oxides, Type 1b. The fabric is intertextic. Weatherable minerals are observed. These include exfoliated, weakly weathered mica pieces and feldSpar. A few papules (included pedorelic) are the result of mica pieces that have been weathered in place and coated with Fe oxide. The shape is angular. Some edge weathering is possible, as may be indicated by the rugged nature of the perimeter. 70-115cm, 821: It is basically a rock fragment that has weathered in place and then been impregnated by Fe oxides, Type 1b. The fabric is intertextic. It is porous and contains a few vughs lined with sesquioxide. The shape is angular and rugged, and edge weathering is observed. A patch of oriented clay is also observed at the peri- meter of the nodule. Weathered or weathering minerals are also 174 commonly present in this nodule. These include mica and possibly some feldspar (Plate 9b). 115-175cmL_822: Similar to the above nodule, this is basically a rock fragment that is being weathered in place and has been impreg- nated by Fe oxides, Type 1b. The fabric is intertextic to granular. The Fe concentration is lower than in the above-described horizon. Edge weathering is observable, but not as prominent as in the previous two nodules from the overlying horizons. Weatherable and weathered minerals include mica and feldspar. Descriptions of Thin Sections of Non- Grave11y* Soils (Segbwema, Pendembu, Masuba, Moa, Gbesebu, and Makundu) Segbwema series (Plate 10), 30-33cm, Al/BZl boundary: The skeletal grains are randomly distributed. They range from coarse to fine sand, with the medium and fine sand grains predominating. The fabric is porphyroskelic and the plasmic fabric is mainly skel- insepic (Plate 10a, b) at 100x. In some places, the plasmic fabric is mosepic. The horizon is porous. The plasma is composed of clay mixed with organic matter and Fe oxide. The channels are single, some of which are coated with a thin layer of ferri-argillans. The voids are of the compound packing type, and vughs are predominantly irregular orthovughs. A few of the pores and vughs have argillans. There are also embedded grain argillans (Plate 10a, b). * Does not contain ironstone gravel. 175 PLATE 10. Thin sections, Segbwema series, B horizon. (a) 30—33cm (Al/821 boundary), plain light, 50X. Note weathering rock fragments in the southeast corner. The plasma is high in Fe oxide as reddish brown coatings on the clay. The flaky rec- tangular materials in the matrix are weathered mica flakes. Also, note the distribution of the channels and the high amount of organic matter in the matrix. Some small channels are filled with mixtures of organic matter and Fe oxide. 176 PLATE 10 (continued) (b) 30—33cm (Al—821), X—nicol, 50X of (a). Note the few embedded grain argillans and the insepic nature of the plasmic fabric. The darkened portion of the rock fragment, southeast corner, may be due to variations in orientation of that portion. 177 PLATE 10 (continued) (c) 178 PLATE 10 (continued) (d) 33-38cm (B21t): (c) plain light, (d) X-nicol, 50x. Insepic— mosepic, plasmic fabric. Channel and vugh argillans and also embedded grain argillans. Note the high concentration of Fe oxide in the matrix and plasma concentration of clay as a result of weathering of mica in place. 179 PLATE 10 (continued) (e) 78—85cm (822t), X-nicol, 150x. Mica flakes weathering in place. Note the different weathering zones and the stages of weathering. The piece in the southwest corner seems to be the most weathered. The reddish brown zone is clay coated with Fe; clay is predomi- nantly kaolinite. The inner zones represent interlayer—illite— chlorite. The lighter speckled areas may represent gibbsite. 180 .xOON .HooHc-x .HuNNmO somm HOV mh .mcflumaumw3 mo mommum msoflum> um mmxmaw Oon .xOON .usmHH :HmHm .luNNmO EOOO-Oa Amy AcoscHucoov OH madam 181 Lithorelics included weathering primary minerals such as feld- spars and mica, some of which have weathered in place to give a plasma concentration of oriented clay minerals. There are also plasma concentrations of Fe oxide. There are quite a few primary minerals in the sand fraction. 33-38cm, 821t: This section is similar to that of the 30-33cm layer. The fabric is agglomeroplasmic and the plasmic fabric is skel-insepic to mosepic at lOOX. More channels, vughs and embedded grain argillans and ferri-argillans are present (Plate 10c, d). There are also a lot of weatherable minerals, mainly mica and feldSpar. 40-55cm, 821t: The skeletal grains are randomly distributed. The fabric is agglomeroPlasmic. The plasmic fabric is skel-insepic to insepic at lOOX. The plasma is composed of clay mixed with Fe oxides. The channels are single and dendritic and there are both orthovughs and metavughs. Some of the vughs and channels contain strongly oriented clay and also pellets of organic matter and Mn pellets. There are embedded grain argillans and also plasma concen- trations of Fe oxides. There are weathered primary minerals which produce clay minerals with preferred orientation. Weatherable minerals such as mica and feldspar are also present. 70-75cm, 822t: The skeletal grains are randomly distributed, and it is similar to the above horizons. The fabric is agglomero- plasmic and the plasmic fabric is vo-masepic. The horizon is porous. The channels are single and dendritic. There are both orthovughs -and metavughs. Some of the channels and vughs have ferri-argillans 182 lining the walls. Some vughs also contain Fe oxides. Some of the channel and vugh argillans show strong orientation. There are plasma concentrations of Fe oxides in the form of mottles. The Fe concen- tration in this horizon is higher than that of the above sections. There are weathered and weathering primary minerals, i.e., mica and feldspar, some of which have produced clay with preferred orienta- tion in the soil matrix. There are also embedded grain argillans. 78-85cm, 822t: The skeletal grains are randomly distributed. This section is similar to that of the above. The fabric is agglomero- plasmic and the plasmic fabric is insepic to vo-mosepic. The plasma is composed of a mixture of clay and Fe oxides. There are plasma concentrations of Mn and Fe oxides. The Fe oxide concentration is in the form of mottles. The channels are single and bifurcating. There are some pores that are coated with strongly oriented clay. There are a few embedded grain argillans. There are weathered primary minerals which have produced clay with preferred orientation in the matrix. Also, some of the weathered mica pieces show various weathering bands (Plate lOe, f, g). Weatherable primary minerals are also present. 98-103cm, 822t: There are many to common skeletal grains which are randomly distributed. This section appears to be more compact than the previous ones and may be transitional to the C horizon. The fabric is agglomeroplasmic and the plasmic fabric is insepic to mosepic. There are plasma concentrations of Fe oxides and some Mn. The Fe oxide is in the form of mottles. The vughs are mainly irregu- lar orthovughs and channels are single and bifurcating. There are 183 a few embedded grain argillans. There are fewer oriented clays in the matrix. Weatherable minerals such as mica and feldspar are present, and weathered pieces of these minerals can also be seen. Pendembu series, 45—48cm, A3/82 boundary: The skeletal grains are randomly distributed. The size ranges from coarse to medium and fine, with the medium sand grains predominating. The fabric is intertextic and the plasmic fabric is silasepic at 200x. The voids are predominantly single packing voids. There are some vesicles. The channels are mostly single. There are Fe oxide concentrations in the form of mottles, and some Mn nodules are also present. Weathered mica pieces (possiblywmuscovite) are present. 75—78cm, B21 horizon: The skeletal grains are randomly dis- tributed as in the A3/82 boundary. The fabric is intertextic- agglomeroplasmic and the plasmic fabric is silasepic to undulic at 200x. The voids are of the single packing type. This layer is compact, and only a few vughs are present. These are mainly irregular metavughs. Also present are vesicles. There are a few Mn nodules and plasma concentrations of Fe in the form of mottles. 93-95cm, 822 horizon: The skeletal grains are randomly dis- tributed as in the above horizons. The fabric is intertextic- agglomeroplasmic, and the plasmic fabric is silasepic to undulic at 200x. There are very few channels and vughs in this layer. It is dense and the packing voids are of the single type. The vughs are mainly_irregular orthovughs. Lithorelics include weathered pri- mary minerals, with resulting clay concentration, that shows preferred orientation. 184 There are also plasma concentrations of Mn nodules and Fe oxides in the form of mottles. 108-110cm, 822 horizon: The skeletal grain distribution is random, with the most common size being that of medium sand grains. The fabric is intertextic to agglomeroplasmic, and the plasmic fabric is undulic at 200x. The voids are of the single packing type; vughs are mainly irregular orthovughs. The horizon is dense. There are plasma concentrations of clay showing preferred orientation. Possibly these are due to the weathering of primary minerals in place. There are also plasma concentrations of Fe oxides. Masuba series (Plate 11), 23-25cm, boundary between Ap and 821t horizon: The skeletal grains are randomly distributed. The size varies from coarse to medium and fine sand, with medium and fine sand grains predominating. The fabric is intertextic and the plasmic fabric is inundulic to silasepic at 200x. The plasma is composed of a mixture of clay, with some Fe oxides and silt size organic matter. There are plasma concentrations of Mn and organic matter. The horizon is dense and pores are mostly micropores. There are a few irregular orthovughs. Lithorelics include chalcedony and weathered mica plates. 40—43cm, 821t horizon: The skeletal grains are randomly distri- buted. The predominant sizes are medium and fine sand. The fabric is porphyroskelic. In some places the skeletal grains show a clustered distribution. Some vugh ferri-argillans and embedded grain argillans are observed. The vughs are mostly orthovughs. The channels are small and single. The plasmic fabric is inundulic to 185 .GONHHon mflnu :H GOEEoo mum mcflmum Hmuoamxm 3mm .xfinumfi mcflucsouusm use Huwucmov Spas mmwufl .mcmHHfiwum cflmum pmbbwnfim cam mcmH OHDQSDOmH .xov .uzmwa cflmHm .HUHNQV EUMVIov IHfimHm pcmeHUGH .xov .HOUOGIN .Apammv 80mmlom Hbv HOV .conwuos m .moHme Obdmmz mo wonsumwm pwuomamm mo cofluomm cane .HH mfiflqm .wcflmum Hmuwamxm can mo wufim “mound .cmHHHOum smu> .xOO .HoUHc-x .AHNOO somOH-OOH mnu muoz .cmHHHoum ems> uamHmHocH new mcmHHHmum :Hmum cucumbeo 30m .xov .Hooflalx .Hmmv Edna-mo H3 H3 186 Abwzcflucoov HH madam 187 silasepic at 200x, but undulic at 400x. A few plasma concentra- tions of Fe oxides and Mn are observed. A pedorelic is in the form of anisotubule,which is composed of a few skeletal grains and unoriented clay-like material plus organic matter (Plate 11a). 50-53cm, 821t horizon: The skeletal grains are randomly distributed within the soil matrix. However, clustering of fine grains is also evident. The grains are predominantly medium to fine sand. The fabric is agglomeroplasmic and is inundulic to silasepic at 200x. The horizon is compact and the pores are mainly micrOpores. Vughs are present, most of which are irregular ortho- vughs. A few vugh ferri—argillans are present. Also, there are both void argillans and embedded grain argillans (Plate 11b) in this horizon. Pedological features include plasma concentrations of sesquioxide (in the form of mottles) and oriented clay (possibly derived in part from the weathering of mica pieces in place, litho- relics). Manganese nodules and lithorelics of chalcedony. 63-66cm, ngtfhorizon: The skeletal grain distribution is similar to that of the 50-53cm layer. Grains are randomly distributed with some clustering of smaller grains in some places. The fabric is agglomeroplasmic and plasmic fabric is insepic to inundulic at 200x. The horizon is porous and there are a few vughs, some of which are coated with ferri-argillans. There are also very few channels. The plasma is composed of silt and clay size material, which is more or less isotropic. Pedorelics include plasma concentrations of Fe * Two slides from the same depth were described. 188 oxides, Mn nodules and weathered mica and weathering mica and feld- spar pieces. The Mn nodule is subrounded to subangular and it is coated with ferri-argillans. A few embedded grain argillans are observable. Also, vugh argillans are present. 63-65cm, 822t* horizon: It is similar to 822t above. Metavughs and a few channels are present. Also present are a few embedded grain argillans, vugh ferri-argillans and channel ferri-argillans. Some of the vughs contain small pieces of quartz grains. Chalcedony is present as lithorelic. Also present are nodules of organic matter . 80-83cm, 822t horizgg; The skeletal grains are randomly distri- buted as described in the above sections. The fabric is agglomero- plasmic and plasmic fabric is undulic to insepic. There are few irregular orthovughs, some of which contain quartz crystals. The channels are few also. Some of the channels and vughs are lined with patches of oriented clay (vugh and channel argillan). There are also a few embedded grain argillans. In general, there is less illuvial clay in this layer than in the above layers. Pedorelics include Mn nodules and Fe oxide mottles. 100-103cm, 822t horizon; The skeletal grain distribution is as discussed in the pervious section. They are comparatively larger than those of the overlying B horizon. The fabric is agglomerOplasmic and the plasmic fabric is undulic to insepic. Also present in this layer are vugh argillans (Plate 11d). * Two slides from the same depth were described. 189 Moa series (Plate 12), 15-l8cm, 821: The skeletal grains are randomly distributed. The size ranges from coarse to medium and fine sand, with the latter two predominating. The fabric is porphyro- skelic (Plate 12), and the plasmic fabric silasepic-insepic at 200x. The plasma is a mixture of clay, Fe oxide and some organic matter. The horizon is porous. Voids are mainly compound packing voids. There are some irregular orthovughs and a few metavughs. The channels are mainly single with a few dendritic ones. There are plasma concentrations of Fe oxides and Mn oxides. Worm or insect castings are also observable. Lithorelics include a few weathered mica pieces. 33-35cm, 821: The skeletal grains are randomly distributed and their predominant sizes are medium and fine sand. The fabric is agglomeroplasmic to porphyroskelic and the plasmic fabric is sila- sepic to insepic at 200x. The composition of the plasma is clay, coated with Fe oxides. The voids are compound packing voids. The channels are mainly single, with a few dendritic ones. Some of the channels are filled with Mn nodules and fine quartz grains and organic matter. The horizon is porous. The vughs are mainly irregular orthovughs. There are plasma concentrations of Mn and Fe oxides. Weathered or weathering pieces of mica are observable. 43-45cm, 821: The skeletal grains are comparatively fewer than in the overlying layers and are predominantly medium and fine sand. Some large grains are evident. The fabric is agglomerOplasmic to porphyroskelic. The plasma fabric is undulic to silasepic at l60X. The vughs are mostly metavughs, some of which contain Fe oxides. A 190 PLATE 12. Thin section, Moa series, 821 horizon. 15-18cm (821), plain light, 80X. Showing Fe oxide coating on channel. Coarser and more sand grains than those in Plate 13, Gbesebu. Also, note plasma concentration of Fe oxides (north— east) and the porphyroskelic nature of the fabric. 191 thin layer of weakly oriented clay lines a metavugh. There are a few clay 'concentrations' with preferred orientation, probably due to the weather of mica 'in situ.‘ Plasma concentrations of Fe oxides are also present in this horizon. The plasma is composed of clay coated with Fe oxide. This layer is less porous. The channels are predominantly single. A channel filled with weakly oriented clay is observable. The channel appears to be below the sesquioxide coating. Some other channels are filled with Mn nodules and organic matter. Pedological features include Mn nodules and some sesqui- oxides, mainly Fe oxide concentrations (mottles) with undifferentiated matrix. The external boundary of the mottles is irregular. 53-55cm, 822: This section is similar to that from 43-45cm. The skeletal grains are fewer than in the sections described for 33-35cm. It is also denser. The channels are few and are single and dendritic. Some of these channels contain Mn nodules. The vughs are fewer than the above horizon, and are mainly irregular metavughs. The fabric is agglomeroplasmic to porphyroskelic. The plasma is composed of clay coated with Fe oxides. A few weathered mica pieces are present. 65-68cm, 822: This section is similar to that of 53-55cm. The grains are randomly distributed. The fabric is agglomeroplasmic to porphyroskelic. The plasmic fabric is undulic to silasepic at 160x. Vughs are irregular orthovughs. The channels are few, single and dendritic. There are plasma concentrations of clay, Fe oxides and Mn. The clay is coated with Fe oxides. 192 Gbesebu series (Plate 13) , 18—28cm, 821b: The skeletal grains are randomly distributed and are few in number. The sizes vary, with the very fine sand grains being most common. The fabric is porphyroskelic and the plasmic fabric is silasepic to insepic at 200x. The plasma is composed of clay mixed with Fe oxides and organic matter. The channels are single and branching, some of which are filled with organic matter (Plate 13a) and others coated with Fe oxide. The vughs are predominantly orthovughs. The matrix is generally dense. Cracks, due to shrinking of the clay, occur in this horizon. Lithorelics include weathered or weathering pieces of mica and plasma concentrations of Fe oxides and Mn oxides. The clay derived from weathered mica shows a preferred orientation. 35-43cm, 821b: The skeletal sand grains are few and randomly distributed, with very fine grains predominating. The fabric is porphyroskelic and the plasmic fabric is insepic at 200x. The plasma composition is clay coated with Fe oxides. The horizon is dense. There are a few channels and irregular orthovughs. Some of the channels are coated with Fe oxides or filled with organic matter. Lithorelics include plasma concentrations of Mn and Fe oxides (Plate 13b, c). 48-53cm, 822b: There are very few skeletal grains which are randomly distributed. The fabric is porphyroskelic and the plasmic fabric is insepic at 200x. Cracks are common and there is evidence of craze planes. The channels are single and dendritic and some metavughs are also present. A few of the vughs and channels are lined 193 PLATE 13. Thin sections, 821 horizon of Gbesebu series. (a) 18-28an(82lb),X-nicol, 80X. Note the filling of organic matter in the small channels and Fe oxide coatings on some of the ped surfaces. The reddish brown area represents Fe oxide concentra— tions. Also, note cracks or channels in the southeast corner. 194 .memam on» we ousum: canons“ ou oamwmmHHm can meccmzo mo maficwH wwwxo mm mcHsonm .xOO .uOOHH :HmHm.HbHNmOsomOumm .msm:> mum mmwum oven: wmumq .mwvflxo on can Hounds oecmmuo Spas pmHme mHmccmso .xOO .HBOHH :HOHm.HaHNmO somO-Om Awwscflucouv ma mfidqm 195 with Fe oxides. Pedological features include plasma concentrations of Mn and Fe oxides. Also weathered mica pieces are present, some of which appear like pieces of 'plasma concentration' of oriented clay. 55-60cm, 822b: This section is similar to the thin section described above. Some of the channels are coated with Fe oxides. Vughs are mostly irregular orthovughs. There are some cracks or channels. The section is comparatively denser. There are plasma concentrations of Mn and Fe oxides. Weathered and weathering mica pieces are evident. 125-130cm, 823 horizon: This section is dense with very few skeletal grains. The fabric is porphyroskelic and the plasmic fabric is insepic at 200x. There are few vughs and channels, and some of the channels contain Fe\oxides and some organic material. Thin linings of weakly oriented clay are observed in a few pores. There are plasma concentrations of Fe oxides. Weathering mica pieces are also present. Makundu series (Plate 14), 45-48cm, A8: The skeletal grains are randomly distributed and are predominantly medium and fine sand grains. The fabric is agglomeroplasmic to porphyroskelic and the plasmic fabric is inundulic to insepic at lOOX. The plasma is composed of clay mixed with organic matter and Fe oxides. The horizon is not porous. The channels are single and bifurcating and the vughs are mostly irregular orthovughs. Pellets of organic matter occur in some of the channels and also as plasma concentrations. There are also plasma concentrations of Fe oxides and patches of 196 PLATE 14. Thin section of Makundu series, 821 horizon. llO—llScm (821), plain light, 40X. Note the vughs and chan- nels, some of which contain a mixture of organic matter and Fe oxides. 197 clay with preferred orientation, possibly due to the weathering of mica pieces 'in situ.‘ 53-58cm, Bl: Skeletal grains are randomly distributed, few, and predominantly fine sand grains. There are a few coarse grains, which were also observed in the AB horizon. The fabric is porphyro- skelic and the plasmic fabric is inundulic to insepic at lOOX. Channels are few; horizon is massive. There are predominantly meta- vughs, the lining material being mainly Fe oxides mixed with organic matter. Organic matter also occurs as a pedorelic. There are also Fe oxide concentrations in the form of mottles. 58-63cm, 81: The skeletal grains are few and are randomly dis- tributed. The fabric is porphyroskelic and the plasmic fabric is insepic. The composition of the plasma is clay mixed with Fe oxide and organic matter. The horizon is dense but there are some cracks. The channels are single and bifurcating; some of them are coated with organic matter mixed with Fe oxide. There are both plasma con- centrations of Fe oxides (which appear as mottles) and Mn nodules. Organic matter also occurs as pedorelics. There are a few grain argillans, and some channels have patches of weakly oriented clay. 77-82cm, 821: The skeletal grains are few and are randomly distributed. There are also a few coarse sand grains. The predominant grain size is fine sand. The fabric is porphyroskelic and plasmic fabric is insepic at lOOX. There are some cracks, and numerous vughs, some of which are irregular orthovughs. There are a few metavughs. Some of these vughs have Fe oxide mixed with organic matter coating their walls. Some of these vughs also contain 198 Mn nodules and a few have soil material probably brought from over- lying horizons as a result of termite or worm activities. There are plasma concentrations of Fe oxide in the form of mottles. Some fecal pellets occur in the plasma as fossil formations. Weatherable minerals are present, mainly mica and a few feldspars. The channels are mainly single and bifurcating. A very few of these channels have patches of weakly oriented clay. There are also 'clay concen- trations', possibly due to the weathering of primary minerals in place, lithorelics. llO-llScm, 821 (Plate 14): The skeletal grains are few and are randomly distributed. The size is predominantly fine sand. The fabric is agglomeroplasmic to intertextic and the plasmic fabric is masepic to insepic. The plasma is composed of clay mixed with Fe and some organic matter. The horizon is less porous. The chan- nels are predominantly single, some of which are coated with organic matter mixed with Fe oxides. The vughs are predominantly metavughs, some of which contain Mn and others are lined with Fe oxides. The clay of the plasma shows orientation in some places. There are plasma concentrations of both Mn and Fe oxides which are in the form of mottles. A thin layer of oriented clay was observed. 163-170cm, 822: The sand grains are randomly distributed and they are predominantly medium and fine in size. There are more coarse grains in this horizon than in the overlying ones. It is a dense horizon with predominantly single channels. The fabric is intertextic anui the plasmic fabric is insepic. This layer resembles a 83 or C1 horizon. Also present are weatherable minerals and plasma 199 concentrations of Fe oxide and Mn. The plasma also contains some organic matter. Discussion of Thin Sections of Gravelly Soils (Baoma, Manowa, Njala, Makeni and Timbo Both thin sections from the soils as a whole and individual ironstone nodules were studied. Nodule is used instead of concre- tions in this discussion. Nodules as defined by Brewer and Sleeman (1964) are glaebules with an undifferentiated initial fabric, which include recognizable rock and soil fabrics. The glaebules in the gravelly soils meet this definition. The lack of illuvial clays in the B horizons of the soil matrix of Baoma, Manowa and Makeni series suggests that clay is not actively moving in these profiles (Plates 2, 5 and 7). The illuvial clays observed are within the ironstone nodules (Baoma, Manowa and Makeni) found in the 8 horizons (see detailed discussion below), indicating that active clay movement occurred at an earlier stage in the soil development. The presence of sesquioxide coatings in channels and on surfaces of nodules found in the sections studied suggests that some sesquioxides are moving within these profiles. Some structural develOpment was observed in the matrix of the thin sections studied, as depicted by the presence of craze planes. The structure is not massive or similar to that of rock structure, as is diagnostic of cambic horizons. These observations, therefore, suggest the presence of an oxic horizon in the control sections of these profiles, a requirement for the order Oxisols. van Vuure and Miediema (1972) reported argillans in the B horizon of the same Makeni profile that was studied in this research, and suggested the order Ultisol. This 200 study does not support this view, which is also not confirmed by clay analyses discussed earlier. In the Njala series, degraded argillans were observed in a few channels (Plate 5a, b) and on the surface of some nodules. However, the bulk of the sections observed lack these degraded argillans and it is doubtful whether the percentage of illuvial clay required for an argillic horizon will be met in the control section. Fine clay analysis (previously discussed) did not indicate the presence of an argillic B horizon in this profile. The order Oxisol is, therefore, suggested for this series also. Clearly evident in the thin section study of the Timbo profile is that clay movement is not an active process. Thin sections examined for the Timbo series were from the A and AB, 821 and 822 horizons. No illuvial clay was observed in channels or vughs. However, the structure is dense and shows some resemblance to rock structure. Also, quite a few weatherable miherals were observed in the soil matrix. Grain counts for the B horizon show enough weatherable minerals that it meets the requirement of a cambic horizon. The profile is believed to have a cambic horizon in the Upper B and, therefore, belongs in the order Inceptisol. Discussion of Ironstone Nodules (Plate 15) Thin sections of nodules from the five gravelly soil series discussed above were studied. The thin sections were made of repre- sentative nodules from each of the horizons described in these profiles. Two main types of nodules were recognized. Type 1 nodules (Plates 5a, b, c; 6a, b, c; 9a, b) have basically a rock structure. PLATE 15. Macromorphology of ironstone nodules: (a) Types 1a and b; (b) Type 2a. (a) Types 1a and lb ironstone nodules. Left to right: first, subsurface and, second, surface horizon,Njala series (Type 1a); third, subsurface and, fourth, surface (Al), Timbo series (Type lb). 202 PLATE 15 (continued) (b) Type 2a. Left to right: first, subsurface; second, surface; third, subsurface cut — Makeni series. Fourth, subsurface and, fifth, surface of Baoma series. Note the large pore on the top of the surface Baoma gravel. 203 Each is apparently a rock fragment which has been weathered in place to release Fe oxide that acts as a cementing agent or a rock fragment that is being weathered and impregnated by Fe oxide. In the latter situation, la, the Fe oxide tends to concentrate on the edges of the nodules, channels into the nodules and, in some cases, the Fe oxide appears as specks in the matrix of the nodules. The nodules from the Njala series are representative of Type la (Plates 5a, b, c and 12). In the Type lb, the Fe oxide tends to concentrate within the matrix in places where primary minerals have been weathered to release Fe oxides. This nodule is porous, possibly as a result of minerals which have been weathered out in place. This kind of nodule tends to have a high concentration of weatherable minerals within them (Plate 9a, b). The nodules from the Timbo soils are representa- tive of this kind of Type 1 nodule. Type lb nodules are what have been described as residual laterite (Alexander and CadYn 1962) . Type 2 nodule basically developed as a result of plasma concen- tration or separation of Fe oxides. Two kinds may be recognized in this type. In 2a, the matrix resembles that of the surrounding soil material (Plates 7a, b and 15). Basically, a concentration of Fe oxide has cemented the soil material together. In 2b (Plate 8c), there is basically a plasma concentration of clay or Fe oxide or clay coated with Fe oxide, with or without a few sand grains, which sometimes act as nuclei (Plate 2b, c, d, f). Type 2 may be con- sidered as indurated plinthite (ironstone), derived from mottles or plasma concentrations of clay coated with Fe oxide. A third kind, 2c, shows preferred oriented clay derived from the weathering of primary 204 minerals 'in situ' and then coated with Fe oxide. These kinds of Type 2 nodules are usually smaller than the other types. It occurs mostly in the coarse sand fraction of gravelly profiles. This frac- tion in the gravelly profiles is usually made up of 50 to 60% iron- stone nodules. In some of the above profiles, at least two of the possible kinds suggested under Type 2 occur in the same horizon (Plate 2f), i.e., the Baoma and Makeni series. It should be noted at this point that other researchers, including Alexander and Cady (l962),have recognized the two broad types of nodules similar to Types 1 and 2 discussed above. Westerveld (1969) identified two types of ironstone nodules in the Njala series based on different criteria. His type distinction is based on the external appearance of the nodules, i.e., SLC (Smooth Laterite Concretions) and RLC (Rough Laterite Concretions). The Type 1 nodules identified in this research predominate in the Njala and Timbo series. Smaller proportions of Types 2a and 2c are also present in the Njala series. Type 2 nodules predominate in the Baoma, Makeni and Manowa series. For Fe oxide concentration, two major conditions are necessary: (1) an approximately equal length of wet and dry season, and (2) fluc- tuating water tables. It seems reasonable to assume that condition (1) plus intense weathering is more prone to produce the Type 1 nodule than condition (2) and that the Type 2 nodules (especially 2a and 2b) tend to be mainly the result of condition (2). This idea may be supported by the definition of plinthite and theories concern- ing its development (Soil Survey Staff, 1975). Type 2 nodules are here considered to be true ironstone, which is the hardened form of 205 plinthite. Type 1 nodules are here considered as pseudo-ironstone. Sys (1968) suggested the name petroplinthite for ironstone nodules. In my view, this name should be applied to the Type 2 nodules dis- cussed above. The term pseudo-petroplinthite is being suggested for the Type 1 nodules. The presence of argillans (ferri-argillans) in the vughs and channels of some of the nodules studied in the Makeni, Manowa and Baoma soils and the absence of similar situations in the soil matrix proper (Plate 2d, e) may be due to one of the following reasons: (a) that the nodules are derived from an earlier soil material in which illuviation took place and have been transported and redeposited in their present location, or (b) that the nodules reflect illuvia- tion processes which took place in the profiles in an earlier stage of develOpment and have been preserved within the nodules. The roundness of the gravels in the upper horizons (Appendix A) tends to support (a) that at least the surface gravels must have been transported from another source and redeposited in their present position. Westerveld (1969) and Odell et al. (1974) have suggested that the gravels of the upper horizons are colluvial in origin. However, based on: the topographic location of the profiles; thin section studies conducted here, which show striking similarity between the surface gravels and those in the subsoil (see description of nodules, Plates 3, 6 and 7); the similarity in clay mineralogy of the A and B horizons of representative profiles (Westerveld [1969] also observed similarity in the mineralogy of the SLC and RLC gravels which he identified in an Njala profile and suggested that with passing of time the RLC would become SLC); and the slaking of the 206 nodules (surface and subsurface) after shaking with distilled water for six hours in a reciprocating shaker (Table 5), the reason proposed in (b) is more likely than (a). The appearance of the nodules in the surface horizon can be explained if the nodules (particularly Type 2) were formed below the soil surface at a time when the soil was wetter or at a zone of fluctuating water table, which became exposed at the surface, after uplift, followed by erosion. The roundness of the surface nodules is probably a result of the action of intense rainfall which caused slaking by solution or abrasion between nearby nodules or external weathering of the nodules. From the preceding discussion, the need for recognition of the ironstone nodules at some level of classification in the Soil Taxonomy (1975) arises. Recognition at the family level is hereby suggested. It is proposed that the term petroplinthic (Sys, 1968) be used in place of skeletal at the family level, if 35% or more of the coarse fragments in the upper 1.25m of the profile is composed of ironstone nodules (both Types 1 and 2), e.g., clayey-petroplinthic oxidic, etc. The term petroplinthic will also reflect the genetic nature of the greater than 2mm fraction as opposed to skeletal, which refers mainly to primary rock or mineral fractions >2mm. Discussion of Thin Sections of Non- Gravelly Soils (Segbwema, Pendembu, Masuba, Moa, Gbesebu and Makundu) Both Segbwema and Masuba soils have illuvial clays in their 8 horizons. Taking the thickness of the horizons into consideration, their 8 horizons meet the requirements of an argillic horizon. Hence, 207 the order Ultisols is suggested in view of their associated low base saturation. This observation is in agreement with the B/A and/or fine clay/total clay ratios of the two profiles. The thin section examinations of Segbwema series show many embedded grain argillans (Plate 10c, d). Plate lOe, f and 9 shows thin sections across weathered mica pieces in the B horizon of Segbwema. Several weathering zones are observed that have high birefringence of oriented clays. Many of those mica—like flakes occur in the sand fractions, especially the very fine to medium sand of the horizons identified in Figure 9c. The profile x-ray analysis of the mica-like flakes in the sand fractions showed that these flakes are not true micas, but weathered material composed of interstratified clay minerals (illite-chlorite), kaolinite and gibbsite. The bands seen in the thin section represent layers of clay minerals. The outermost layer is assumed to be kaolinite, and the inner layers are believed to be the interstrati- fied clay minerals and gibbsite. Also, embedded grain argillans are present in the Masuba series, Plate 11b. The pedotubule shown in Plate 11a is the result of soil material from an overlying horizon filling a cavity created by the activities of soil organisms. A few argillans were observed in the Pendembu and Makundu pro- files. However, the amount of illuvial clay is very small and it is doubtful whether these soils have argillic horizons. The B/A and/or fine clay/total clay ratios of these soils do not indicate the presence of argillic horizons. 208 No argillans were observed in the Moa and Gbesebu series, con— firming the absence of argillic horizons in these profiles. The dense appearance of the matrix of Pendembu, Moa, Gbesebu and Makundu suggest possible cambic horizons. Grain counts of the very fine and fine sands show the presence of a cambic horizon with weatherable minerals within the control section of the Moa, Gbesebu and Makundu profiles. Therefore, these three profiles meet the requirement of Inceptisols. The occurrence of Fe oxide on ped surfaces and in channels in Moa, Gbesebu and Makundu profiles suggests that Fe oxide is actively moving in these profiles. This is in agreement with earlier obser— vations for extractable Fe oxides in these profiles (Figure 36). Proposed Classification of the Soils Diagnostic Horizons The most important diagnostic horizons recognized in this study are umbric and ochric epipedons and argillic, oxic and cambic sub- surface horizons. Detailed discussions of these horizons are given in Soil Taxonomy (1975). Umbic epipedon: This is a surface horizon that is dark and has more than 0.6% organic C. It is usually 25cm or more thick and has a base saturation of <50% and less than 250ppm P Munsell colors 205. for this horizon are values darker than 3.5 and chromas of less than 4 when moist. Five of the eleven profiles have umbric epipedons. Ochric epipedons: This is a surface horizon that is too high in value and chroma, too low in organic C, or too thin to be an 209 umbric epipedon or other related surface horizon. Six of the eleven profiles have ochric epipedons. Argillic horizon: This is a subsurface horizon in which illuvial lattice silicate clays have accumulated to a significant extent. If an eluvial horizon remains and if there is no litho- logic discontinuity between it and the argillic horizon, the argillic horizon contains more total clay and more fine clay than the eluvial horizon. The requirements for an argillic horizon are: If the eluvial horizon has <15% total clay, the argillic horizon should contain a minimum of 3% more clay. Ratio of fine clay to total clay is greater in the argillic horizon by one-third or more than that of the overlying and underlying horizons. If the eluvial horizon contains 15-40% clay, the argillic horizon should have a minimum of 1.2 times more clay than the eluvial horizon or one-third more fine clay to total clay ratio than the overlying and underlying horizons. If the total clay content of the eluvial horizon is 40% or more, the argillic horizon should have at least 8% more clay than the eluvial horizon, or meet the requirement as described above for fine clay to total clay. If the total clay in the eluvial horizon is >60%, 8% more fine clay is required in the argillic horizon. These clay increases should occur within a vertical distance of 30cm or less, and the argillic horizon should be at least 15cm thick. Since clay has moved, the presence of clay skins in pores and on ped faces is diagnostic of argillic horizons. Also, thin sections should show 1% or more of oriented clay. 210 In the soils studied, identification of clay skins in the field was extremely difficult, as they were poorly developed. Laboratory examination of thin sections from ten of the eleven profiles studied showed that only two profiles, Segbwema and Masuba series, have enough illuviated clays to meet the requirements of an argillic horizon in their 8 horizons. Other oriented clays observed are mainly within ironstone nodules (Baoma, Manowa and Makeni series). If only the illuvial/eluvial clay* ratio of total clay is used to identify argillic horizons (Table 2), then Njala, Baoma and Manowa series will also have argillic horizons. However, the fine clay to total clay ratio together with thin section observations do not support the presence of argillic horizons in these profiles. Oxic horizon: This is a mineral subsurface horizon in an advanced stage of weathering. It is an altered horizon that is at least 30cm thick. It consists of a mixture of hydrated oxides of Fe or Al or both with variable amounts of 1:1 lattice clay, and resistant minerals such as quartz + zircon. It contains more than 15% clay and the texture should be sandy loam or finer. The fine- earth fraction has a total of less than 10me per 100gm of clay of 1N NH OAC extractable bases plus Al extractable with 1N KCl. The 4 C.E.C. of the fine-earth fraction is <16 me/lOOgm clay by NH OAC, 4 unless there is appreciable content of Al interlayered chlorite. The horizon should have less than 3% weatherable mineral, or <6% muscovite, in the 20—200u fraction and less than 5% by volume of * 1 c ayl. 211 rock structure. Boundaries between horizons are diffuse or gradual and clay skins are very few or absent. The subhorizons of all the soils studied have one or more char- acteristics of an oxic horizon. Some of them, however, have other diagnostic characteristics that exclude them from the oxic horizon or do not have all the characteristics of an oxic horizon. Baoma, Manowa, Njala, Makeni and Pendembu series all meet the above criteria for oxic horizons. Moa, Gbesebu, Makundu and Timbo soils have diag- nostic properties of oxic horizons in the subsoils but are excluded because they have too high a percentage of weatherable minerals. Moa, Gbesebu and Makundu soils have more than 6% mica, which is predominantly muscovite. Muscovite in the Makundu series is slightly weathered to give both the characteristic 10A° peak for mica and 7.2A° peak for kaolinite. The Timbo series has >6% mica in subhorizons. However, x-ray analyses (see discussion on x-ray analyses) of the flakes from the sand fractions show that the material identified by optical methods as mica is not true mica but a mixture of interstratified illite- chlorite, kaolinite, and some gibbsite. If this material is accepted as mica (for classification purposes), then Timbo series will be excluded from the oxic horizon or otherwise it will be considered to have an oxic B horizon. Cambic horizon: This is a slightly altered B horizon that lacks characteristics of an argillic horizon, and with no cementation or induration. Evidence of alteration includes strong chroma, redder hues or higher clay contents than the underlying horizons, gray colors associated with a regular decrease in organic matter with 212 increasing depth, lack of properties of umbric epipedons, textures of loamy fine sand or finer, soil structure or the absence of rock structure in at least half of the volume, and >3% weatherable minerals or >6% muscovite. Gbesebu, Moa, Makundu and Timbo* series meet the mineralogical requirement. Also, the Timbo series has >5% rock fragments in its subsoil. Classification of the Soils Baoma and Manowa series: These soils are found in soil province L on summits and convex slopes. They are well drained. Manowa soils are very gravelly (Type 2a) from the surface to depths greater than 175cm. The gravel content in the Baoma series is comparatively low, especially in the surface 58cm. Manowa has an umbric epipedon and Baoma an ochric epipedon. These soils have oxic B horizons. Manowa is classified in the clayey skeletal, oxidic, isohyperthermic family of Typic Umbriorthox. Baoma series is classified as clayey skeletal, oxidic, isohyper- thermic family of Typic Haplorthox. Njala series (P109): This soil is well drained and is found on stable upland surfaces of soil province G. It is gravelly from the surface to depths greater than 155cm. The gravel is predominantly Type la (see thin sections) - shale-like fragments that have been impregnated by Fe oxide. At lower depths (>155cm) the gravels are * Not true mica (see discussion on x-ray analyses). 213 softer. The percent gravel decreases with depth. The surface diagnostic horizon is umbric and the subsurface meets the require- ments of an oxic horizon. Thin sections show only relics of oriented clay. The percent total clay increases gradually with depth. The I/E horizon ratio of total clay increases with depth, and fine clay shows decreases with depth rather than a zone of accumulation, common in argillic horizons. The soil is classified as clayey-skeletal, kaolinitic, isohyperthermic, family of Typic Umbriorthox. Makeni series: This soil is well drained and occupies a similar physiographic position to the Njala series. It accounts for a high percentage of the upland soils in soil province J. It is gravelly from the surface to >168cm. This is mainly Type 2a gravel. Some weathered bedrock gravels (Type lb) are found in the deep layers. The diagnostic surface horizon is umbric (0-25cm) and the sub- surface horizon is oxic (25-168cm), as confirmed by thin sections and fine clay analyses. This soil is classified as clayey-skeletal, oxidic, isohyperthermic, family of Typic Umbriorthox. Segbwema series: This soil is well drained and it occurs on steep 510pes in soil province L. It has an ochric epipedon (0-33cm), comparatively high cation exchange capacity that is >16 mg/100g clay in the upper B horizon, high percentage of weatherable minerals, and an argillic horizon (33-150cm) that is identified by oriented clay in thin sections and a fine clay to total clay ratio in the argillic horizon of >1/3 that of the overlying and underlying horizons. It is classified as clayey, kaolinitic, isohyperthermic, family of 214 Orthoxic Tropudults (or may be placed in the great group Kandiu- dults [Moormann et al., 1977]). Timbo series (P19): This soil is well drained and it occurs on moderate to steep slopes. It is very gravelly from the surface to >175cm. Most of the gravels are decomposed bedrock fragments, some of which occur at the surface. This soil is high in weatherable minerals, predominantly mica.* The diagnostic surface horizon is umbric (0-48cm) and the subsoil has a number of properties diagnostic for oxic horizons. However, the greater than 5% rock structure and the high percentage of mica (if accepted as one) exclude the Timbo series from the oxic horizon. It can be classified as clayey- skeletal, oxidic, isohyperthermic family of Ustoxic Dystropepts. Ustoxic is used because Timbo has about four months of drought period (as defined earlier). If the weathered flakes are not considered as mica and the rock fragments are disregarded, then the soil can be classified as Typic Umbriorthox. Pendembu series: This imperfectly drained soil occurs on Foot- slopes and Upper Tributary Terraces. The surface horizon is ochric. Thin section studies of several layers of the B horizon did not show oriented clays.** Percent clay shows a gradual increase with depth, a pattern similar to that of Njala, Makeni, Baoma and Manowa soils. The I/E horizon clay ratio is less than 1.2 in the upper two 8 * Not true mica (see discussion on x-ray analyses). ** Clay coatings were reported observed in the field by Sivarajasingham (1968). 215 horizons. Also, the ratio of fine clay to total clay does not indi- cate the presence of an argillic horizon. Also, the percent of weatherable mineral in the 20—200u fraction is less than that required for a cambic horizon. This soil is therefore classified in the fine—loamy, oxidic, isohyperthermic family of Epiaquic Haplorthox. Masuba series (P9): This is a moderately well-drained soil that occurs on the lower part of tributary stream terraces. The texture is sandy clay loam throughout the profile, but there is higher percent clay in the B horizon that meets the requirement of I/E hori- zon ratio of 1.2 or greater in the illuvial horizon. Also, thin sections show oriented clays in the B horizon. The diagnostic sur- face horizon is ochric. This soil is classified in the fine-loamy, mixed, isohyperthermic family of Typic Paleudults. If there are >10% weatherable minerals in the 20-200u fraction, this soil would be classified as Orthoxic Tropudults. Moa series: This soil occurs on alluvial river terraces of large streams in the Moa basin of soil province L. It is moderately well to well drained. The surface horizon is ochric (0-15cm) and the subsoil has some properties of the oxic horizon. However, grain counts show >6% weatherable minerals in the 20-200u fraction. The dominant mineral is muscovite mica. The soil is classified in the fine, kaolinitic, isohyperthermic family of Fluventic Oxic DystrOpepts. Gbesebu series (N125): This soil is found on Alluvial Flood- plains and is moderately well drained. This soil has properties in 216 the B horizons that are characteristic of an oxic horizon, but has >6% mica in the 20-200u fraction, which excludes it from the oxic horizon. The mica is mainly muscovite that is slightly weathered. X-ray diffraction of the mica in the sand fraction gave the char- acteristic lOA° peak for mica and a 7.2A° peak for kaolinite, showing that part of the mica has been weathered to kaolinite. The mica observed here is different from that found in the Segbwema and Timbo series. Total clay analysis shows a slight accumulation in the B horizon. Fine clay analysis and thin section study did not support the presence of an argillic horizon. The surface epipedon is ochric and the classification is very fine,oxidic, isohyperthermic family of Fluventic Oxic Dystropepts. Makundu series (P104): This soil occupies a physiographic position similar to that of the Gbesebu. It is found in soil province J. The drainage is moderately well to well drained. The diagnostic surface horizon is umbric. The B horizon has some prOper- ties diagnostic of an oxic horizon, as in the case of the Gbesebu series, but there are >6% weatherable minerals, predominantly musco- vite mica, in the 20-200u fraction. Total clay distribution is similar to that of Gbesebu series. Thin sections did not show sig- nificant amounts of oriented clay. This soil belongs to the very fine, kaolinitic, isohyperthermic family of Fluventic Oxic DystrOpepts. General Discussion The above classifications for the gravelly soils do not take into account the presence of the ironstone nodules (petroplinthite and pseudo-petroplinthite) that are common in these soils. At the 217 family level the term skeletal, however, indicates the presence of coarse fragments >2mm, in excess of 35% in the profile. The iron- stones are different from the rock fragments found in temperate regions and in the U.S. in particular, where Soil Taxonomy (1975) was developed. These nodules are usually part of the 'soil system' as they are derived from it or reflect the genetic processes within the profile. Also, in this study it has been shown that these nodules have active surfaces (see phosphate adsorption) that can fix P. Therefore, there may be a need for the nodules to be considered in higher categories than family in Soil Taxonomy (1975). However, additional research is needed to justify their inclusion at a higher category of classification. The proposed classifications of the eleven profiles studied and their previous classifications are given in Table 9. If the proposal for modification at the family level is con- sidered, the gravelly soils would be classified as follows: Baoma series Typic Haplorthox clayey-petroplinthic, oxidic Manowa series Typic Umbriorthox clayeyfpetroplinthic, oxidic Njala series Typic Umbriorthox clayeyipetroplinthic, kaolinitic Makeni series Typic Umbriorthox clayey-petroplinthic, oxidic Timbo series Ustoxic Dystropepts clayey-skeletal, oxidic 218 xonuuonmm Hm snosmuv oabfixo .xfimoauocam muasbsoamm oamwa bazvmwmm HOmflxo Oaxo oausoo answbcom Hmumomouummo UHxObD HOV x03» mumomouumwo Hmamv oapflxo .Hmuoamxmuzmxmao IHOanED owmxa owxoums HOmHuQOUCH ownsmo Dawns: OQEHB Hmumomouum>o 0083 Moo x05 wagons—one . .OOOOOH oauflcaHomx .wwmeo nonmmm owumomoua owxozuuo HOmHuHD UHHHHmuo oeunoo maosnmwm xonuuowunfis Hmmv oabflxo .HmummemuwozmHo muHsbsonm beams Gamay Homflxo oaxo Dawns: “nexus can nuaneasmHmm xonuuoflunfia . Amoamv uHcHHomx .HmummemuxmmmHo owxosuuo owmwa Howflxo oaxo Dawns: madmz AxoauuoHunSD owmha HOV muass xonuuoflunsa AH Damsmmv oawao .Hmumamxmu>o>mHo Issoamm owxosuuo vamha HOmeO owxo bangs: 030cm: Axonuuonmm oflooooua Hoe xofiuoHdmm . «HOOOOH OHUon .Hmumawxmlwo>mHo muazbsoamm owmha owmza Homflxo Oaxo oaunoo maomm O>HHEOm HOhmH msouonsm umbho couwuom :ONHHom mmfluom ..Hm um Hambov bmmomoum mommu5mnsm moouusm aflom msowmnsm oaumocmowo oaumocuMHQ msoH>oum Amnmav >Eocoxme Hflom ocu ou mafipwooom mmawmoum co>on gnu mo cofiumowwflmmmao venomoum .m wands 219 .cofiuomou capflom .wEHmmH wusuoummfimu baaumcunom>£0mfi Ham .1 xocuuo mumwmouu -Hnbso =oHuomo -mso ono HOOHNV OHuHcHHomx .OOHO Num> -ous= poucHHm .oHucm>sHm HemHudoocH OHQEOO OHHHEO seesaw: mummmouu mumomouu -m>o 0onoo nmso 0on HONHzO UHbeo .ocflw >Hw> UHuco>sHm .owucm>=Hm HomflummocH UHQEOU oflucoo sbmmobw Hmumomouum>o UHNOpD owuco>sHm mumomouu uov xonbuoH nm>Q Oaxo Hm :QMDQMV UHuHCHHomx .mcwm imam oaumomoue .oflucm>sam HomflummocH UHQEOU oauzoo mo: mummmouu -m>o onoco mbHsOOOHmm Home OOxHe .ssmoH-OOHO =0HnucHHO= OHQNO HOOHOHO UHHHHOHO oHuaoo magma: OmHHEmm HOOmH mdoumnsm umbuo coufluom couawom mmflumm ..Hm um HHOUOV bomomoum oommudmbsm mommusm Hflom msoumbsm oaumocmmwo oaumocmmflo msoH>oum Aposcflucoov m wanna 220 Proposed Hypothesis for the Genesis of the Soils Differences in stages of development of the soils studied are the results of differences in kinds of parent materials and the rate and direction of physical, chemical and biological processes. Three main stages in the development of the profiles are pro- posed. These are: (l) weathering of primary silicate minerals, (2) formation and illuviation of clay, and (3) accumulation of Fe and the formation of plinthite and ironstone nodules. Weathering of Primary Silicate Minerals The rate of weathering in Sierra Leone is very rapid. This is favored by the high temperatures and abundant rainfall. Primary silicates are rapidly weathered to release bases (mainly Ca and Mg), Si, Al, and Fe. The bases are easily lost from the soil, by leaching with excess moisture. The Si and Al combine to form 1:1 secondary clay minerals (mainly kaolinite). Kaolinite is the predominant clay mineral present in the soils studied. The changes from primary silicate to kaolinite can take place within a short distance (Plate 10e, f, g). Two diagnostic soil horizons are related to the degree of weathering. These are the cambic and the oxic horizons. A cambic horizon represents a lesser degree of weathering and it is characterized by the presence of appreciable amounts of weatherable minerals. An oxic horizon repre- sents a strongly weathered condition. It is characterized by the presence of few or no weatherable minerals and low cation exchange capacity. The soils of the Upland Surfaces of Highly Weathered Material (Baoma, Manowa, Njala and Makeni series) have oxic horizons. These 221 soils are on stable landscapes, on which infiltration predominates. The abundant supply of moisture that moves through the profile, coupled with high temperatures, resulted in the strongly weathered profiles of these soils in which the released bases and Si have been depleted. In the soils of the Steep Hills and Slopes (Segbwema and Timbo series), runoff predominates over infiltration. This resulted in less weathered and developed profiles, where appreciable amounts of weatherable minerals are still present. This explains the presence of a cambic horizon in the Timbo series. Weatherable minerals are also present in the Segbwema series. However, profile development has advanced to the stage at which illuvial clay is present in the B horizon. Erosion is common on soils of the Steep Hills and Slopes. New surfaces are continuously exposed, which in some cases contain rock fragments, e.g., Timbo series. These exposed rock fragments are then subjected to the sequence of weathering and release of bases. Si, Al, and Fe . Under good vegetative cover, erosion is reduced and more moisture tends to move through the profile. This gives rise to a more advanced stage of profile development, similar to that of the Segbwema series. The presence of an oxic horizon in the Pendembu series of the Colluvial Footslopes and Upper River Tributary Terraces is a reflec- tion of the nature of the weathered material from which this soil is developed. The soils of the Alluvial Terraces and Floodplains (Moa, Gbesebu and Makundu series) have cambic horizons. This reflects the 222 nature of the eroded materials derived from the catchment area. They include weathered materials high in kaolinite and sesquioxides but also some weatherable minerals from soils on the steep slopes. Illuviation of Clay Two processes are involved in the formation of an illuvial clay horizon. These are: (l) eluviation of clay from the surface horizons and (2) accumulation of the eluviated clay (illuviation) in the subsoils. Eluviation is favored by silica as a dispersing agent, or lack of excess Fe and excess infiltration over runoff and evaporation. This condition prevails during the rainy season. Illuviation (depo- sition from suspension) is favored by dry conditions. Both wet and dry conditions occur in Sierra Leone, giving rise to the development of eluvial and illuvial horizons. The illuvial horizon so formed is referred to as the argillic horizon. This horizon is characterized by the presence of argillans or clay skins on ped surfaces. In the soils of the Steep Hills and Slopes, an argillic horizon is present in the Segbwema series. In the soils of the Colluvial Footslopes and Upper River Tributary Terraces, Masuba series also has an argillic horizon. The presence of an argillic horizon represents an intermediate stage between the cambic and the oxic subsurface development of the profiles studied. In the soils of the Upland Surfaces of Highly Weathered Material (Baoma, Manowa, Njala and Makeni series), illuvial clays are believed to occur at depths greater than five feet. The possible 223 reason is the result of deep movement of percolating water carrying eluvial clay, during the wet season. Accumulation of Fe and the Formation of Plinthite and Ironstone Nodules Intense weathering of silicate minerals releases, among other things, Fe. Mobilization of the released Fe is favored by low redox potential, and this condition can be achieved under fluctuating water table or waterlogged conditions. In the soils studied, mobilization of Fe is favored during the rainy season. The transportation of Fe occurs within relatively short distances in the profiles. During dry periods the mobilized Fe is oxidized (usually around large pores). This oxidized Fe subsequently accumulates to give bright mottles. Plinthite is a more strongly developed mottled horizon which in addition to Fe oxides includes highly weathered mixtures of clay, quartz and other diluents. It is usually red in color. It becomes irreversibly hardened on exposure to wetting and drying. During the hardening process segregation and crystallization of Fe occurs. The hardened mottles are referred to as ironstone nodules. High rain- fall and prolonged dry seasons, such as those in Sierra Leone, are favorable for the formation of plinthite and ironstone. In the soils of the Upland Surfaces of Highly Weathered Material (Baoma, Manowa, Njala and Makeni), ironstone nodules occur from the surface horizon to depths greater than five feet. Below this depth the nodules are soft. The occurrence of ironstone nodules at the surface may be the result of: (l) transportation and deposition from another source or (2) erosion of surface gravel-free layers, to eXpose subsoil material. d 224 The round shape of the surface gravels tends to support the idea of transportation from other sources (Sivarajasingham, 1968; Westerveld, 1969). Data obtained in this study and discussed earlier, however, support the idea of erosion and in situ development of the ironstone nodules. Also, argillans described in old channels present in the iron- stone nodules of Baoma, Manowa and Makeni series reflect the past genesis of the profiles. They also suggest that Fe accumulations in these profiles occurred at a later stage in the profile develop- ment,than illuviation of clay. The removal and accumulation of Fe represents an advanced stage in the profile development of the soils. At this stage, excess Si released by weathering of primary silicates has been removed from the profiles. Fe movement seems to be an active process in the soils of the Alluvial Terraces and Floodplains (Plates 12 and 13). The sequence in the genesis of the profiles studied can be summarized as follows: (1) Parent material + cambic horizon -+ argillic horizon rich in primary (Inceptisol) illuviation (Ultisol) silicates + oxic horizon removal of Si, movement and accumulation (Oxisols) of Fe.Extreme weathering, and breaking down of kaolinite to release gibbsite (2) Parent material + cambic horizon + oxic horizon rich in kaolinite (Inceptisols) (Oxisols) and sesquioxides The second sequence is proposed for soils of the Alluvial Terraces and Floodplains. 225 Significance to Land Use and Management The usefulness of the research conducted here to characterize the soils depends on its applicability for crOp production and other land use practices adopted by landowners and planners. At the lower levels of classification, soil properties that sig— nificantly influence the use and management of the soils are taken into consideration. These properties include: slope (erosion hazard), surface texture (including gravels or stones), unfavorable layers such as gravel or rock, moisture holding capacity, nutrient status, and wetness. These may be used as phase separations at any category level with Soil Taxonomy (1975). In the soils studied, the most important prOperties that require attention, aside from nutrient status, are erosion hazard, soil moisture holding capacity and unfavorable layers of gravels (including surfaces). Erosion Hazard This is a serious problem in Segbwema and Timbo series. These soils are well drained and are usually located on slopes >6%. This poses a moderate to severe erosion hazard, particularly if the land is cleared. The problem is more severe in Segbwema series, which is usually on steeper slopes than the Timbo series. Also, the high per- centage of surface gravels in the Timbo series absorbs the impact of the intense rain characteristic of Sierra Leone, thereby reducing the detachability of the finer soil particles, hence minimizing soil loss. These soils preferably should be left under continuous vege- tative cover, as they are too steep and erosive for cultivated crops. 226 Erosion hazard is not a serious problem in Baoma, Manowa, Njala and Makeni series, especially in those that occur on 0—3% slopes. However, on the sloping variants of these soils (3-15% slopes), continuous cultivation is not recommended as it will cause serious erosion problems, despite the surface gravels. A modified form of shifting cultivation, with longer fallow periods, will help to build a more stable soil structure and minimize erosion. Since a permanent vegetative cover is necessary to control erosion on these soils, forestry or permanent tree crOps such as coffee, cocoa (Plate lb) or citrus can be cultivated. Pendembu, Masuba, Moa, Gbesebu and Makundu series all have little or no erosion problems. They are imperfectly to well drained soils and are found on fairly flat land surfaces (0-3% slopes). However, soil loss by erosion could be a serious problem if the surface soils are not properly managed, i.e., by correct timing of tillage operations. Destruction of surface structure and creation of a compact layer below the plow layer will reduce infiltration and cause surface runoff, particularly in Gbesebu, Makundu, and Moa series that have very fine textures. Available Moisture Holding Capacity This is an important factor that affects the use of the soils studied for agricultural production. Soil texture and organic matter content are soil properties that influence the amount of water that is retained against gravity and that will become available for utili- zation by plants. These properties have been discussed previously and are listed in Tables 2 and 3. Available moisture capacity is greatest in soils that contain much silt. The Segbwema series has “‘I— 227 a medium water holding capacity (Table 10) and experiences a drought period of three months. Timbo soil is very gravelly and has a more sandy surface horizon than Segbwema series. As a result of these two factors, it retains less water for utilization by crops. Timbo has a very low available moisture holding capacity and eXperiences four months of dry period. Manowa, Njala and Makeni series have clay to clay loam textures in their control sections but are also very gravelly (SO-80% gravels by weight). The latter characteristic con- tributes to the fact that these soils have a low or very low available water holding capacity. They experience about four to five months of drought period. Cultivation of a second crop annually is possible in these soils with irrigation. This, however, is a very expensive operation and is not recommended, as these soils have low natural productivity. The Baoma soil is less gravelly, has a low moisture holding capacity and a drought period of about four months. The Pendembu and Masuba series have comparatively sandy profiles. They are imperfectly drained and moderately well drained, respectively. Masuba has medium water holding capacity and a drought period of three months. The water holding capacity of Pendembu series is slightly lower than that of Masuba. Also, Pendembu series is water- logged for about one to two months during the rainy season, and only experiences a short drought period. Both of these soils are naturally poor in nutrients. With adequate fertilization they can be used for continuous crop production. A second crop is possible annually with irrigation. The Moa, Gbesebu and Makundu series all have medium or high available moisture holding capacity and a short drought period of 228 Table 10. Groupings for available moisture capacities* Units of avail- able moisture Available moisture capacity per holding capacity to unit of depth 1524mm (60 inches) Designations _ (mm) (inches) Soil Series Very low 0.01 - 0.03 15-51 0.6-2 Manowa, Makeni and Timbo Low 0.03 - 0.06 51—102 2-4 Baoma, Pendembu and Njala Medium 0.06 - 0.11 102-178 4-7 Makundu, Moa, Masuba and Segbwema High 0.11 - 0.18 178—279 7-11 Gbesebu Very high 0.18 - 0.25 279-381 ll-lS * Taken from Odell et a1. (1974) 229 one to two months. The textures are clayey (very fine or fine), but their silt contents are moderate (15, 30 and 26%, respectively). Gbesebu and Makundu soils may be flooded for short periods during the rainy season, and Moe may be waterlogged for a short period. These soils are the most productive soils among the groups studied. Two crops can be grown in these soils annually with minimum irrigation. Influences of Ironstone Gravel (Unfavorable Layer of Gravel) This is of importance in Baoma, Manowa, Njala, Makeni and Timbo soils. The serious effects of the gravel on crop production are: (a) inhibit proper develOpment of roots, (b) reduce the contact between roots and the finer soil particles, and (c) contribute to the reduction of the amount of available water that is retained by the soil per unit volume. The gravels do not, however, pose serious problems to plowing, because of their comparatively small size (about 3cm or less in diameter). The above limitations affect the productivity of these soils and the amount of nutrients that can be retained for plant use. Since it has been shown in this study that the gravels retain very little soil P, phosphate fertilization rates should be adjusted to take the gravels into account. Since most of the fertilizer recom- mendations for these soils are based on analyses of the <2mm fraction, the diluting effect of the gravels should be taken into consideration. A significant advantage of the gravels is their mulching effect on the surface soils, against intense rainfall. The gravels adsorb the impact of the rain, thereby reducing the severe effect on the 230 erodability of the soils. They also increase infiltration and decrease runoff, as well as decrease erosion losses. Another possible advantage is the decrease in the amounts of PO adsorption 4 per unit volume of plow soil (see discussion on P fixation). Suggested Crops and Their Management Needs The most important single crop grown in Sierra Leone is rice. It is usually cultivated on uplands as well as in swamps. Among the soils studied here, the best upland rice producing ones are the Gbesebu, Makundu, Moa, Masuba and Pendembu series. Continuous crOpping can be practiced on these soils, and irrigation is not necessary during the rainy season. Since the soils are acid and poor in nutrients, a program of fertilization will be necessary for optimum yields. In situations where these soils are periodi- cally flooded or waterlogged, bunding to control flood water may be necessary. Njala, Makeni, Manowa and Baoma soils can also be used for rice production. However, since these soils are very poor in nutrients and there is also a higher risk of erosion hazard, organic matter management is very important. Continuous cultivation is not recom- mended. The traditional shifting cultivation with fallow periods is strongly recommended, as it will improve the organic matter and nutrient supply and also minimize erosion. Timbo and Segbwema soils are the least suitable for rice produc- tion because of the high erosion risks. Other crops, such as maize, coffee, cocoa, groundnuts and cassava, grow well under good management on Gbesebu, Moa, Makundu, 231 Masuba and Pendembu soils. Cassava grows well on Baoma, Manowa, Njala and Makeni soils, but the size and shape of the tubers are sometimes adversely affected by the gravels. A crop rotation system of rice followed by groundnuts, or rice followed by forage, may be useful for the above soils if they are to be continuously cultivated. Permanent tree crops (such as oil palm, coffee or cocoa), pasture, or forestry are recommended for sloping phases of Baoma, Manowa, Njala and Makeni soils and also for Segbwema and Timbo soils. Additional Research Needs Insufficient information is presently available on the charac- teristicsof soils in Sierra Leone to aid in their proper placement in the Soil Taxonomy. Some of the needed characterizations for selected soils have been looked into in this study, and some problems that need additional work have emerged. It is shown in this study that removal of free Fe oxides from soils before particle size analyses is superior to not removing the Fe, and it also gives a better indication of the genesis of the profiles. This procedure is laborious. Therefore, working out a factor that can be used to multiply the moisture content at 15 bar to give an estimate of the percent clay can be meaningful. For testing that possibility, additional representative profiles would have to be studied. The possible use of ratios of the difference between extractable dithionite Fe and oxalate Fe to total clay (Fe removed) in the 203 separation of Oxisols from Inceptisols or Ultisols in Sierra Leone 232 has been suggested. Additional work is required to study the extent to which this ratio is useful. Ironstone nodules in the gravelly soils studied have been shown to adsorb P. However, the nature of the adsorption mechanism is not certain and additional research is needed to determine the nature of the adsorption process, whether it is a surface adsorption or diffusion of P into the nodules through micropores on the surface. If the latter mechanism is predominant, the gravels may result in much more P adsorption than this study indicates. It is important also to evaluate this relative to crop responses and fertilizer requirements for efficient production. These ironstone nodules have been shown to slake on shaking in a reciprocating shaker. It would be interesting to know whether this same process takes place in the field under intense rainfall or whether certain management practices foster this slaking. Could this mechanism account for the lower percentage of gravels in the upper soil horizons? Or are they being diluted by concentrations of fine-earth in the plow layer by soil fauna, particularly termites, ants and earthworms? This may be important as some of these nodules have been shown to have primary minerals that have been protected from weathering. Finally, x-ray analyses of mica-like flakes in the sand frac- tions of some of the profiles have shown that these flakes are not true mica. It would be useful to know the state of mica-like flakes in other soils, and in which sand fractions true micas exist. This information would be useful for proper classification of the soils, and to assess their agricultural potential and management needs. SUMMARY AND CONCLUS ION Eleven representative profiles from four physiographic groups in Sierra Leone were studied. Figure 8 shows the relationship between the physiography and the classification of the soils. The soils of the Upland Surfaces of Highly Weathered Material are all Oxisols. Segbwema and Timbo series of the Steep Hills and Slopes are Ultisols and Inceptisols, respectively. On the Colluvial Foot- lepes and Tributary Terraces, Pendembu soils belong to the order Oxisols and Masuba series to the order Ultisols. All the soils of the Alluvial Terraces and Floodplains belong to the order Inceptisols. Other significant observations and relationships include the following. The percentage clay values are higher when free Fe oxide is removed before particle size analyses than when the free Fe oxide is not removed. The percentage clay values obtained with free Fe oxide removed are also commonly higher than the clay values estimated by the factor 2.5 x 15 bar moisture content. The value of the sum of percent clay (after Fe removed) + percent Fe203d would require a factor of 3.0 x 15 bar H20 to estimate clay content in the soils studied. Some differences in the vertical distribution pattern of percent clay with depth were observed after the removal of free Fe oxide. The percentage fine clay, <0.2u, is usually higher than that of coarse clay. The ratio of fine clay to total clay for soils of the Upland 233 234 OXISOLS INCEPTISOLS ULTISOLS ULTISOLS OXIS OLS INCEPTISOLS Upland Surface of Highly Weathered Material Steep Hills and Slopes Colluvial Footslopes and Tributary Terraces UOUJD Alluvial Terraces and Floodplains Figure 8 Relationship between physiography and classification of soils. 235 Surfaces of Highly Weathered Material (Baoma, Manowa, Njala and Makeni) decreases with depth. A similar pattern is observed for the Pendembu series on the Colluvial Footslopes and Tributary Terraces. All of these soils belong to the order Oxisols. The silt/clay ratio decreases with removal of free Fe oxide. The ratio is the lowest in the soils of the Upland Surfaces of Highly Weathered Material, suggesting a more advanced stage of ferrallitic- pedogenesis in these profiles. The highest ratio is obtained for the Segbwema series of the Steep Hills and Slopes, indicative of less ferrallitic weathering, suggesting it may be the youngest soil profile studied. Percent sand and silt also tends to decrease with the removal of free Fe oxides. The 15 bar moisture content tends to follow the same distribution pattern as the clay content. Within each of the four groups of soils-studied, percent organic matter in the surface horizon increases with the moisture regime from ustic to udic. There is also a significant correlation (0.89) at the 1% level between percent clay (Fe oxide removed) and organic C for the surface horizon. The critical value of 516 me C.E.C./100 g clay seems to give a clearer split of the oxic and non-oxic soils in this study than does a critical value of 12 me C.E.C./100 g clay. The critical value of $10 for the ratio of the sum of the exchangeable cations plus exchangeable Al to percent clay (also one of the criteria used to identify an oxic horizon) does not seem to be useful in the separa- tion of the soils studied. Values of <10 are also obtained for soils with known argillic or cambic horizons, e.g., Segbwema and Masuba or Gbesebu series. 236 Dithionite-citrate-bicarbonate solution extracted more Fe oxide than did acid ammonium oxalate and Na—pyrophosphate solutions. Acid ammonium oxalate did not consistently extract more Fe oxide than Na- pyrophosphate solution. In the soils of the Upland Surfaces of Highly Weathered Material, dithionite extracted Fe increased with depth in the profile. In the soils of the Steep Hills and Slopes, dithionite Fe decreased with depth in the Segbwema series and showed a subsoil maximum in the Timbo series. Zones of maxima of dithionite extracted Fe oxide are also present in the soils of the Colluvial Footslopes and Upper River Tributary Terraces and soils of the Alluvial Terraces and Floodplains. The distribution pattern of oxalate extracted Fe oxide is a gradual decrease with depth in soils of the Upland Surfaces of Highly Weathered Material, and for soils of the Steep Hills and Slopes. In the soils of the Colluvial Footslopes and Upper River Tributary Terraces, oxalate extracted Fe oxide decreased with depth in the Pendembu series and remained fairly constant in the Masuba series. Zones of maxima are present in the soils of the Alluvial Terraces and Floodplains. In general, the distribution pattern for Na-pyrophosphate extract- able Fe oxide showed a subsoil maxima in all the soils studied except for the Segbwema and Baoma series. The correlation between percent organic C (Walkley-Black) and Na-pyrophosphate extracted Fe oxide is low. The distribution pattern of dithionite extracted Fe oxide appears to be useful in distinguishing the soils studied in relation to 237 geomorphic surfaces. The ratio of Fe203ox/Fe203d appears to be useful in determining the degree of profile development. The ratio of the difference between dithionite extractable Fe oxide and oxalate extracted Fe oxide to that of percent clay (Fe removed) seems to have some promise as one of the criteria in sepa- rating the oxic and non-oxic soils. A critical value of 0.06 (6%) is suggested for the soils studied. The amount of P adsorbed by the ironstone nodules of the Njala and Makeni series is comparatively much smaller than that adsorbed by the respective fine-earth fractions. The amount of P adsorbed by the nodules increased with time of equilibration. The ironstone nodules tend to have a diluting effect on the amount of P that is fixed by the whole soil, e.g., 5325kg/hec P can be fixed by the fine- earth of the surface horizon of the Makeni series and the gravels fixed 203kg/hec P in 48 hrs. The combined value is reduced to 1597kg/hec if the diluting effect of the ironstone nodules is taken into consideration. In the Makeni series, the ironstone nodules account for 8.2% of the P that can be fixed by the whole surface soil and 7;5% of the whole subsoil. The gravels are 70% and 76% of these soil layers. In general, in the Makeni, Segbwema and Gbesebu series, P adsorbed by the surface soil is higher than that adsorbed by the subsurface soil. The reverse is true in the Masuba series. The influence of Fe oxide (by dithionite and oxalate) on the adsorption of P is greater in the subsurface horizon than in the surface horizon, where organic matter, A1 oxide (oxalate) and organic complexes of Fe and A1 are more important. 238 The predominant clay mineral in the total and fine clay frac- tion, as revealed by x-ray diffractograms, is kaolinite, which is less ordered in the fine clay. There is a striking similarity between the clay minerals of the ironstone nodules of the Makeni and Njala series and that of their respective fine-earth fractions. X-ray analyses of mica-like flakes in the medium and fine sand of the A, B and C horizons of the Segbwema series and fine sand of the B horizon of Timbo series show that the flakes are mainly interlayer {illite—chlorite' and kaolinite. The interlayer material does not readily fix potassium. Thin sections of the soil horizonssampled showed that most of them lack an argillic horizon. Thin sections of the ironstone nodules revealed two main types. Type 1 is derived from rock frag- ments and Type 2 is probably derived from plinthite. Subdivisions within the types are also recognized. The subdivisions in Type 1 represent nodules derived from granite and mudstones. In Type 2, the two main subdivisions represent nodules with similar matrix to that of the surrounding soil and nodules which are basically clay concentrations that have been coated with iron oxides. Argillans that have been protected in the nodules are observed along old channels, indicating past genetic processes in the profiles. Also, there is a striking similarity between the internal S-matrix of the surface and subsurface nodules of the gravelly soils studied. The roundness of the surface gravels is more likely the result of solution than of transportation. The term petroplinthic is suggested in place of skeletal at the family level of classification, when the >2mm fraction is mainly >35$ 239 ironstone nodules, to reflect the genetic nature of these nodules. Soils of the Upland Surfaces of Highly Weathered Material are less suited for continuous cultivation than soils of the Colluvial Footslopes and Tributary Terraces and soils of the Alluvial Terraces and Floodplains. The soils of the Steep Hills and Slopes are the least suited for continuous cultivation, because of serious erosion hazards. Proper organic matter management is important for soils with oxidic mineralogy. Percentage fine and total clay (when Fe oxide is removed), free Fe oxide/clay ratio (Fe removed), and thin section studies appear to be three of the most important bases for classifying the soils studied here within Soil Taxonomy (1975). BIBLIOGRAPHY BIBLIOGRAPHY Aleksandrova, L. N. 1960. The use of sodium pyrophosphate for isolating free humic substances and their organic-mineral compounds from soils. Soviet Soil Sci. 2:190-197. Alexander, E. B. 1974. Extractable iron in relation to soil age on terraces along the Truckee River, Nevada. Soil Sci. Soc. Amer. Proc. 38:121-124. Alexander, L. T. and J. G. Cady. 1962. Genesis and hardening of laterite in soils. USDA Tech. Bull. 1282:1-89. Ashaye, T. I. 1969. Sesquioxide status and particle size distri- bution in twelve Nigerian soils derived from sandstones. African Soils 14:85-94. Aubert, G. 1963. Soil with ferruginous or ferralitic crust of tropical regions. Soil Sci. 95:235-242. Baldwin, M., C. E. Kellogg and J. Thorp. 1938. Soil Classification. Agric. Yearbook. Soils and Man. 979-1001. Barshad, I. 1964. Chemistry of Soil Development, p. 1-70. In F. E. Bear (ed.), Chemistry of the Soil. (Reinhold, New York). Bascomb, C. L. 1968. Distribution of pyrophosphate extractable iron and organic carbon in soils of various groups. J. Soil Sci. 19:251-268. Bennema, J. 1963. The red and yellow soils of trOpical and sub— tropical uplands. Soil Sci. 95:250-257. Black, C. A. (ed.) 1965. IHethods of Soil Analysis. Parts I and II. Agronomy Monograph 9. Amer. Soc. Agron. 1,572 p. Blume, H. P. and V. Schwertmann. 1969. Genetic evaluation of profile distribution of aluminum, iron and manganese oxides. Soil Sci. Soc. Amer. Proc. 33:438-444. Brewer, R. 1956. A petrographic study of two soils in relation to their origin and classification. J. Soil Sci. 7:268-279. Brewer, R. 1964. Fabric andIVineral Analysis of Soils. R. E. Krieger Publishing Co., Huntington, New York (1976 reprint). 482 p. 240 A 241 Brewer, R. and J. R. Sleeman. 1964. Glaebules: their definition, classification and interpretation. J. Soil Sci. 15:66-78. Brown, G. 1961. The.X-ray Identification and Crystal Structures of ClayIVinerals. Mineralogical Society, London. 543 p. Bromfield, S. M. 1965. Studies of the relative importancecyf iron and aluminum in the sorption of phosphate by some Australian soils. Aust. J. Soil Res. 3:31-44. Buol, S. W. and F. D. Hole. 1961. Clay skin genesis in Wisconsin soils. Soil Sci. Soc. Amer. Proc. 25:377-379. Buol, S. W. and F. D. Hole. 1959. Some characteristics of clay skins in the B horizon of a Gray-Brown Podzolic soil. Soil Sci. Soc. Amer. Proc. 23:239-241. Buol, S. W., F. D. Hole and R. J. McCracken. 1973. Soil Genesis and Classification. The Iowa State University Press, Ames. 360 p. Cady, J. G. 1965. Petrographic microscope techniques, pp. 604-631. In C. A. Black (ed.), Methods of Soil Analysis (Part I). Agronomy Monograph 9. Amer. Soc. Agron., Madison, Wisconsin. Caillier, S. and S. Henin. 1963. Mineralogie, des Argiles. Masson et Cie. Paris-Vie. Clarke, J. I. 1966. Sierra Leone in Maps. University of London Press Ltd. 119 p. Cline, M. G. 1949. Profile studies of normal soils of New York. I. Soil profile sequences involving Brown Forest, Gray-Brown Podzolic and Brown Podzolic soils. Soil Sci. 68:259-272. Crompton, E. 1960. The significance of the weathering/leaching ratio in the differentiation of major soil groups with par- ticular reference to very strongly leached Brown Earths on the hills of Britain. Trans. 7th Int. Congr. Soil Sci. 4:406-412. Daniels, R. 8., H. F. Perkins, 8. F. Hajek and E. E. Gamble. (In press). Morphology of discontinuous phase plinthite and criteria for its field identification in the southeastern U.S. Denison, I. A., W. H. Fry and P. L. Gile. 1929. Alteration of musco- vite and biotite in soils. USDA Tech. Bull. 128 p. Desphande, T. L., D. J. Greenland and J. P. Quirk. 1964. R018 of iron oxide in bonding of soil particles. Nature, London 201:107-108. 242 Desphande, T. L., D. J. Greenland and J. P. Quirk. 1968. Changes in soil properties associated with the removal of iron and aluminum oxides. J. Soil Sci. 19:108-122. De Villiers, J. M. 1965. Present soil forming factors and processes in tropical and subtropical regions. Soil Sci. 99:50-57. D'Hoore, J. 1954. Proposed classification of the accumulation zones of free sesquioxides on a genetic basis. Soils Afr. 3:66-81. D'Hoore, J. 1955. The description and classification of free sesqui- oxide accumulation zone; Trans. 5th Int. Congr. Soil Sci. 4:39-44. Dickman, S. R. and R. H. Bray. 1940. Colorimetric determination of phosphate. Ind. Eng. Chem. Anal. Ed. 12:665-668. Dixey, F. 1922. The physiography of Sierra Leone. Geog. Jour. 60:41-65. Ellis, 8. S. 1952. Genesis of a tropical red soil. J. Soil Sci. 3:52-62. Eswaran, H. 1972. Micromorphological indicators of pedogenesis in some tropical soils derived from basalts from Nicaragua. Geoderma 7:15-31. Eswaran, H. and N. G. Raghu Mohan. 1973a. The microfabric of petro- plinthite. Soil Sci. Soc. Amer. Proc. 37:79-82. Evans, J. W. 1910. The meaning of the term Laterite. Geol. Mag. (V) 7:189-190. F.A.O., UNESCO. 1974. Soil Map of the WOrld l:5,000,000. Vol. 1 Legend, Paris. Fermor, L. L. 1911. What is Laterite? Geol. Mag. 5:454-462. Fox, C. S. 1936. Buchanan's Laterite of Malabar and Kanara. Records Geol. Survey India 69 (4):389-422. Fox, R. L. and E. J. Kamprath. 1970. Phosphate sorption isotherms for evaluating phosphate requirements of soils. Soil Sci. Soc. Amer. Proc. 34:902-907. Franzmeier, D. P., B. F. Hajek and C. H. Simonson. 1965. Use of amorphous material to identify spodic horizons. Soil Sci. Soc. Amer. Proc. 29:737-743. Friedman, G. M. 1958. Determination of sieve-size distribution from thin section data for sedimentary petrological studies. J. Geol. 66:394. 243 Gallaher, R. N., H. F. Perkins and K. H. Tan. 1974. Classification, composition and mineralogy of iron glaebules in a southern coastal plain soil. Soil Sci. 117:155-165. Grim, R. E. 1968. Clay Mineralogy (2nd ed.). McGraw—Hill Book Co., London. 596 p. Hallsworth, E. C. 1963. An examination of some factors affecting the movement of clay in an artificial soil. J. Soil Sci. 14:360-373. Hamilton, R. 1964. A short note on droplet formation in iron crust, p. 277-298. In Jongennus (ed.), Proc. 2nd working' Meeting on Soil Micromorphology. Elsevier, Amsterdam. Harter, R. D. and D. E. Baker. 1977. Application and misapplication of the Langmuir equation to soil adsorption phenomena. Soil Sci. Soc. Amer. Proc. 41:1077-1080. Haseman, J. F., E. H. Brown and C. D. Whitt. 1950. Some reactions of phosphorus with clays and hydrous oxides of Fe and Al. Soil Sci. 70:257-271. Ho, R. 1957. Corrugated iron pan. Nature (London) 180:1276-1277. Holland, T. H. 1903. On the constitution, origin and dehydration of laterite. Geol. Mag. (IV) 10:59-69. Hsu, P. H. 1964. Adsorption of phosphorus by aluminum and iron in soils. Soil Sci. Soc. Amer. Proc. 28:474-478. Hsu, P. H. 1965. Fixation of P by Al and Fe in acid soils. Soil Sci. 99:398-402. Humbert, R. P. 1948. The genesis of laterite. Soil Sci. 65:281-290. Ingles, O. G. and D. R. Willoughby. 1967. An occurrence of hisinger- ite with evidence of its genesis. Soil Sci. 104:383-385. Jackson, M. L. 1956. Soil chemical analysis - advanced course. Dept. of Soils, Univ. of Wisconsin, Madison. 991 p. Jenny, H. 1941. Factors of Soil Formation. McGraw-Hill Book Co., Inc., New York. 281 p. Jenny, H. and G. D. Smith. 1935. Colloid chemical aSpects of clay pan formation in soil profiles. Soil Sci. 39:377-389. Jordens, E. R. 1972. Chemical and micro-morphological study of the genesis of a Typic Plikhaquult of Sierra Leone, West Africa. Thesis for the Ir (M.Sc.) degree, Agric. Univ. of Wageningen, The Netherlands. 74 p. 244 Juo, A. S. R., F. R. Moorman and H. 0. Maduakor. 1974. Forms and pedogenetic distribution of extractable iron and aluminum in selected soils of Nigeria. Geoderma 11:167-179. Karmanova, L. A. 1975. Effect of parent materials and of group differences between soils on the composition and distribution of various forms of iron. Soviet Soil Sci. 75-82. Kellerman, V. V. and I. G. Isyurupa. 1966. Role of iron and organic matter in the aggregation of a soil's clay component. Soviet Soil Sci. 940-947. Kellogg, C. E. 1949. Preliminary suggestions for the classification and nomenclature of great soil groups in tropical and equa- torial regions. Commonwealth Bull., Soil Sci. Tech. Communi- cation 46:76-68. Kilmer, V. J. and L. T. Alexander. 1949. Methods of making mechani- cal analysis of soils. Soil Sci. 68:15-24. Kittrick, J. A. and M. L. Jackson. 1956. Electron microscope obser- vation of reactions of phosphate with minerals leading to a unified theory of phosphate fixation in soils. J. Soil Sci. 7:81-89. Kubiena, W. L. 1938. Micropedology. Collegiate Press, Inc., Ames, Iowa. 224 p. Kubiena, W. L. 1956. Red earth formation and lateritization: their differentiation by micromorphological characteristics. 5th Intn Soil Sci., Paris, Com. V, 247-249. Kurtz, L. T. 1942. Elimination of fluoride interference in the Molybdenum blue reaction. Ind. Eng. Chem. Anal. Ed. 14:855- 856. Lundblad, KL 1934. Studies on Podzol and Brown Forest soils. 1. Soil Sci. 34:137-153. Maignien, R. 1959. Soil cuirasses in tropical West Africa. Soils Marbut, C. F. 1935. Soils of the United States. In USDA Atlas of American Agriculture, Part 3, Advance Sheet No. 8. Marbut, C. F. 1935. Soils: their genesis and classification. Pub. 1951 by Soil Sci. Soc. Amer., Madison, Wisconsin. Marshall, C. E. and C. D. Jeffries. 1946. Mineralogical methods in soil research. I. The correlation of soil types and parent materials with supplementary information on weathering processes. Soil Sci. Soc. Amer. Proc. 10:397-405. 245 Martin, F. J. and H. C. Doyne. 1927. Laterite and Lateritic soils in Sierra Leone. J. Agri. Sci. 17:530-547. Martin, F. J. and H. C. Doyne. 1930. Laterite and Lateritic soils in Sierra Leone II. J. Agric. Sci. 20:135-143. McKeague, J. A. 1967. An evaluation of 0.1 M pyrophosphate and pyrophosphate-dithionite in comparison with oxalate as extractants of accumulation products in Podzols and some other soils. Can. J. Soil Sci. 47:95-99. McKeague, J. A., J. E. Brydon and N. M. Miles. 1971. Differentia- tion of forms of extractable iron and aluminum in soils. Soil Sci. Soc. Amer. Proc. 35:33-38. McKeague, J. A. and J. H. Day. 1966. Dithionite and oxalate- extractable Fe and A1 as aids in differentiating various classes of soils. Can. J. Soil Sci. 46:13-22. McKeague, J. A. and R. J. St. Arnaud. 1969. Pedotranslocation— Eluviation-Illuviation in soils during the Quaternary. Soil Sci. 107:428—434. McKenzie, E., L. T. Kurtz and S. W. Melsted. 1977. Phosphorus fertility of some tropical soils in Sierra Leone. Soil Sci. Soc. Amer. Proc. 41:378-381. Mehra, O. P. and M. L. Jackson. 1960. Iron oxide removal from soils and clays by a dithionite-citrate cystem buffered with sodium bicarbonate, pp 317-327. In Clays and Clay Min. Proc. 7th Natl. Conf., Monograph 5, Earth Science Series, Pergamon Press, New York. Miedema, R. 1971. Comparison between the chemical and physical properties of termite mounds and surrounding soils, with special references ma Macrotermesspp. mounds. Thesis for the Ir. (M.Sc.) degree, Agric. Univ. of Wageningen, The Netherlands. 115 p. Mohr, E. C. J. and F. A. van Baren. 1954. Tropical Soils. The Royal Tropical Inst., Amsterdam Interscience, New York. Moormann, F. R., et a1. 1977. Report of the Committee on the Clas- sification of Alfisols and Ultisols with Low Activity Clays. National Soil Survey Conference, Orlando, Florida. Mortland, M. M., K. Lawton and G. Uehara. 1956. Alteration of bio- tite to vermiculite by plant growth. Soil Sci. 82:477-481. Odell, R. T. and J. C. Dijkerman. 1967. Properties, classification, and use of tropical soils, with special reference to those in Sierra Leone. Njala University College, University of Sierra Leone. 137 p. 246 Odell, R. T., J. C. Dijkerman, W. van Vuure, S. W. Melsted, A. H. Beaver, P. M. Sutton, L. T. Kurtz and R. Miedema. 1974. Characteristics, classification and adaptation of soils in selected areas in Sierra Leone, West Africa. A report of cooperative research by the Agronomy Dept. of Njala Univ. College, Univ. of Sierra Leone, and the Univ. of Illinois. Bull. 4 and 748, respectively, 194 p. Maps and appendix. Olsen, 8. R. and F. S. Watanabe. 1957. A method to determine a phosphorus adsorption maximum of soils as measured by the Langmuir isotherm. Soil Sci. Soc. Amer. Proc. 21:144-149. Ozanne, P. G. and T. C. Shaw. 1967. Phosphate sorption by soils as a measure of the phosphate requirements for pasture growth. Aust. J. Agric. Res. 18:601-612. Parfitt, R. L. 1972. Amorphous material in some Papua, New Guinea soils. Soil Sci. Soc. Amer. Proc. 36:683-686. Parfitt, R. L. 1977. Phosphate adsorption on an Oxisol. Soil Sci. Soc. Amer. Proc. 41:1064—1067. Pendleton, R. L. and S. Sharasuvana. 1946. Analysis of some Siamese laterites. Soil Sci. 62:423-440. Pollet, J. D. 1951. The geology and mineral resources of Sierra Leone. Colonial Geol. and Mineral Resources 2:3-28. Prescott, J. A. 1931. The soils of Australia in relation to vege- tation and climate. Council for Sci. and Res., Australia, Bull. 52. Prescott, J. A. and R. L. Pendleton. 1952. Laterite and Lateritic soils. Commonwealth Soils Tech. Comm. 47. Pullman, R. A. 1967. A morphological classification of Laterite ironstone and Ferruginized rocks in Northern Nigeria. Nigerian J. Sci. 1:161-174. Rhodes, E. R. 1975. Phosphate sorption isotherms of some Sierra Leone soils. J. Sci. Food Agric. 26:895-902. Rhodes, E. R. and P. M. Sutton. 1978. Active iron ratio of some soils from three physiographic units in Sierra Leone. Soil Sci. 125:326-328. Rich, C. I. and S. S. Obenshain. 1956. Distribution of free iron oxide in coastal plain soils of Virginia. 6th Congr. Int. Soil Sci. Soc. 2:447-451. Ruhe, R. V. and J. G. Cady. 1954. Latosolic soils of Central Africa interior high plateaus. Trans. 5th Int. Congr. Soil Sci. 4:401-407. 247 Saunders, W. M. H. 1965. Phosphate retention by New Zealand soils and its relationship to free sesquioxide, organic matter and other soil properties. New Zea. J. Agric. Res. 8:30-57. Schmidt-Lorenze, R. 1964. Contribution to micromorphology of the accumulation of iron and A1 oxides in the decomposition of clay in the Laterite of Keralu and Ceylon, pp. 279-289. In Jongerius (ed.), Proc. of the 2nd working Meeting of Soil Micromorphology: Elsevier, Amsterdam. Sherman, G. D. and L. T. Alexander. 1959. Characteristics and genesis of Low Humic Latosols. Soil Sci. Soc. Amer. Proc. 23:168-170. Sherman, G. D., Y. Kamahiro and Y. Matsusaka. 1953. The role of dehydration in the development of Laterite. Pac. Sci. 7: 438-446. Sherman, G. D., Y. Matsusaka, H. Ikawa, and G. Uehara. 1964. The role of amorphous fractions in the properties of tropical soils. Agrochima 7:146-163. Sivarajasingham, S. 1968. Soil and land—use survey in the Eastern Province of Sierra Leone. F.A.O. Report TA 2584, Rome. 73 p. and map. Sivarajasingham, S., L. T. Alexander, J. G. Cady, and M. G. Cline. 1962. Laterite. Advance in Agronomy vol. 14:1-60. Smith, H. and L. P. Wilding. 1972. Genesis of argillic horizons in Ochraqualfs derived from fine textured till deposits in Northwestern Ohio and Southwestern Michigan. Soil Sci. Soc. Amer. Proc. 36:808-815. Soil Survey Staff. 1960. Soil Classification, A Comprehensive System. 7th Approximation. U.S. Dept. Agr., Washington, D.C. Soil Survey Staff. 1972. Soil Survey Investigation Report 1. U.S. Dept. Agr., Washington, D.C. Soil Survey Staff. 1975. Soil Taxonomy. U.S. Dept. Agr., Washington, D.C. Sree Ramulu, U. S., P. F. Pratt and A. L. Page. 1967. Phosphorus fixation by soils in relation to extractable iron oxides and mineralogical composition. Soil Sci. Soc. Amer. Proc. 31: 193-196. Steele, J. G. and R. Bradfield. 1934. The significance of size distribution in the clay fraction. Amer. Soil Survey Assoc. Bull. 15:88-93. Stephen, I. 1952. A study of rock weathering with reference to the soils of the Malvern Hills. 1. Weathering of biotite and granite. J. Soil Sci. 3:20-33. 248 Stephen, I. 1960. Clay orientation in soils. Sci. in Progress 48:332-333. Stobbs, A. R. 1963. The Soils and Geography of the Boliland Region of Sierra Leone. Govt. of Sierra Leone, Freetown. 45 p. and maps. Stoops, G. 1968. Micromorphology of some characteristic soils of the Lower Congo (Kinshasa). Pedologie 18 (1):110-149. Sys, C. 1968. Suggestions for the classification of tropical soils with lateritic materials in the American classification. Pedologie 18:189—198. Tamura, T. 1958. Identification of clay minerals from acid soils. J. Soil Sci. 9:141—147. Tanner, C. B. and M. L. Jackson. 1947. Nomographs of sedimentation time for soil particles under gravity or centrifugal accelera- tion. Soil Sci. Soc. Amer. Proc. 12:60-65. Thorp, J. and M. Baldwin. 1938. New nomenclature of the higher categories of soil classification as used in the Dept. of Agr. Proc. Soil Sci. Soc. Amer. 3:260-268. Thorp, J. and G. D. Smith. 1949. Higher categories of soil classi- fication. Order, suborder and great soil group. Soil Sci. 67:117-126. Truog, E., J. R. Taylor, R. W. Pearson, M. E. Weeks and R. W. Simonson. 1937. Procedure for special type of mechanical and mineralogi- cal soil analysis. Soil Sci. Soc. Amer. Proc. 1:101-112. Udo, E. J. and F. D. Uzu. 1972. Characteristic of phosphorus adsorp- tion by some Nigerian soils. Soil Sci. Soc. Amer. Proc. 36: 879-883. van Vuure, W. and R. Miediena. 1973. Soil survey of the Makeni area. Northern Province. Sierra Leone. Njala Univ. College, Univ. of Sierra Leone. 104 p., maps. van Vuure, W., R. T. Odell and P. M. Sutton. 1973. Soil survey of the Njala area, Sierra Leone. Njala Univ. College, Univ. of Sierra Leone, Bull. 3. 117 p. and maps. Waldock, E. A., E. S. Capstick and A. J. Browning. 1951. Soil con- servation and use in Sierra Leone. Sessional paper No. 1. 124 p. van Wambeke, A. R. 1962. Criteria for classifying tropical soils by age. J. Soil Sci. 13:124-132. van Wambeke, A. R. 1967. Recent developments in the classification of the soils of the tropics. Soil Sci. 104:309-312. 249 Westerveld, D. H. 1969. Morphological and mineralogical differences between two types of iron concretions in a soil of Sierra Leone, West Africa. Thesis for the Ir (M.Sc.) degree, Agric. Univ. of Wageningen, The Netherlands. 25 p. Williams, F. G., N. M. Scott and M. J. McDonal. 1958. Soil proper- ties and phosphate sorption. J. Sci. Food Agric. 9:551-559. Woodruff, J. R. and E. J. Kamprath. 1965. Phosphorus adsorption maximum as measured by the Langmuir isotherm and its rela- tionship to phosphorus availability. Soil Sci. Soc. Amer. Proc. 29:148-150. Young, A. 1976. Tropical Soils and Soil Survey. Cambridge Univ. Press. 468 p. APPENDICES APPENDIX A PROFILE DESCRIPTIONS (BAOMA, MANOWA, NJALA, MAKENI, SEGBWEMA, TIMBO, PENDEMBU, MASUBA, MOA, GBESEBU, MAKUNDU) Profile 144801A, Baoma sandy clay loam Description after Sivarajasingham (1968) Location Physiography Relief Vegetation Drainage Parent material A1 0-5 inches 0-13 cm Lab. No. 828572 B2t 5-23 inches 13-58 cm Lab. No. 528571 On the right-hand side of the road from Daru Junction to Moa Barracks, about 150 feet (46m) past the Girl's School at the first bend of the road to the left. Undulating upland. Upper gentle slope. Cocoa 7 to 15 years old in very poor health, poor management, inadequate shade, Open stand, and heavy weed growth; few tall trees of the former secondary forest remain. Well drained. A thick layer of locally transported material derived from laterite crust and partially weathered and fresh rock of a previous landscape. Dusky red (2.5YR 3/2); sandy clay loam; moderate fine subangular blocky; light and porous: soft, slightly sticky, slightly plastic; many fine and medium roots; clear, smooth boundary to horizon below. Red (10R 4/6) to dark red (10R 3/6); clay; less than 5% hardened plinthite glaebules of the kind in the layer below: moderate medium and fine sub- angular blocky; porous; friable, slightly sticky, and slightly plastic; common fine and medium roots; abrupt, smooth boundary to horizon below. IIB31 23-43 inches 58-109 cm Lab. No. 528570 11332 43-67 inches 109-170 cm Lab. No. $28569 251 Main gravel layer; red (10R 4/6); gravelly clay; 40% black-coated and uncoated, medium, round, black, dense hardened plinthite glaebules, and 10% coarse, dense, very hard, fresh rock pebbles; moderate fine subangular blocky; porous; friable, slightly sticky, slightly plastic; few fine roots; diffuse, smooth boundary to horizon below. Red (10R 4/6); gravelly clay; 40% hardened plinthite glaebules as in IIB31 but finer in size; moderate fine subangular blocky; slightly porous; friable, slightly sticky, slightly plastic; no roots. Profile Kpuabu l, Manowa sandy clay loam Description after Location Physiography Relief Vegetation Drainage Parent material Al 0-10 inches 0—25 cm Lab. NO. 828558 A3 10-21 inches 25-53 cm Lab. NO. 828557 Sivarajasingham (1968) Kpuabu Cocoa Experiment Station; about 450 feet (137m) from the Kenema-Joru road on the road to the Station Office, and about 150 feet (46m) on the right-hand side from the Station Office road. Accordant, flat-topped hill of the dissected lateritic upland. Upper, convex 5% slope. Cocoa plantation under many tall trees of original secondary vegetation; good grass cover. Moderately well drained. A thin layer of gravel-free material over a thick, very gravelly layer of locally transported material. Very dark grayish brown (lOYR 3/2); sandy clay loam; moderate medium and fine subangular blocky; porous; friable, slightly sticky, slightly plastic; common fine, few medium, and very few coarse roots; clear, smooth boundary to horizon below. Dark brown (lOYR 3/3); very gravelly sandy clay; 70% yellow-coated, nodular, coarse and medium, dense, red and yellow, hardened plinthite glaebules; weak, fine subangular blocky aggregates with no strong interface; friable, sticky, slightly plastic; few fine and very few medium roots; gradual, smooth boundary to horizon below. E21 21-35 inches 53-89 cm E22 35-70 inches 89-178 cm Lab. No. 528556 Remarks 252 Dark yellowish brown to yellowish brown (10YR 4/4—5/6); very gravelly sandy clay; 60% yellow- coated and uncoated, round, fine, dense, red and black, hardened plinthite glaebules; weak fine subangular blocky aggregates with no strong inter- face; friable, sticky, slightly plastic; very few fine and medium roots; gradual, smooth boundary to horizon below. Strong brown (7.5YR 5/8); very gravelly clay; 75% yellow-coated nodular, coarse, dense, red and yellow, hardened plinthite glaebules; weak fine subangular blocky aggregates with no strong inter- face; porous; friable, sticky, slightly plastic; very few fine and medium roots. This soil is very gravelly and droughty and would be expected to be unsuitable for cocoa. The cocoa planted in 1959 appears as stunted trees of very poor health, with many vacant patches because of low survival rate of the planted seedlings. Management is very good, but shade appears to be excessive. Profile N109, Njala very gravelly clay loam Described by H. Breteler on December 29, 1966 Location Physiography Relief Vegetation Drainage Parent material A1 0-14 inches 0-35 cm Lab. No. 529072 Southwestern corner of the Oil Palm Station of Njala University College. From the junction of the Kania boundary road and the village path to Pujehun, 84 feet (26m) along the boundary line uphill towards the Taia River. Pit N109 is located between palms 1155 and 993. Dissected erosion surface; on the slope between the highest erosion surface and the upper river terrace. Lower part of convex 8% slope, on mapping unit 3, Njala, sloping. Oil palm plantation. Well drained. Colluvial material high in hardened plinthite glaebules. Dark brown (10YR 3/3); very gravelly (70% by volume) clay loam; gravels are mainly round and nodular, %" to k" in diameter, hardened plinthite glaebules with outside colors of yellowish red A3 14-21 inches 35—53 cm Lab. No. 829074 E21 21-49 inches 53-125 cm Lab. No. 829076 B22 49-62 inches 125-157 cm Lab. No. 529076 253 (5YR 4/6) and inside colors of yellowish red and very dusky red (5YR 4/6, 4/8, 5/8 and 10R 2/2); weak to moderate very fine and fine subangular blocky, breaking into very weak, very fine granular very friable; many fine, medium, and coarse pores; many fine, medium, and coarse roots; gradual, smooth boundary to horizon below. This A1 horizon is thick enough and just dark enough to qualify as an umbric epipedon (see Section 4). Most pro- files of Njala soils have thinner or lighter colored A1 horizons, which are ochric rather than umbric. Dark yellowish brown (10YR 4/4); very gravelly (80%) clay; gravels are mainly round and nodular, %" to 3" in diameter, hardened plinthite glaebules with outside colors of yellowish red and very dusky red (5YR 5/8 and 10R 2/2) and inside colors of dark red, red, and reddish yellow (10R 3/6, 4/8 and 7.5YR 6/8); weak very fine, fine, and medium subangular blocky, breaking into weak to moderate very fine, fine and medium granular; very friable; many fine, medium, and coarse pores; many fine, medium, and coarse roots; gradual, smooth boundary to horizon below. Strong brown (7.5YR 5/8); very gravelly (70%) clay; gravels are mainly round and nodular, %" to 3" in diameter, hardened plinthite glaebules with out- side colors of red (2.5YR 4/8—5/8) and inside colors of red and very dusky red (10R 4/6 and 2/2); very weak fine, medium, and coarse angular to subangular blocky, breaking into weak to moderate very fine and fine granular; friable; many fine, medium, and coarse pores; common fine, medium, and coarse roots; gradual, smooth boundary to horizon below. Yellowish red (5YR 5/8); very gravelly (50%) clay; gravels are mainly round and nodular, %" to 1" in diameter, hardened plinthite glaebules with out- side colors of yellowish red (5YR 5/8) and inside colors of red and very dusky red (10R 4/6 and 2/2); very weak fine, medium, and coarse angular to sub- angular blocky, breaking into weak to moderate very fine and fine granular; friable; common fine, medium, and coarse pores; common fine, medium, and coarse roots. 254 Profile P2, Makeni very gravelly sandy clay loam Described by W. van Vuure and R. Miedema on March 8, 1968 Location Physiography Relief Vegetation Drainage Parent material A1 0—10 inches 0-25 cm Lab. No. 529804 B21 10—20 inches 25-50 cm Lab. No. 529805 B22 20-67 inches 50-170 cm Lab. No. 829806 Topographic map of Sierra Leone, scale 1:50,000, sheet 43, coordinates HE 274-865; on traverse A, near augerhole 4. Dissected erosion surface, sloping. Slopes 14% to SSW and 10% to SE. Secondary bush, 4 to 10 years old. Well drained. Gravelly to very gravelly weathering products of Precambrian granite and acid gneiss. Dark brown (10YR 3/3); very gravelly sandy clay loam; structure and consistence not observable because of gravel content; common macro- and many meSOpores; few fine distinct charcoal mottles; common coarse, many medium and fine roots; common ant and termite activity; 74% uncoated, fine and medium, rounded, dense, red and yellow, hardened plinthite glaebules; clear, smooth boundary to horizon below. Yellowish red (5YR 4/8); very gravelly clay; very weak, very fine angular blocky; consistence is not observable because of high gravel content; common macro— and mesopores; common coarse many medium and fine roots; low biological activity; 77% uncoated fine and medium, rounded and nodular, dense, red and yellow, hardened plinthite glae- bules; gradual, smooth boundary to horizon below. Yellowish red (5YR 5/8); very gravelly clay; very weak, very fine angular blocky; firm, slightly sticky and plastic; common macro— and mesopores; few medium, common fine roots; 80% yellow-coated, medium and coarse, nodular and angular, dense, red, hardened plinthite glaebules, and a few very fine quartz gravels. Profile 145005, Segbwema gravelly sandy clay loam Description after Sivarajasingham (1968) Location On a very high hill on the right-hand side of the road from Mano Junction to Segbwema Junction. The path leading to the pit starts from the village of Niahun and goes southwards. Physiography Relief Vegetation Drainage Parent material A1 0-13 inches 0—33 cm Lab. No. 828564 B21 13-28 inches 33-71 cm E22 28—60 inches 71-153 cm Lab. No. 828563 C1 60-94 inches 153-239 cm Lab. No. 828562 255 Very high hills. The pit is on the middle part of a very steep (42%), straight slope of a very high hill. The land was in upland rice during 1965; in 1966 it was under a low succulent to woody herbaceous vegetation with many wild oil palms. Well drained. Residual, presumably from rock of granodioritic composition. Strong brown (7.5YR 5/6); gravelly sandy clay loam; strong fine subangular blocky and granular; medium density and porosity; friable, slightly sticky, slightly plastic; common fine and medium roots; clear, smooth boundary to horizon below. Red to weak red (10R 4/6-4/4); heavy sandy clay loam, slightly gritty; strong medium subangular blocky; porous; friable, slightly sticky, slightly plastic; few fine roots; gradual, smooth boundary to horizon below. Red (2.5YR 4/6) with few coarse red (10R-7.5R 4/6) mottles; clay loam with fine white specks of decomposing feldspar indicating its saprolitic nature; strong medium subangular blocky; porous friable, slightly sticky, slightly plastic; few fine roots; diffuse, smooth boundary to horizon below. Red (2.5YR 4/8 and 10R 4/8) in equal amounts present as coarse faint mottles; also contains white decomposing feldspar and black decomposing hornblende; sandy clay loam; weak fine subangular blocky; porous; nonsticky, slightly plastic; few fine roots. Profile P19, Timbo gravelly sandy clay loam Described by J. M. Cawray, A. A. Thomas and R. Miedema on March 27, 1968 Location Physiography Topographic map of Sierra Leone, scale 1:50,000, sheet 43, coordinates HE261-852; near Timbo along the motor road from Makeni to Panlap. Dissected erosion surface. Relief Vegetation Drainage Parent material All 0-12 inches 0-30 cm Lab. No. 529818 A12 12-19 inches 30-49 cm Lab. No. 529819 AB 19-28 inches 49-70 cm Lab. No. 529820 1321 28-42 inches 70—100 cm Lab. No. 529821 B22 43-70 inches 110-179 cm Lab. No. 829822 256 Slope 6% to south. Cassava and short weeds and grasses. Well drained. Gravelly weathering products of Precambrian granite and acid gneiss. Very dark grayish brown (10YR 3/2); gravelly sandy clay loam; weak fine subangular blocky; slightly hard, friable, slightly stick, and slightly plastic; many macro- and mesospores; few fine distinct charcoal particles; common coarse and medium roots; 44% fine and medium, uncoated nodular, red, hardened plinthite glae- bules, and a few decomposed rock fragments; clear, smooth boundary to horizon below. Dark brown (10YR 3/3); very gravelly sandy clay loam; weak fine subangular blocky; slightly hard, friable, slightly plastic; many macro- and meso- pores; few fine distinct charcoal particles; common coarse, medium, and fine roots; 50% medium and fine, uncoated nodular, very hard, porous, yellow and red, decomposed rock fragments; clear, wavy boundary to horizon below. Yellowish red (5YR 4/6); gravelly sandy clay loam; weak fine subangular blocky; friable, slightly sticky, and slightly plastic; many macro- and mesopores; few coarse, medium, and fine roots; 47% uncoated, very hard, porous, red and yellow, decomposed rock fragments; few feldspars and micas, especially in the decomposing rock pieces; gradual, wavy boundary to horizon below. Yellowish red (5YR 4/8); gravelly sandy clay loam; weak medium angular and subangular blocky; friable, slightly stick, and slightly plastic; many macro- and mesopores; few coarse, medium, and fine roots; 44% coarse, medium, and fine, uncoated, soft to hard, porous, red and yellow, decomposed rock fragments; few micas and feldspars; diffuse, smooth boundary to horizon below. Yellowish red (5YR 5/8); gravelly sandy clay loam; weak medium angular blocky; sticky and plastic; common macro- and mesopores; few coarse, medium, and fine roots; 20% coarse and medium, uncoated, soft to hard, porous, red and yellow, decomposed rock fragments, with feldspars and micas. 257 Profile Kpuabu 2, Pendembu fine sandy loam Description after Sivarajasingham (1968) Location Physiography Relief Vegetation Drainage Parent material A1 0-7 inches 0-18 cm Lab. No. 828552 A3 7-18 inches 18-46 cm Lab. No. 828551 Blt 18—37 inches 46-94 cm 132:: 37-54 inches 94-137 cm Lab. No. 828550 Kpuabu Cocoa Experiment Station, about halfway between the nursery buildings and the stream. Accordant, flat-topped hill of the dissected lateritic upland. Long, gentle, concave slope of about 2%. Cocoa plantation under adequate shade of many tall trees of original secondary forest. Although the soil is deep and gravel-free and would have been expected to be suitable, the cocoa planted in 1959 shows many large, vacant patches. Imperfectly drained. A thick layer of gravel-free, locally transported, leached parent material. Very dark gray (10YR 3/1); fine sandy loam; weak, medium, and fine, subangular blocky; friable, slightly sticky, nonplastic; very few fine pores; common fine and medium roots; clear, smooth boundary to horizon below. Dark grayish brown (2.5Y 4/2-10YR 4/2); sandy clay loam; dense clods breaking to weak fine subangular blocky aggregates with no characteris- tic interface; friable, slightly sticky, slightly plastic; very few pores and few large burrow holes; common fine and medium and few coarse roots in the first 5 inches, then decreasing gradually with depth; clear, smooth boundary to horizon below. Yellowish brown (10YR 5/4) to light olive brown (2.5Y 5/4); sandy clay loam; dense clods as in A3 horizon; friable to firm, slightly sticky, slightly plastic; few pores with clay coatings along the pore walls; few fine and medium roots; gradual, smooth boundary to horizon below. Yellow (2.5Y 7/6) to brownish yellow (10YR 6/6); sandy clay loam; massive clods breaking to weak very fine subangular blocky aggregates with no characteristic interface; friable to firm, slightly p1astic,slightly sticky; few pores with clay coatings along the pore walls; very few fine and medium roots; gradual, smooth boundary to horizon below. B3t 54-72 inches 137-183 cm Clg 72—80 inches 183-203 cm 258 Pale yellow to yellow (2.5Y 7/6) with common medium, faint yellowish brown (10YR 5/6) and few medium, prominent red (2.5YR 4/8) mottles; sandy clay loam; massive clods as in B2 horizon; firm, slightly sticky, slightly plastic; many fine pores with clay coatings; very few very fine roots; gradual, smooth boundary to horizon below. White (2.5Y 8/2) with common medium, prominent yellowish brown (10YR 5/6) and strong brown (7.5YR 5/8) mottles; sandy clay; massive; wet; firm, sticky, slightly plastic; the strong brown mottles are firm to hard and may be considered as incipient plinthite glaebules; no roots. Profile P9, Masuba sandy clay loam Described by R. Miedema and A. A. Thomas on March 20, 1968 Location Physiography Relief Vegetation Drainage Parent material AP 0-7 inches 0-19 cm Lab. NO. 529810 Bl 7-22 inches 19-57 cm Lab. NO. 829811 Topographic map of Sierra Leone, scale l:50,000, sheet 43, coordinates HE278-872; on traverse E, 525 feet (160m) from profile P8. Lower part of stream terrace near valley edge. Slope 0 to 3%. Farm with cassava, Kandi trees and weeds, and many wild oil palms. Moderately well drained. Gravel-free, transported alluvial/colluvial material. ‘ Very dark grayish brown (10YR 3/2); sandy clay loam; weak fine to medium angular blocky; very hard; common macro- and many mesopores; few distinct fine charcoal mottles; many coarse, medium, and fine roots; many large and medium ant holes; clear, smooth boundary to horizon below. Pale brown (10YR 6/3); sandy clay loam; weak fine to medium angular blocky; very hard; many macro- and mesopores; few distinct fine charcoal mottles; common fine distinct reddish yellow to strong brown (7.5YR 5.5/8) to red (2.5YR 5/8) mottles; common coarse, many medium and fine roots; less than 10% uncoated, nodular, coarse, porous, red, hardened plinthite glaebules, with few quartz grains; many large and medium ant holes; gradual, smooth boundary to horizon below. B2 22-67 inches 57-170 cm Lab. No. 529812 259 Very pale brown (10YR 6.5/3); sandy clay loam;. weak fine angular blocky; firm; many macro- and meSOpores; many distinct fine and medium yellowish red (5YR 5/8) and reddish yellow (7.5YR 6/8) mottles; common distinct fine and medium char- coal mottles; few coarse, common medium, and many fine roots; less than 10% gravel, similar to B1 horizon, and one quartz stone; common worm holes with dark coatings. Profile Kpuabu 3, Moa clay Description after Location Physiography Relief Vegetation Drainage Parent material A1 0-6 inches 0-15 cm Lab. No. 828555 B21 6-21 inches 15-53 cm Lab. No. 828554 Sivarajasingham (1968) Kpuabu Cocoa Experiment Station; near the path from the nursery buildings to the wooden bridge over the stream. Bottomland (river terrace). Middle of a narrow, level terrace adjoining a stream whose bed has incised about 10 feet (3m) below the terrace surface. A thick stand of cocoa planted in 1960, with dense foliage forming a close canopy adequate shade of many tall trees of the original secondary forest. ‘ Moderately good. The land may be flooded two or three times a year for durations of one or two weeks. Flood water drains rapidly from the sur- face layers, but even during the height of the dry season the water table is encountered within 6 or 7 feet (2m) below the surface. A thick layer of clayey river alluvium. Very dark grayish brown (10YR 3/2); clay; strong fine subangular blocky and fine granular; porous; friable, slightly sticky, slightly plastic; ter- mites and earthworms present; many fine and medium roots; clear, smooth boundary to horizon below. Strong brown to dark brown and brown (7.5YR 5/6- 4/4); clay; strong fine subangular blocky; porous; friable, sticky, slightly plastic; common fine and medium roots; gradual, smooth boundary to horizon below. B22 21-31 inches 53-79 cm B3 31—59 inches 79—150 cm Lab. No. 528553 Clg 59-71 inches 150-180 cm 260 Strong brown (7.5YR 5/6-5/8) with few fine, dis- tinct yellowish red (5YR 4/8) to red (2.5YR 4/8) mottles; clay; strong medium and fine subangular blocky; porous; friable, sticky, slightly plastic; common fine and medium roots; gradual, smooth boundary to horizon below. Brownish yellow (10YR 6/8) with few, medium dis- tinct strong brown (7.5YR 5/8) and red (2.5YR 4/8) mottles; yellow (2.5Y 7/6) mottles are more numerous with increasing depth; clay; strong medium subangular blocky; porous; friable, sticky, slightly plastic; few fine and medium roots; gradual, smooth boundary to horizon below. Mottled white (N 8/ ), yellow (2.5Y 7/6), yellowish brown (10YR 5/8), and strong brown (7.5YR 5/6) in a variegated pattern; sandy clay; wet, massive clods; the strong brown mottles are firm to hard and may be considered as incipient plinthite glaebules. Profile N125, Gbesebu silty clay Described by H. Breteler on January 18, 1967 Location Physiography Relief Vegetation Drainage Parent material A1 0—4 inches 0-10 cm Lab. No. 829066 From the extreme southwestern corner of the Oil Palm Station of Njala University College, at the junction of the Kania boundary road and the path along the Taia River near surveyor stone No. PB-B 829, thence 322 feet (98m) down the steep lepe towards the river to pit N125, near the river bank on a natural levee. Natural levee of the Taia River, on the present floodplain or first terrace. Nearly level, convex slope. Old secondary bush with much grass. Moderately well drained; may be flooded for several weeks during the wet season. Clayey alluvium. Dark brown (10YR 4/3); silty clay; weak very fine and fine subangular blocky, breaking to weak very fine granular; very friable; many fine, medium, and coarse pores; many fine, medium, and coarse roots; clear, smooth boundary to horizon below. A3 4—7 inches 10-18 cm Lab. NO. 829067 B21 7-19 inches 18-48 cm Lab. No. 829068 BZZb 19-25 inches 48-63 cm Lab. No. 829069 B23 25-63 inches 63-160 cm Lab. No. 529070 261 Dark yellowish brown (10YR 4/4); clay; weak very fine and fine angular to subangular blocky, breaking to weak very fine granular; friable; many fine, medium and coarse pores; many fine, medium, and coarse roots; clear, smooth boundary to horizon below. Strong brown (7.5YR 5/6) with many fine and medium faint yellowish red (5YR 5/6) mottles; clay; weak very fine and fine blocky, breaking into weak very fine granular; friable; many fine, medium, and coarse pores; common fine, medium, and coarse roots; mica flakes; clear, smooth boundary to horizon below. Strong brown to yellowish brown (7.5YR-10YR 5/6) with many fine, medium, and coarse faint yellowish red (5YR 5/6) mottles; this is a buried A horizon with common fine, medium, and coarse charcoal mottles; clay weak to moderate fine and medium blocky, breaking into weak to moderate very fine and fine granular; firm; many fine, medium, and coarse pores; common fine and medium roots; mica flakes; clear, smooth boundary to horizon below. Strong brown (7.5YR 5/6) with many fine, medium, and coarse faint yellowish red (5YR 5/6) mottles; clay; weak to moderate fine and medium blocky, breaking into weak to moderate fine granular; firm; many fine, medium, and coarse pores; few fine and medium roots; mica flakes. Profile P104, Makundu clay Described by J. M. Cawray and R. Miedema on June 18, 1968 Location Physiography Relief Vegetation Drainage Parent material Topographic map of Sierra Leone, scale 1:50,000, sheet 43,coordinates HE124-912; 300 feet (91m) from Mabole River, on the road from Makundu to the river. Nearly level river terrace. Concave, very gentle 1% slope to the south. Dense secondary bush, with many wild oil palms. Moderately well to well drained. Alluvium from the Mabole River. All 0-8 inches 0-20 cm Lab. No. 529841 A12 8-16 inches 20-41 cm Lab. No. 829842 AB 16-21 inches 41-53 cm Lab. No. 829843 Bl 21-28 inches 53-71 cm Lab. No. 829844 I321 28-43 inches 71-108 cm Lab. No. 529845 E22 43-74 inches 108-188 cm Lab. No. 829846 262 Very dark gray (10YR 3/1); clay; weak fine angular and subangular blocky; friable, slightly sticky, and slightly plastic; many macro- and mesopores; few medium distinct charcoal mottles; many coarse, medium, and fine roots; much ant, termite, and worm activity; clear, smooth boundary to horizon below. Very dark grayish brown to dark brown (10YR 3/2.5); clay; weak fine angular and subangular blocky; friable, slightly sticky, and slightly plastic; many macro- and mesopores; few medium distinct charcoal mottles; many coarse, medium, and fine roots; much ant, termite, and worm activity; clear, smooth boundary to horizon below. Dark yellowish brown (10YR 4/4); silty clay; weak fine angular and subangular blocky; friable, sticky and plastic; many macro- and mesopores; few medium distinct charcoal mottles; many coarse, medium, and fine roots; much ant, termite and worm activity; clear, smooth boundary to horizon below. Yellowish brown (10YR 5/6); clay; moderate fine and medium angular and subangular blocky, firm, sticky, and plastic; many macro- and mesopores; few medium distinct charcoal mottles; few fine faint yellowish red (5YR 4/6) mottles; common coarse, many medium and fine roots; much ant, termite, and worm activity; gradual, smooth boun- dary to horizon below. Brownish yellow (10YR 6/6); clay; moderate fine and medium angular and subangular blocky; firm, sticky, and plastic; common macro- and mesopores; common medium distinct yellowish-red (5YR 4/8) mottles; few medium, common fine roots; much ant, termite and worm activity; gradual wavy boundary to horizon below. Brownish yellow (10YR 6/6); clay; moderate angular and subangular blocky; firm, sticky and plastic; common macro- and mesopores; many medium prominent yellowish red (5YR 5/6) mottles; common fine roots; common ant, termite, and worm activity. 263 EoHme No bosom was mean Ho>mum mcoumcouH m\m mum.mnm\m mum.m mmlo «cwxmz omumoo nuumoo pom enemas pom Hw>muo mcoumcouw new menu .Hm>mum xoou nmumnummz o\v mwm.mnm\m mwm omauooa Esfiome Nuumso 30m popcsoundm ou um~smc< pom mafia .Ho>muo ocoumconH v\m mwm.m ooanoaa Ho>mum xoou poumsumm3 mo Edfipoe mwooflm 30m m moam nmncsounsm cu amasmca new wane .Hm>mum mcoumcouH «\m muoanvxm mmm.m caauoo Edwoma “Madman ou bosom pom mcflm Hw>mum mcoumcouH m\m mwmlv\v mwm ooIOm Edwpms uuumso 3mm muumso H\m mwoa 30m umHsmcm new venom new menu can; Hm>mum wcoumcouH me.m mumuo\m mam cmuma Hm>mum uuumso Hmaomcm pom xoou poumnumw3 uuumoo H\m mwoa 3mm daomvva Noumso .pcsou >chmz womb Bow .Hw>mum mcoumcouH m\m mwmlv\v mum.m mano msomm macaw «mNHm Ho>muw mo Axfluume Hmcnoucflv AEoV mwfiumm Hwom cofiuflmomsoo HO GCHM uoHoo Hammcsz nummo conwuom AomzHe ozc «302m: .Hzmxaz .azoamv mmqoooz mzoemzomH mo wooqommmozomoaz m XHozmmmd 264 mmumoo paw EDfipoE uuumso 3mm pom umHomc< .ocwm Hw>mum mcoumcouH ©\mnq\m mwm.m mmalmm mmumoo pom Eswows “nausea .wcam Hm>mum wcoumcouH v\m mum.m .m\v mwm.m mmumm EoHme “magnum new menu Hm>mum mcoumcouH o\m mom a mxv mwm.~ mmumm Esflpmfi “madman pom bosom pom wcfim Hm>mum wcoumcouH m\m moanv\m mwm.m mmuo 630cm: mmumoo pom endows “masmc< .ocam Hm>muo oCODmcouH v\m mwm.m moalmaa mmumoo on Edfloma unasmc< .mch Hm>mum oCODmcouH o\m mum.m mfialmm mmumoo pom Edwpwa Hmasmcm Ou UGCCSOHQSm .mCHm Hw>mum mcoumCOHH ©\m w «\m mwm.m mmumm endows popcsoundm paw venom pom mcwm Ho>mum ocODmcouH o\m mwm.m w m\m mwm.m mmlmm wmmzm *mNflm Hm>muo mo Axfiuums accumucfiv AEov mofluom Hwom COAUHmOQEOU Ho coax uoHoo Madman: comma CONfluom Aomscflucoov m xHozmmmc 265 .Eom.mA wmumoo .Eom.mlmm.a ESHUOE .EOmN.va:Hh .1 omumoo mucmEmmHm umHsmc< ocm Esflpmz xoou mononumwz o\m mwm.m .m\v moa mnaumaa mmumoo mucmammum “masmca 6cm esonmz x00» nmumnummz m\m mwm.m .v\m m>m.m maanok omumoo Ho>mum 6cm mucosomum umH56c< 6cm endow: xoou meaumnummz 6\v mom .m\m m>m onumv omumoo Ho>muo NDHMDU 3mm pom EdflpwE mafia Amcflumsummzv unasmc< .mcwm Hm>mnm xoom m\v moa w m\v mum.m mvuom mmumoo pom Esflcmfi Ho>wum xoou amasmca .mcmm 36o m can mcoumcouH o\m mew.» a «\v mwm.m omuo oases mmumoo 6cm EsHme smasmca .mcom Hm>mum mcoumcouH «\m mwm.m .o\v mmm.m omfiumma mmmnm «gnaw Hm>muo mo Axfluuma Hmcuwucflv AEoV mwfiumm HHom cofluflmomfiou Mo ocflx 4' 'i" .1 1 :‘lu‘ln. I . ' I ‘. '1 g... l.'lu....x'.l‘li-;.ll uoHou Haomcsz cummo cowauom Aomscflucoov m xHozmmmc APPENDIX C RATIO OF EXTRACTABLE IRON AND ALUMINUM OXIDES TO OTHER VARIABLES F6203d- Fe Si]. Soil Fe203d §§2939§_ F229§9§_ A1202_ ox clay,Fe Fesil = Total Series Md* Md* A1203d removed Fe - Fed** Baoma Al 1.33 0.39 0.90 0.58 0.27 11.34 B2 1.55 0.36 1.20 0.59 0.22 11.28 IIB31 1.70 0.36 1.30 0.44 0.21 11.60 IIB32 1.50 0.35 1.20 0.47 0.22 13.06 IIB33 --- --- --- --- --- --- Manowa Al 0.11 0.75 1.10 0.67 0.12 4.64 A3 0.09 0.82 1.35 0.84 0.13 5.34 B21 0.07 0.92 1.17 0.79 --- --- B22 0.05 1.16 1.30 0.86 0.13 6.30 Njala A1 0.31 0.08 0.24 2.96 0.05 2.63 A3 0.45 0.06 0.38 2.10 0.04 2.38 B21 0.58 0.06 0.51 1.30 0.04 2.31 B22 0.67 0.05 0.62 1.00 0.02 1.64 Makeni A1 0.21 0.05 0.16 1.75 0.12 6.58 B21 0.28 0.05 0.23 3.43 0.17 11.01 B22 0.31 0.04 0.27 2.61 0.17 11.50 Segbwema A1 0.68 0.33 0.35 0.51 0.03 1.7 Bth 0.64 0.27 0.37 0.45 -—- --- BZZt 0.58 0.23 0.35 0.40 0.14 4.56 Cl 0.30 0.12 0.16 3.30 0.13 31.10 266 Appendix C (continued) 267 Fe203d~ Fe Si]. Soil 222939 3222325 2229325. 512.92% clamFe Fesil = Total Series Md* Md* Md* A1203d removed Fe - Fed** Timbo All 0.31 0.07 0.25 1.45 0.23 7.68 A12 0.43 0.06 0.37 1.54 0.19 8.02 AB 0.42 0.05 0.37 1.36 0.24 9.95 BZlb 0.53 0.04 0.49 0.53 0.26 9.97 B22b 0.48 0.04 0.44 0.24 0.34 10.79 Pendembu A1 0.19 0.05 0.15 1.20 0.03 1.00 A3 0.24 0.05 0.19 0.89 0.04 1.16 821 --- —-- --- 0.63 --- --- B22 0.21 0.02 0.20 0.51 0.04 1.34 823 --- --- --- 0.19 --- --- B3 --- --- --- 0.35 --- --- Masuba AP 0.12 0.04 0.11 0.55 0.09 2.08 Bth 0.17 0.04 0.16 0.29 0.09 2.34 BZt 0.14 0.04 0.13 0.70 0.10 2.51 .r~4<>_a Al 0.42 0.08 0.34 0.87 0.06 3.20 821 0.45 0.07 0.39 0.44 0.07 3.85 B22 0.51 0.05 0.46 0.24 --- --- B3 0.52 0.07 0.45 0.17 0.07 3.78 C 0.51 0.06 0.45 0.10 --- --- Gbesebu Al 0.37 0.12 0.25 2.60 0.02 1.74 A3 0-45 0.19 0.26 2.00 0.02 1.23 BZlb 0.49 0.19 0.31 3.20 0.03 2.12 BZ2b 0.43 0.18 0.26 3.20 0.03 2.38 823 0.39 0.13 0.26 2.40 0.04 2.82 Makundu All 0.26 0.03 0.23 0.15 0.10 6.65 A12 0.31 0.10 0.22 2.24 0.09 6.47 AB 0.31 0.07 0.24 1.15 0.10 6.72 B1 0.32 0.08 0.25 3.21 0.10 6.87 B21 0.32 0.03 0.29 0.65 0.10 7.23 822 0.30 0.04 0.27 0.71 1.09 6.48 * Mean difference of percent clay after Fe removed and before Fe removal. ** Total iron - Fe dithionite = Silicate Fe, Fesil’