Mineral and elemental distribution in soils formed on the River Niger floodplain, eastern Nigeria.

Introduction

Studies on the mineral and elemental distribution within soils provide vital information for the assessment of their genesis and behaviour. Teveldal et al. (1990) reported that, in some soils, the depth distribution of minerals in the fine fractions does not depict the correct relative stability of minerals, since physical disintegration of coarser fractions continually feeds the finer fractions with fresh materials. Also, St. Arnaud and Whiteside (1963) earlier showed that physical weathering may affect the distribution of minerals in the soils in ways that complicate the study of chemical weathering. From studies of a chronosequence of soils derived from alluvium, Harris et al. (1980) observed that the distribution of minerals in the coarse fraction reached stability earlier than the distribution in the fine fraction. Esser et al. (1992) indicated that physical weathering and mineral distribution patterns in the sand fraction of the Indiana dunes exert a significant influence on the distribution of minerals in the silt and clay fraction.

Evaluation of weathering potential in a soil should therefore include not only the clay and silt fractions but also the sand fractions (Keilen etal. 1976; Akhtar etal. 1995). To achieve this, the fractionation technique is often employed. It is very common in soil organic matter (SOM) studies and has been adapted by many researchers (Watson and Parsons 1974: Koutika etal. 1997). Others (Wilson and Logan 1976; Wilson 1976; Gudmundsson and Stahr 1981; Morras 1995) have used the technique to study distribution in cation exchange capacity (CEC) and nutrients within the soils. In Nigeria, Unamba-Oparah etal. (1987) observed that the knowledge of mineral distribution in the soils is useful for the understanding of the relationships between the minerals and the nutrient capacity of the soil. In Germany, Keilen et al. (1976) demonstrated that certain elements, mainly cations, were lost through weathering while others accumulate.

Studies on the distribution of minerals within fractions in Nigerian soils are non-existent. Available studies are mainly on the mineralogy of clay fractions and whole fine-earth fractions. The objectives of this study were to determine the occurrence, distribution, and weathering transformations of minerals in 5 floodplain soils of the River Niger. We selected 5 soils deposited at various stages on the floodplain, with the assumption that they represent an age sequence.

Materials and methods

Soils

The location, climate, and the vegetation pattern of the study area have been reported by Igwe and Stahr (2004). The characteristics of the soils used for this study are shown (Table 1). The detailed characteristics have also been discussed (Igwe and Stahr 2004). The oldest and old depositions (Table 1) represented by profiles 1 and 2 have less silt in their particle size fractions than the other profiles and arc well sorted with clay. In profiles 3, 4, and 5, less coarse sand is obtained. This phenomenon is attributed to the stage of material deposition and perhaps rapid physical weathering due to abrasion and grinding of coarse materials by the action of the river. The soil pH is generally low (3.8 4.9). Total carbon and total nitrogen are low in the soils, an indication of rapid mineralisation of SOM even under partially submerged conditions (Igwe et al. 1999). The low level of SOM in these soils is also reflected in the low CEC. This is a general problem for these soils and adjacent upland soils (Kang and Juo 1981). Most soils of the floodplain are derived from highly weathered upland soils of the river watershed. Therefore, the CEC and the major soil properties of the floodplain soils are a reflection of the sedimentary materials in which the soils have formed. The dithionite-extractable Fe ([Fe.sub.di]) of the soils especially for the profiles representing the middle, recent, and very recent depositions were generally higher than those of oldest and older depositions (Table 1).

Generally, the soils have high bulk density of 1.50 1.85 Mg in:. Structural degradation of the soil, irregular packing of the sand grains due to constant flooding and drying, reworking and sorting of the soils resulting in pedocompaction, may have been the reason for the relatively high bulk density. The soils are all described as Inceptisols in the USDA Soil Taxonomy (Soil Survey Staff 1999). Specifically. profile 1 is classified as Typic Tropaquept and profile 2 Typic Endoaquept while profile 3 is Dystric Durochrept. Profiles 4 and 5 are Aquic Eutropept and Fluventic Eutropept, respectively.

Geology

The underlying geological materials are mainly recent alluvial deposits (Orajaka 1975). These materials have been transported from the upstream of the River Niger and its tributaries. These geological materials also include those described by Ofoegbu (1985). According to him, these materials were from the Benue Trough and are mainly weathered materials from the Basement Complex. Oyawoye (1972) observed that the Basement Complex consists mainly of quartzofeldspathic migmatites and gneisses with occasional quartzites, marbles. and amphibolites.

Field study

Five profile pits were dug along depositional stages of the floodplain m such a way that the profiles were not less than 2 km from each other (Fig. 1). The 5 profiles were located in such a way that each represented the 5 depositional stages: thus, oldest, older, middle, recent. and very recent depositions were represented by profiles 1,2,3,4, and 5, respectively. The soils have developed under an aquic moisture regime and an isohyperthermic temperature regime. Most of the soils show iron oxide concentration with depth from the topsoil or directly below the topsoil. Gleysation was the dominant visible pedogenic process in the soils. The soil profiles are young and have diagnostic horizons (e.g. cambic) to qualify them as Inceptisols.

[FIGURE 1 OMITTED]

Laboratory methods

Particle size distribution of the <2-mm fractions was measured by the hydrometer method as described by Gee and Bauder (1986). Bulk density was determined by clod method (Blake and Hartge 1986). The soil pit value was measured in 1:2.5 suspensions of soil in 0.1 M KCl. Total organic carbon was determined by loss on ignition method using the LECO equipment. Dithionite-citrate-bicarbonate [Fe.sub.di], [Al.sub.di]. and [Mn.sub.di] were determined by the extraction method of Mehra and Jackson (1960). The ammonium oxalate extractable [Fe.sub.ox], [Al.sub.ox], and [Mn.sub.ox] were determined by Schlichting and Blume (1966) method.

For the determination of minerals and elemental oxides in sand, silt and clay fractions. <2-mm fractions were mechanically fractionated in deionised water with sieves after shaking for 24 h without chemical treatments. The sand fraction was separated with sieve and the silt and clay fractions were separated through a series of high-speed centrifugation. X-ray diffractometry (XRD) of fine sand and silt was determined as powder, and that of clay was determined with oriented clay specimens using the Siemens D500 diffractometer with Ni-filter and CuK[alpha]-radiation. Prior to mounting the oriented clay, subsamples were first pre-treated as follows: saturation with Mg, solvation with glycerol saturation with K, and subsequent heating of the samples to 110[degrees]C, 350[degrees]C, and 550[degrees]C. The aim was to be able to isolate the different mineral forms, because of their variable properties under different temperature regimes. The semi-quantitative evaluation of minerals in the soil fractions was determined using the computer package DIFFRAC AT V3.3 Siemens 1993. This method assumes that the available minerals in the sample sum to 100% and the individual mineral is a fraction of the total. Also in these fractions, and whole soils, oxides of elements were determined using Siemens SRS 200 X-ray fluorescence (XRF). In this method the <2.00-mm soil samples and the fractions were ground to fine powder, out of which 2.666g was weighed and further ground and mixed with 1.333g cellulose material with the trade name "Spectromelt C10 Merck'. This material acts as binding agent to the soil sample to be analysed, The sample was then transferred to tablet making machine and turned into tablet with a pressure of 30t for 5 min. The soil/cellulose sample in a ratio of 2:1 was then analysed using XRF against standards to obtain the concentration of the elements which was further converted to their oxide forms using factors. The elements determined in this manner are referred to as total elements and are expressed as oxides. These standards were of soil and rock materials of various mineralogies purchased commercially from Breitlander Eichproben und Labormaterial GmbH (Hans-Sachs-Strasse 12, D-59077 Hamm, Germany, www.breitlander.com). The standards, of which there were 50, were prepared in the same way as the soil samples and analysed with XRF using the Cr-Tube at the intensity of 50 kV and 50 mA. This method has been adapted in some other studies (Singer et al, 1998; Stahr et al. 2000). Thereafter, the output was calibrated using a Siemens computer package 'Spectra 300 V2.0' (Siemens 1993). Scanning electron microscopy (SEM) LEO 420 equipment was used to examine some quartz and feldspar grains in the silt and title sand fractions.

Data analysis

The values of soil minerals and elements measured were subjected to principal component analysis using the SYSTAT on SPSS 10 computer package (SPSS 1999). This method was able to decompose many soil properties relating to kaolinite to only 5 principal components. For improved interpretation, varimax rotation was applied to the eigenvectors or loadings. The ultimate goal in rotation is to obtain some theoretically meaningful factors and if possible the simplest factor structure. Varimax rotation had the effect of reallocating variances to higher components as well as increasing the number of components order than the few components produced before rotation. Therefore, with varimax rotation, the maximum number of components is produced.

Results and discussions

Dynamics of Fe oxide

The values of [Fe.sub.di] and [Fe.sub.ox] for the soils are shown (Table 1). The [Fe.sub.di] is the crystalline plus the amorphous or poorly crystalline forms of Fe, whereas the [Fe.sub.ox] is the amorphous form. [Fe.sub.2][O.sub.3] in these various forms was prominent in these soils (Table 1 and Table 2). The dominance of crystalline forms of Fe over amorphous forms (Table 1) may have been due to the age of these soils or the age of the sediments in which these soils formed. Schwertmann and Kampf (1983) showed …