Printer Friendly

Variation in soil strength, bulk density and gravel concentration along a toposequence in Abeokuta, south-western Nigeria.

Abstract. This study was carried out at Abeokuta, south-western Nigeria, to understand the variation in soil strength, gravel distribution, and bulk density along a toposequence. In 2003, a 120-m transect on a fallowed land was sampled at every 1 m for topsoil bulk density measurement by excavation (3278 [cm.sup.3] pits), while soil strength was measured at every soil depth increment of 25 mm to 0.50m depth. Total dry ([[rho].sub.t]) and fine earth (<2 mm) ([[rho].sub.f]) bulk densities were determined. Soil water content was also determined. Gravel was divided into classes of 2-4, 4-8, 8-16, and >16 mm. In 2006, four 100-m transects were considered; two each on adjacent fallowed and cultivated lands. Soil strength and water content were measured. The fine earth fraction of topsoil ranged from 62 to 90.6%. Gravel in the 2-4 mm class was dominant with a range of 0.8-35.7%. Thus, cores [greater than or equal to] 50mm could be used in the topsoil to obtain reliable estimates of bulk density.

Total bulk density ([[rho].sub.t]) was reduced by 4-19% when corrected for gravel to obtain [[rho].sub.f]. Soil strength of the lower slope was highest in 2003 (1981-4482 kPa) and lowest in 2006 (1546 kPa). In spite of the apparent significant influence of water content on soil strength, the relationship was weakly expressed by regression analysis, as only 35% of variation in soil strength was explained by water content at 0.10-0.15 m soil depth in 2003. No relationship was found in 2006; the cultivated segment had higher soil strength (2045 kPa) than the fallowed segment (1970 kPa) even though the water contents were similar. Also, only the 2-4 mm gravel significantly influenced [[rho].sub.t]. Land use, soil depth, and slope position significantly affected soil strength. Root-limiting soil strength (>2000 kPa) would certainly be encountered below 0.20 m soil depth in the wet season irrespective of land use. Management of this gravelly landscape must be based on the heterogeneous nature of soil physical properties along the toposequence, and this could be made effective by grouping the soils according to slope position and taking interest in the few portions of the landscape with extreme values of gravel distribution and high soil strength.

Additional keywords: penetrometer resistance, rock fragments, fallow, cultivation, compaction.

Introduction

Soils along a toposequence vary strongly according to their position on the slope, particularly in the tropics (Landon 1991; Schaetzl and Anderson 2005; Salako et al. 2006a; Fasina et al. 2007). This variation becomes more complicated with the presence of rock fragments. Rock fragments are unattached pieces of rock >2 mm in diameter that are strongly cemented or more resistant to rupture (Soil Survey Staff 1993). According to the Soil Survey Staff (1993) classification, pebbles are in the size range of 2-75mm with subclasses as fine pebbles (2-5mm), medium pebbles (5-20mm), and coarse pebbles (20-75 mm). Soils with abundant pebbles are gravelly, and the relative abundance of each of the 3 classes enables description as fine gravelly, medium gravely, and coarse gravelly, respectively. For most of the soils formed on basement complex in southwestern Nigeria, there is an extensive distribution of pebbles or gravel in the profiles, making them highly heterogeneous, laterally and vertically, in physical properties (Moormann et al. 1975; Lal 1987; Lal and Shukla 2004).

Presence of rock fragments in soil profiles increases soil strength and prevents even normally deep-rooted plants, such as trees, from exploring the soil deeply (Salako et al. 2002, 2007; Rodrigue and Burger 2004; Fasina et al. 2007). Apart from impeding plant root growth, they also create problems during soil sampling for estimation of properties such as bulk density and soil C pool (Harrison et al. 2003) as well as soil hydraulic properties (Sauer and Logsdon 2002; Salako et al. 2007). Thus, the 'standard' core method may not provide accurate estimates of soil bulk densities in gravelly soil, and there may be a need to make corrections for the presence of gravel (Blake and Hartge 1986; Lal and Shukla 2004). However, the presence of rock fragments on surface soil can be an advantage in soil conservation, as they serve as mulch that prevents soil loss through erosion and water loss by evaporation (Poesen and Lavee 1994; Nyssen et al. 2001). Gravel mulch technology is widely applied in semi-arid and arid areas.

Cotching and Belbin (2007) found that there was a well defined relationship between penetrometer resistance and soil water content, especially at 0.15-0.30m soil depth for structurally degraded soils, with penetrometer resistance increasing as soil water decreased, whereas in well-structured soils, variations in soil wetness had less influence on penetration resistance. The implication of this was that root growth would be impeded more as soil dried in the structurally degraded soil than in the well-structured soil. Also, Sojka et al. (2001) reported that soil strength relationships with water content and bulk density were poor but could improve if data were segregated by depth; this improvement was better with soil water than bulk density.

Generally, soil bulk densities of the gravelly soils of southwestern Nigeria are <1.6 g/[cm.sup.3] (Lal 1987, 1997; Hulugalle and Ezumah 1991; Salako et al. 2006a, 2006b, 2007). Only a few of these soil bulk density determinations used the excavation method (e.g. Lal 1997). Rather, the more convenient core method (Blake and Hartge 1986; Harrison et al. 2003) was usually adopted. Lal (1996) reported that the mean of upland surface soil bulk densities of forests in Ibadan, south-western Nigeria, was 0.73-1.3 g/[cm.sup.3], while soil penetrometer resistance was 32-91 kPa. Salako et al. (2007) obtained penetrometer resistances between 500 and 1500 kPa for topsoils on similar gravelly soil at Abeokuta, south-western Nigeria. In the inland valleys or floodplain, Ogban and Babalola (2003) reported a soil bulk density range of 0.56-1.67 g/[cm.sup.3]. These bulk density measurements were generally not corrected for the gravel content, even when it might be desirable to do so (Franzen et al. 1994; Brye et al. 2004; Lal and Shukla 2004). Oyedele and Aina (1998) reported 2-49% of gravel content in soil profiles at Efon Alaye, while Salako et al. (2006a) reported 4-36% gravel content in soil profiles at Ibadan and Alabata in south-western Nigeria. Ley et al. (1989) found that gravelly soils exhibited hardsetting properties with increases in soil strength as water content decreased, and this was more pronounced for cultivated (mechanised or tilled) than for less disturbed soils.

Extensive in situ measurements are necessary to account for variability that exists in the field on the relationships between soil strength, bulk density and water content (Sojka et al. 2001). The aim of the present study, therefore, was to determine the variations in soil strength (penetrometer resistance), bulk density, and soil water content along a toposequence in southwestern Nigeria.

Materials and methods

Study site

This study was carried out in August and October 2003, and in October 2006 at the University of Agriculture, Abeokuta, Alabata road (7[degrees]17'N, 3[degrees]26'E), south-western Nigeria. The location was the same as that described by Salako et al. (2007), with a fallow vegetation of about 7 years by 2003. Mean annual rainfall for 23 years (1983-2005) was 1176 mm, with a bimodal distribution. Peak rainfall was in July (177 mm) and September (195 mm). The climate is humid, with maximum daily temperature ranging from 31[degrees]C in August to 39[degrees]C in February. The vegetation is a forest-savanna mosaic, with the soils formed over the basement complex. Gravelly Alfisols dominate the landscape while Inceptisols and Entisols can be found at the lower slope position or footslope (Ahn 1970; Moormann et al. 1975; Salako et al. 2006a).

According to Salako et al. (2007), for this study site, sand content at 0-0.45 m depth for the lower slope (82%) was significantly higher than that of the upper slope (72.6%), whereas the clay content (16.7%) of the upper slope was higher than that of the lower slope (9.2%); gravel concentration was higher at the lower slope position (20.7%) than the upper slope position (14.4%). Salako et al. (2007) reported further that soil pH was significantly higher at the upper (6.1) than the lower slope (6.0), although organic C (1.43-2.01%), exchangeable K (0.37-0.53 cmol/kg) and Ca 3.8-4.3 cmol/kg) contents were similar.

Soil gravel distribution, bulk density and penetrometer resistance measurements and water content in 2003

A 120-m-long transect was cut through the natural fallow vegetation, very close to the site reported by Salako et al. (2007), in August 2003. The transect ran from the upper slope position, through the middle slope, to the lower slope position which was close to a stream. Soil was excavated up to 0.20 m depth and collected for dry bulk density determination at every 1-m interval of the transect. The dimensions of each excavated hole were measured with a ruler; mean volume of soil sampled was 3278 [cm.sup.3] (s.d. 30 [cm.sup.3]). The bulk field soil sample was weighed ([M.sub.t]). After this, a subsample was taken to determine the gravimetric water content ([[theta].sub.g]). Thus, the dry weight ([M.sub.d]) of the field sample, [M.sub.t], was obtained as:

[M.sub.d] = [M.sub.t]/1 + [[theta].sub.g] (1)

The remaining excavated field samples were air-dried and re-weighed before sieving to separate the fine-earth (<2 mm) and gravel (>2 mm) fractions. The gravel fraction was sieved further into 2-4, 4-8, 8-16, and >16 mm classes. Each fraction was weighed after sieving. Thus, proportions of fine earth and the different classes of gravel were calculated.

Also, samples of each class of gravel were weighed in the laboratory before placement in a graduated beaker of water to determine the volume of water displaced by each set of samples. From these data, the density of each class of gravel was calculated as mass of gravel divided by its volume. The average density of 2.20 g/[cm.sub.3] of all the classes was used to obtain the total volume of gravel excavated in the field by calculation as total mass of gravel divided by average gravel density. Based on these data, total soil bulk density (with gravel) was corrected for gravel to obtain the bulk density of the fine earth fraction as follows:

[[rho].sub.t] = [M.sub.d]/[V.sub.t] (2)

[[rho].sub.f] = ([M.sub.d] - [M.sub.gr])/([V.sub.t] - [V.sub.gr]) (3)

where [[rho].sub.t] is total dry soil bulk density (i.e. with gravel) in g/[cm.sub.3]; [[rho].sub.f] is the fine-earth bulk density (g/[cm.sup.3]); [M.sub.d] is oven-dry weight of excavated soil (g); [M.sub.gr] is total mass of gravel (g); [V.sub.t] is volume of excavated soil ([cm.sup.3]); and [V.sub.gr] is the volume of excavated gravel ([cm.sup.3]).

This procedure of calculating fine-earth bulk density is similar to that of Brye et al. (2004), although they assumed a gravel density of 2.8 g/[cm.sup.3] because of the lithology of their environment.

In October, 2003, penetrometer resistance was measured at every 1-m interval using the RIMIK CP20 cone penetrometer (Agridry Rimik Pty Ltd 1994) along the same transect used for soil bulk density sampling but at corresponding points which were not disturbed. The penetrometer, with apex cone angle of 30[degrees], was set to read penetrometer resistances in kPa to a datalogger at every 25 mm depth up to 0.50 m depth. The maximum recordable reading was 5000 kPa. This maximum reading was assumed to have been attained wherever the reading terminated due to the presence of impenetrable layer or gravel. Three measurements were taken at each 1-m interval up to 120m from the upper to the lower slope positions. Measurements were taken after the drainage of excess rainfall or theoretical attainment of field capacity moisture (Hillel 1998; Jury and Horton 2004) as penetrometer readings were very difficult to obtain when the soil was dry. Penetrometer readings were downloaded into a computer for further analysis. Gravimetric water content was determined at every 5-m interval of penetrometer measurements at 0-0.05, 0.05-0.10, 0.10-0.15, and 0.15-0.20 m soil depths. Field soil samples taken at these depths were weighed and oven-dried at 105[degrees]C before reweighing to determine gravimetric water content (Hillel 1998; Jury and Horton 2004).

Determination of penetrometer resistance and soil water content in October 2006

In October 2006, 2 transects, each, were cut through the natural fallow and cultivated segments of the landscape from the upper to the lower slope positions. The two transects on each segment were 10m apart. The fallowed segment was about 10 years under fallow by 2006 and the transects cut were on new lines. The cultivated segment, which was adjacent to the fallowed segment, was cleared mechanically in June/July 2000 (when the soil was wet or very moist; Table 1) and had since been under mechanised cultivation, particularly to sunflower (Helianthus annuus) and sesame (Sesamum indicum) (V. I. Olowe, pers. comm.).

In 2006, factors in the study were land use (fallowed and cultivated), slope position (upper, middle and lower slopes), and soil depth (0-0.15, 0.15-0.30, and 0.30-0.50 m) using the same penetrometer and its settings as in 2003. Penetrometer readings were recorded at every 1-m interval up to 100 m distance along the slope. However, only 1 measurement was taken per interval to 0.50 m depth. Penetrometer readings were taken within a day after a rainfall event with measurements of 2 transects, one fallowed and one cultivated, completed in a day. Soil samples were taken at 0-0.15 and 0.15-0.30 m depths to determine soil water content gravimetrically for every 10-m interval of all transects. These were converted to volumetric water content with total soil bulk density data.

Data analysis

All of the penetrometer data for the 0-25 mm soil depth were excluded from data analysis, as readings were influenced by surface litter and were highly variable. Also, a few bulk density and gravel concentration data were not included in analysis. The transects were divided into 3 slope positions based on distances observed on the field, namely, upper, middle, and lower slope positions. Also, penetrometer readings were averaged according to depth ranges of 0-0.20, 0.20-0.35, and 0.35-0.50m for the 2003 data. For the 2006 data, penetrometer readings were averaged according to depth ranges of 0-0.15, 0.15-0.30, and 0.30-0.50 m. Thus, gravel concentration and bulk density data collected in 2003 were compared with the 0-0.20m average penetrometer resistance, while soil water content measured in 2006 was compared with 0-0.15 and 0.15-0.30 m penetrometer resistances.

Descriptive statistics (mean and coefficient of variation) were used and 1-way analysis of variance was carried out (Analytical Software 1998) to determine differences between slope positions and soil depth in 2003. In 2006, data were analysed to determine the effects of land use (fallowed and cultivated), slope position (upper, middle and lower slopes), soil depth (0-0.15, 0.15-0.30, and 0.30-0.50 m) and their interactions on soil strength and water content using SAS GLM Procedure, with paired comparisons (t-tests) of means (SAS Institute 1999). Regression analyses between soil strength and bulk density, gravel content, and water content were also carried out.

Results and discussion

Distribution of gravel in topsoil

The fine earth fraction was the dominant constituent of topsoil (0-0.20m depth) with mean values of 62-91% (Table 1). However, there were locations within the middle and lower slope positions with up to 74-95% gravel content in the topsoil. The means suggest that extremely high gravel concentrations were not extensive within these slope positions. Furthermore, very gravelly topsoils (>35%, Soil Survey Staff 1993 classification) were found in all slope positions but more in the middle and lower slopes.

Gravel in the 2-4 mm class was dominant when compared with other gravel classes (Table 1). For each class, the middle and lower slopes had higher proportions of gravel than the upper slope. The trend observed was a decreasing proportion of gravel with increasing size from 2 to >16 mm. The coefficient of variation increased with gravel size, especially at the upper slope position. It was also observed that at some locations on the landscape there were no gravel sizes >4mm. Thus, fine pebbles or gravel (2-5 mm; Soil Survey Staff 1993) were common in the topsoil compared with medium (5-20mm) to coarse (20-75 mm) pebbles or gravel, which were rare. The topsoil was, therefore, fine gravelly. The wide variability in gravel distribution could result in variable productivity of soil along the landscape and render fertiliser application ineffective in some parts of the landscape if the fertiliser drops on gravel. Salako et al. (2007) reported the mean concentration of gravel at the upper and lower slope positions at the same site as ranging between 8 and 27% within 0-0.45 m soil depth, with the topsoil having 8-19% concentration. At Ibadan and Alabata, near Ibadan, mean gravel concentration in the soil profile was between 4 and 36% (Salako et al. 2006a). In another location in south-western Nigeria, Oyedele and Aina (1998) reported that gravel concentration ranged from 4 to 44% at 0-0.15 m depth and 2 to 49% at 0.30-0.50 m depth. Salako et al. (2006a) stated that for these gravelly Alfisols, gravel concentration was highest at 0.20-1.02 m depth. This deduction was consistent with presence of stonelines in subsoils of this region (Ahn 1970; Poesen and Lavee 1994).

Total and fine earth bulk density of topsoil

The total soil bulk density (with gravel) was higher in the middle than the upper and lower slopes (Table 1). Also, the lower slope had a significantly higher total bulk density than the upper slope. The fine-earth bulk density was still highest in the middle slope but this was only significantly higher than the lower slope bulk density. Coefficients of variation of bulk densities were <20%. At the lower slope, the fine earth bulk density was as low as 0.43 g/[cm.sup.3] (Table 1) because some of the samples included alluvial deposits, with variable materials including organic matter.

The decrease in soil bulk density after correction for gravel was more pronounced in the middle (16% reduction) and lower slopes (19% reduction) than the upper slope (4% reduction). Most of the topsoil bulk density values observed with or without correction for gravel (Table 1) were not root-limiting according to the classification of the Soil Survey Staff/Natural Resource Conservation Service (2001), which places root-limiting values of soil bulk density for soils with <35% clay content between 1.54 and 1.69 g/[cm.sup.3]. Topsoil clay contents were 8.7% at the lower slope and 11.5% at the upper slope at 0-0.20 m depth (Salako et al. 2007).

Most of the available bulk density data for south-western Nigeria are based on core method, without correction for gravel content. Generally, bulk densities were < 1.6 g/[cm.sup.3] (Lal 1987, 1997; Hulugalle and Ezumah 1991; Oyedele and Aina 1998; Salako et al. 2006a, 2006b, 2007). Bulk densities as low as 0.56 g/[cm.sup.3] could be obtained on floodplains in the region (Ogban and Babalola 2003). Thus, bulk densities observed at the different slope positions, with or without gravel, were within typical ranges for each slope position.

Soil strength

In 2003, mean soil strength of the lower slope was higher at every depth range of 0-0.20, 0.20-0.35, and 0.35-0.50 m depths than the upper and middle slopes (Table 2). Coefficients of variation were, however, generally high, except for soil strength measured at 0.35-0.50 m in the lower slope position. Gravelly impediments disrupted penetrometer resistance measurements in all slope positions. Minimum values of soil strength were <1600 kPa. Soil strength at 0-0.50 m depth was generally <2000 kPa at the upper slope position, suggesting that there was no risk of root growth being impeded by high soil strength under moist to wet conditions. It could, however, be tortuous due to the presence of gravel.

The mean gravimetric water content of 0.17 observed for the upper slope in 2003 was significantly higher (P < 0.0001) than 0.14 for the middle slope and 0.13 for the lower slope (Table 2). Also, the mean gravimetric water content at 0-0.05 m, which was 0.16, was significantly higher than 0.14 for 0.10-0.15 m depth and 0.14 for 0.15-0.20 m depth but was similar to 0.16 observed for 0.05-0.10 m depth. The 0.10-0.15 m depth had similar water content to the 0.05-0.10 and 0.15-0.20 m depths.

In 2006, land use, slope, and soil depth, and their interactions had significant effects on soil strength (Table 3). The cultivated segment of the landscape had significantly higher soil strength than the fallowed segment (Table 4). Also, soil strength increased significantly with depth for 0-0.15, 0.15-0.30, and 0.30-0.50 m depths. The upper slope position had significantly higher soil strength than the middle and lower slope positions, while the middle slope also had higher soil strength than the lower slope in 2006. The interaction of soil depth and slope position showed that each soil strength value was significantly higher than the next value when ranked from the highest to the lowest value, except for the middle and lower slope 0-0.15 m depths, which had similar values.

Small-depth increment of soil strength showed that most of the significant differences in soil strength along the slope occurred in the subsoil (Fig. 1). Considering the fallowed segment alone in 2006, the mean soil strength for the middle slope position (2318 kPa) was similar to that of upper slope position (2302 kPa) but the 2 slope positions had significantly higher (P < 0.0001) soil strength than the lower slope position (2020 kPa). For the cultivated segment alone in 2006, the upper slope had a significantly higher (P = 0.0006) soil strength of 3037 kPa than the middle slope with 2444 kPa and lower slope with 1551 kPa. Furthermore, the middle slope position had a higher soil strength than the lower slope.

Influence of soil water, gravel concentration, and bulk density on soil strength

In 2006, the cultivated and fallowed segments, as well as soil depths of 0-0.15 and 0.15-0.30 m had similar water contents (Table 5). However, the lower slope position had a significantly higher water content than the upper slope but similar to the middle slope position. The interactions showed that the differences were due mainly to the very low water content at 0.15-0.30 m depth at the upper slope position. The same trends were observed with gravimetric water contents.

In 2003, soil strength (dependent variable) was only significantly (P = 0.00072) and negatively related to gravimetric water content between 0.10 and 0.15 m depth (coefficient of determination, [r.sup.2] = 0.35; n = 19) with a log-log-transformation of data. In 2006, no significant relationship (P > 0.05) was observed between soil strength and water content (gravimetric and volumetric), either by pooling data together (n = 71) or by segregating them according to depth (n = 18) or land use or by logarithmic transformation.

The relationship between the fine earth bulk density and total bulk density is as follows:

[[rho].sub.f] = 0.378 + 0.577 [[rho].sub.t], P < 0.0001, [r.sup.2] = 0.28, n = 121 (4)

where Or and pt are fine earth and total bulk densities (g/[cm.sup.3]), respectively, as in Eqns 2 and 3. A forward stepwise regression analysis of total bulk density (dependent variable) against proportions (%) of <2, 2-4, 4-8, 8-16, and >16mm particle classes (independent variables) showed that only the <2 mm fraction influenced soil total bulk density, [[rho].sub.t], significantly (P < 0.0001, [r.sup.2] = 0.19; n = 120) in a negative way (slope, b = -0.004). Also, a stepwise regression analysis of soil strength (kPa) against [[rho].sub.t], [[rho].sub.f], and proportions (%) of <2, 2-4, 4-8, 8-16, and >16 mm particle classes showed that only the gravel size between 2 and 4 mm significantly influenced soil strength (P < 0.0001, [r.sup.2] = 0.48, n = 96) with a positive slope value of 79.203.

The significant increase in soil strength with soil depth (Tables 1-4; Fig. 1) could be due to overburden pressure (Sands et al. 1979). Such a pressure, particularly shear stress, could also explain the relatively intermediate values of soil strength in the middle slope position (Tables 1, 4, Fig. 1). In 2003, the lower slope position had the highest soil strength (Table 1, Fig. 1), whereas in 2006, it had the least (Table 4, Fig. 1). High or root-limiting soil strength >2000 kPa (Soil Survey Staff 1993) was common from 0.20 m depth downward (Fig. 1). Land use had a significant influence on soil strength in 2006 (Tables 2 and 3) and the segregation of data for separate analysis of variance by land use still showed that the fallowed segment followed the overall trend observed with pooled data, in which the upper slope had the highest soil strength and the lower had the least in 2006. Thus, the cultivated segment only accentuated the differences in soil strength among the slope positions. Considering the fallowed segment, the change in the trend of soil strength values between 2003 and 2006 could be due to maturity of the fallow vegetation, which would influence root growth, hence soil structure (Salako et al. 2002). Higher soil strength of the cultivated segment in spite of similar water content to the fallowed plot (Tables 4 and 5) was due to soil compaction due to mechanised clearing and tillage operations (Franzen et al. 1994), which must have been aggravated by the fact that clearing was done in June/July 2000 (V. I. Olowe, pers. comm.) when the soil was very moist or wet (Ghuman and Lal 1992). Thus, for equivalent moisture content, the cultivated segment had significantly higher soil strength than the fallowed segment. Another factor was that the soils were prone to hardsetting as soon as they were opened up for cultivation (Ley et al. 1989), and this would increase densification and strength of cultivated soils. Ley et al. (1989) reported that increase in soil strength as water content decreased was more pronounced for cultivated (mechanised or tilled) than for less disturbed soils. Cotching and Belbin (2007) stated that in fields with degraded soil structure (e.g. cultivated fields) in Tasmania, Australia, there were highly significant relationships between soil strength and soil wetness at 0.15-0.30m depth, while in well-structured soils (e.g. fallowed fields) variation in soil wetness had less effect on soil strength. The implication of this was that there would be greater resistance to root growth at drier soil water contents in the degraded structure than in the well-structured soils. This deduction is supported by the data in Fig. 1 because the fallowed and cultivated soil strength data were collected under similar water state (Table 5), yet there were higher soil strengths for the cultivated than the fallowed segment (Table 4; Fig. 1). The fact that land use, soil depth, and slope position had significant effects individually and interactively on soil strength (Tables 2 and 4) suggests that variations in these factors could complicate the relationships between soil strength, soil bulk density, and water content. Robertson et al. (1993) noted that spatial pattern and scale of soil variability can differ markedly among edaphically identical sites and the differences can be related to disturbance history.

[FIGURE 1 OMITTED]

The influence of soil water on soil strength for the topsoil was not explicit as shown by regression analyses and analysis of variance (Tables 3 and 4). For instance, only 35% of variation in soil strength was explained by gravimetric water content and this was only at 0.10-0.15 m soil depth in 2003. In 2006, no significant relationship was found between soil strength and soil water. This shows that in situ relationships between soil strength and soil water are very complex, being influenced by other factors which cannot be explained by a simple linear regression between soil strength and water (Sojka et al. 2001). Cotching and Belbin (2007) observed that soil water content only explained 23-41% of the variation in soil strength significantly (P < 0.001) in Tasmania, Australia.

Variability in soil bulk density along the slope was also poorly explained by the fine earth fraction ([r.sup.2] = 0.19), which was the only parameter that was relevant in a stepwise regression analysis that included particles or gravel <2, 2-4, 4-8, 8-16, and >16 mm. Furthermore, the fine earth bulk density was significantly related to total bulk density but with only 28% of its variation explained by total bulk density. The implication of these results was that the high variability of gravel distribution (Table 2) could still confound relationships between bulk density and soil particles, as this would affect the arrangement of soil particles, hence soil structure. Bulk density does not explain how soil particles are arranged, in spite of its being used as a soil structural parameter. The effect on soil structure will invariably affect soil water, especially in a moist to saturated state (Hillel 1998), thereby confounding further the relationship of soil strength, bulk density, and soil water. Since the most important particle class influencing soil bulk density was the fine earth fraction (<2 mm), and the 2-4 mm gravel was also significant in its influence on soil strength, it follows that reliable measurements of topsoil bulk densities could be obtained with cores [greater than or equal to] 50 mm diameter, commonly used in south-western Nigeria. Coarse gravel or pebbles are more common in the subsoil than topsoil, based on visual observation of profiles at Ibadan and Alabata, southwestern Nigeria (Salako et al. 2006a). It is at these depths, particularly from 0.20 to 1.02 m that significant errors can be introduced when sampling for bulk densities with soil cores. Densification of subsoils was also feasible in the soils because gravel is enmeshed in a clayey subsoil, with clay content that could range from 19 to 36% (Salako et al. 2006a; Fasina et al. 2007), under an overburden pressure. However, high soil strength and bulk density of the subsoil would be of advantage for engineering construction (Soil Survey Staff 1993; Soil Survey Staff/Natural Resource Conservation Service 2001).

Salako et al. (2007) observed that the relationship between bulk density and soil strength was not significant at the upper and lower slope positions of this site, while soil water content explained 19-26% of the variation in soil strength at the lower slope position. In their study, they noted that water content could become irrelevant as gravel concentration increased from 10 to 50% and gravel size would only influence soil strength in interacting with water content and gravel concentration. This supports the deduction that the relationships were more complex than could be explained with linear equations. Sojka et al. (2001) stated that the difficulties in developing comprehensive relationships of soil strength and bulk density, and the overriding dependence of soil strength on the dynamic variable of water content, implies great uncertainty when using bulk density for assessment of root growth or crop performance, unless adequate sampling and analysis are carried out. Large variations of some soil physical properties such as gravel concentration (Table 1) imply that precise management of the soil according to physical properties of each slope position would enhance soil productivity. Buchter et al. (1991), using a geostatistical approach, observed a spatial dependence of 1 m for soil physical properties on a slope. This suggests that the approach used in this study reflected substantially the spatial dependence of soil strength and bulk density on the landscapes. This study showed that short range variability of soil physical properties had implications for variability in plant growth on the toposequence, as observed by Meredieu et al. (1996). Gravel distribution and soil strength would not only influence root growth and distribution but also influence ease of cultivation of soil on the landscape and effectiveness of applied agro-chemical inputs or soil amendments.

Conclusion

This study showed that the fine earth fraction (<2 mm) was the dominant component of soil in topsoil when compared with gravel content on a landscape in south-western Nigeria. Among the gravel classes which ranged from 2 to >16 mm, the 2-4 mm gravel was dominant. Few portions of the different slope positions (upper, middle, and lower slopes) were observed to be very gravelly (>35% gravel). This high variability in gravel concentration must be taken into consideration in soil management, particularly in application of soil fertility amendments, and also in undisturbed core sampling. Core samples ([greater than or equal to] 50-mm-diameter cores) at the topsoil would still be generally representative of the landscape if the few gravelly portions are avoided because of the dominance of the fine earth fraction and the fact that it was only this fraction and 2-4 mm gravel that explained significantly (though weakly) variations in soil strength and bulk density. Correction for gravel resulted in 4-19% reduction in bulk density.

Land use, slope, and soil depth, and their interactions had significant effects on soil strength. The relationships between strength, soil bulk density, and water content were weakly expressed, though significant. For equivalent soil moisture content, the cultivated segment had significantly higher soil strength than the fallowed segment. High or root-limiting soil strength values (>2000 kPa) would be encountered below 0.20 m depth in the wet season, irrespective of the land being under fallow or cultivation. However, high soil strength and bulk density would be of advantage for engineering construction. Precise management of different portions of the landscape based on the variability in soil physical properties in each slope position is required to optimise soil productivity.

Acknowledgments

The authors wish to thank Iro A., Oumarou A., Monday Joseph, Bello Ayoade, Bisi Sonuga, and Temitope. A. Adewole, formerly of the Department of Soil Science and Land Management, University of Agriculture, Abeokuta, for their technical and field support. The paper was written when the first author was on sabbatical leave in the Department of Environmental Sciences, University Ca'Foscari of Venice (UNIVE), Venice, Italy, under the Training in Italian Laboratory Programme (TRIL) of the International Centre for Theoretical Physics (ICTP), Trieste, Italy. He expresses his gratitude to Profs. G. M. Zuppi (UNIVE) and G. Furlan (TRIL-ICTP) for their encouragement and support.

Manuscript received 15 May 2007, accepted 18 October 2007

References

Agridry Rimik Pty Ltd. (1994) 'Rimik CP cone penetrometer.' (Agridry Rimik Pty. Ltd: Toowoomba, Qld)

Ahn PM (1970) 'West African soils.' (Oxford University Press: Oxford, UK)

Analytical Software (1998) 'STATISTIX for Windows. Version 2.' (Analytical Software: Tallahassee, FL)

Blake GR, Hartge KH (1986) Bulk density. In 'Methods of soil analysis. Part 1. Physical and mineralogical methods'. (Ed. A Klute) pp. 363-375. (ASA-SSSA: Madison, WI)

Brye KR, Morris TL, Miller DM, Formica SJ, Van Eps MA (2004) Estimating bulk density in vertically exposed stoney alluvium using a modified excavation method. Journal of Environmental Quality 33, 1937-1942.

Buchter B, Aina PO, Azari AS, Nielsen DR (1991) Soil spatial variability along transects. Soil Technology 4, 297-314. doi: 10.1016/0933-3630(91)90008-B

Cotching WE, Belbin KC (2007) Assessment of the influence of soil structure on soil strength/soil wetness relationships on Red Ferrasols in north-west Tasmania. Australian Journal of Soil Research 45, 147-152. doi: 10.1071/SR06113

Fasina AS, Omolayo FO, Falodun AA, Ajayi OS (2007) Granitic derived soils in humid forest of southwesthern Nigeria--Genesis, classification and sustainable management. American-Eurasian Journal of Agricultural and Environmental Sciences 2, 189-195.

Franzen H, Lal R, Ehlers W (1994) Tillage and mulching effects on physical properties of a tropical Alfisol. Soil and Tillage Research 28, 329-346. doi: 10.1016/0167-1987(94)90139-2

Ghuman BS, Lal R (1992) Effects of soil wetness at the time of land clearing on physical properties and crop response on an Ultisol in southern Nigeria. Soil and Tillage Research 22, 1-11. doi: 10.1016/0167-1987(92)90018-7

Harrison RB, Adams AB, Licata C, Flaming B, Wagoner GL, Carpenter E Vance ED (2003) Quantifying deep-soil and coarse-soil fractions: avoiding sampling bias. Soil Science Society of America Journal 67, 1602-1606.

Hillel D (1998) 'Environmental soil physics.' (Academic Press: San Diego, CA)

Hulugalle NR, Ezumah HC (1991) Effects of cassava-based cropping systems on physico-chemical properties of soil and earthworm casts in a tropical Alfisol. Agriculture. Ecosystems & Environment 35, 55-63. doi: 10.1016/0167-8809(91)90076-A

Jury W, Horton R (2004) 'Soil physics.' 6th edn (John Wiley & Sons Inc.: New Jersey)

Lal R (1987) 'Tropical ecology and physical edaphology.' (John Wiley: Chichester, UK)

Lal R (1996) Deforestation and land-use effects on soil degradation and rehabilitation in western Nigeria. I. Soil physical and hydrological properties. Land Degradation and Development 7, 19-45. doi: 10.1002/(SICI)1099-145X(199603)7:1<19::AID-LDR212 >3.0.CO;2-M

Lal R (1997) Long-term tillage and maize monoculture effects on a tropical Alfisol in western Nigeria. 1. Crop yield and soil physical properties. Soil and Tillage Research 42, 145-160. doi: 10.1016/S0167-1987(97) 00006-8

Lal R, Shukla MK (2004) 'Principles of soil physics.' (Marcel Dekker, Inc.: New York)

Landon JR (1991) 'Booker tropical soil manual: a handbook for soil survey and agricultural land evaluation in the tropics and subtropics.' Paperback edn (Booker Tate Limited/Longman Scientific & Technical: London)

Ley GJ, Mullins CE, Lal R (1989) Hard-setting behaviour of some structurally weak tropical soils. Soil and Tillage Research 13, 365-381. doi: 10.1016/0167-1987(89)90044-5

Meredieu C, Arrouays D, Goulard M, Auclair D (1996) Short range soil variability and its effects on red oak growth (Quercus rubra L.). Soil Science 161, 29-38. doi: 10.1097/00010694-199601000-00005

Moormann FR, Lal R, Juo ASR (1975) 'The soils of IITA.' (International Institute of Tropical Agriculture: Ibadan, Nigeria)

Nyssen J, Halle M, Poesen J, Deckers J, Moeyersons J (2001) Removal of rock fragments and its effects on soil loss and crop yield, Tigray, Ethiopia. Soil Use and Management 17, 179-187. doi: 10.1079/SUM200173

Ogban PI, Babalola O (2003) Soil characteristics and constraints to crop production in inland valley bottoms in southwestern Nigeria. Agricultural Water Management 61, 13-28. doi: 10.1016/S0378-3774(02)00170-1

Oyedele DJ, Aina PO (1998) A study of soil factors in relation to erosion and yield of maize on a Nigerian soil. Soil and Tillage Research 48, 115-125. doi: 10.1016/S0167-1987(98)00110-X

Poesen J, Lavee H (1994) Rock fragments in topsoils: significance and processes. CATENA 23, 1-28. doi: 10.1016/0341-8162(94)90050-7

Robertson GR Crum JR, Ellis B (1993) The spatial variability of soil resources following long-term disturbance. Oecologia 96, 451-456. doi: 10.1007/BF00320501

Rodrigue JA, Burger JA (2004) Forest soil productivity of mined land in the Midwestern and Eastern coal field regions. Soil Science Society of America Journal 68, 833-844.

Salako FK, Dada PO, Adejuyigbe CO, Adedire MO, Martins O, Akwuebu CA, Williams OE (2007) Soil strength and maize yield after topsoil removal and application of nutrient amendments on a gravelly Alfisol toposequence. Soil and Tillage Research 94, 21-35. doi: 10.1016/j.still.2006.06.005

Salako FK, Kirchhof G, Tian G (2006b) Management of a previously eroded Alfisol with herbaceous legumes: soil loss and properties under mound tillage. Soil and Tillage Research 89, 185-195. doi: 10.1016/j.still.2005.07.010

Salako FK, Tian G, Kang BT (2002) Indices of root and canopy growth of leguminous cover crops in the savanna zone of Nigeria. Tropical Grasslands 36, 33-46.

Salako FK, Tian G, Kirchhof G, Akinbola GE (2006a) Soil particles in agricultural landscapes of a derived savanna in southwestern Nigeria and implications for selected soil properties. Geoderma 137, 90-99. doi: 10.1016/j.geoderma.2006.06.017

Sands R, Greacen EL, Gerard CJ (1979) Compaction of sand soils in radiata pine forests. I. A penetrometer study. Australian Journal of Soil Research 17, 101-113. doi: 10.1071/SR9790101

SAS Institute (1999) ' SAS System for Windows. Version 8.' (SAS Institute: Cary, NC)

Sauer TJ, Logsdon SD (2002) Hydraulic and physical properties of stony soils in a small watershed. Soil Science Society of America Journal 66, 1947-1956.

Schaetzl RJ, Anderson S (2005) 'Soils: Genesis and geomorphology.' (Cambridge University Press: Cambridge, UK)

Soil Survey Staff (1993) 'Soil survey manual. Soil Conservation Service.' US Department of Agriculture Handbook 18. (USDA: Washington, DC)

Soil Survey Staff/Natural Resource Conservation Service (2001) 'National soil survey handbook title 430.VI.' (USDA/NRSC: Washington, DC)

Sojka RE, Buscher W J, Lehrsch GA (2001) In situ strength, bulk density and water content relationships of a Durinodic Xeric Haplocalcid soil. Soil Science 166, 520-529. doi: 10.1097/00010694-200108000-00003

F. K. Salako (A,C), P. O. Dada (B), J. K. Adesodun (A), F. A. Olowokere (A), and I. O. Adekunle (A)

(A) Department of Soil Science and Land Management, University of Agriculture, PMB 2240, Abeokuta, Nigeria.

(B) Department of Agricultural Engineering, University of Agriculture, PMB 2240, Abeokuta, Nigeria.

(C) Corresponding author. Email: kfsalako@yahoo.ie or fsalako@ictp.it
Table 1. Means and coefficients of variation of fine earth, gravel
concentration, and bulk density in 2003 at Abeokuta, south-western
Nigeria

U, Upper slope position, n = 40; M, middle slope position, n = 40;
L, lower slope position, n = 41. Within size or density category,
means followed by the same letter are not significantly different
(P > 0.05)

 Fine earth and gravel distribution

 Gravel
 Fine earth (%) concentration (%)

 2-4 mm

 U M L U M L

Mean 90.6a 65.6b 62.0c 6.1 16.7 20.6
Coefficient of 7.2 33.7 26.0 52.3 45.1 28.5
 variation (%)

 Fine earth and gravel distribution

 Gravel concentration (%)

 4-8 mm 8-16 mm

 U M L U M L

Mean 1.4 11.0 10.3 1.0 3.7 3.9
Coefficient of 158 85.3 74.5 185 88.2 72.5
 variation (%)

 Fine earth and gravel
 distribution

 Gravel
 concentration (%)

 > 16 mm

 U M L

Mean 1.2 3.1 3.0
Coefficient of 222 155 138
 variation (%)

 Bulk density with and without gravel

 Fine earth
 With gravel (without gravel)
 (g/[cm.sup.3]) (g/[cm.sup.3])

 Upper Middle Lower Upper Middle Lower

Mean 1.21c 1.42a 1.35b 1.16ab 1.19a 1.09b
Coefficient of 8.10 13.19 11.22 9.45 18.59 18.94
 variation (%)

Table 2. Soil strength and gravimetric water content at various slope
positions and soil depth during soil strength determination in 2003

Within strength category, means followed by the same letter are not
significantly different (P > 0.05). Maximum reading of 5000 kPa,
which was the maximum capacity reading of the equipment, was assigned
at depths at which penetrometer could no longer penetrate the soil
due to contact with rock fragments

Soil strength (kPa)

 0-0.20m (n = 224)

 Upper Middle Lower

Mean 979c 11706 1981a
Coefficient of 74.36 50.38 40.61
 variation (%)

 0.20-0.35m (n = 192)

 Upper Middle Lower

Mean 1482c 1945b 3452a
Coefficient of 66.47 60.61 30.52
 variation (%)

 0.35-0.50m (n = 192)

 Upper Middle Lower

Mean 1987c 29646 4482a
Coefficient of 49.37 46.83 17.38
 variation (%)

Gravimetric water content

 Upper slope (n = 7)

Soil depth (m): 0-0.05 0.05-0.10 0.10-0.15 0.15-0.20

Mean 0.19 0.17 0.16 0.16
Coefficient of 18.8 15.2 6.8 16.8
 variation (%)

 Middle slope (n = 8)

Soil depth (m): 0-0.05 0.05-0.10 0.10-0.15 0.15-0.20

Mean 0.14 0.15 0.13 0.13
Coefficient of 13.1 6.4 13.3 13.8
 variation (%)

 Lower slope (n = 4)

Soil depth (m): 0-0.05 0.05-0.10 0.10-0.15 0.15-0.20

Mean 0.15 0.15 0.12 0.11
Coefficient of 13.7 11.1 10.8 9.8
 variation (%)

Table 3. F-value and probability of significance for effects of
land use, slope position, soil depth and their interactions on
penetrometer resistance (kPa) in October 2006

 Degree of
Source of variation freedom F-value P > F

Land use 1 3.6 0.050
Slope 2 124.13 <0.0001
Land use x slope 2 45.32 <0.0001
Depth 2 1490.99 <0.0001
Land use x depth 2 13.91 <0.0001
Slope x depth 4 19.32 <0.0001
Land use x slope x depth 4 24.07 <0.0001

Table 4. Soil strength (kPa) as affected by land use, slope
position, and soil depth in October 2006

Within comparisons, means followed by the same letter are not
significantly different (P > 0.05)

 Upper slope

Land use 0-0.15 0.15-0.30 0.30-0.50

Cultivated 831 2580 4516
Fallowed 1167 1946 3166
Mean (soil depth)
Mean (slope) 2368a
Mean (soil depth x slope) 999g 2263d 3841a

 Middle slope
 Soil depth (m)

Land use 0-0.15 0.15-0.30 0.30-0.50

Cultivated 712 2145 3573
Fallowed 867 1868 3374
Mean (soil depth) 849c 1863b 3310a
Mean (slope) 2090b
Mean (soil depth x slope) 789h 2006e 3474b

 Lower slope

Land use 0-0.15 0.15-0.30 0.30-0.50 Mean

Cultivated 735 1008 2305 2045a
Fallowed 779 1632 2925 1970b
Mean (soil depth)
Mean (slope) 1564c
Mean (soil depth x slope) 757h 1320f 2615c

Table 5. Volumetric soil water content ([cm.sup.3]/[cm.sup.3])
during soil strength determination in October 2006

Within comparisons, means followed by the same letter are not
significantly different (P > 0.05)

 Upper slope

Land use 0-0.15 m 0.15-0.30 m

Cultivated 0.15 0.08
Fallowed 0.14 0.11
Mean (soil depth)
Mean (slope) 0.12b
Mean (soil depth x slope) 0.15ab 0.09b

 Middle slope

Land use 0-0.15 m 0.15-0.30 m

Cultivated 0.12 0.13
Fallowed 0.19 0.13
Mean (soil depth) 0.17a 0.14a
Mean (slope) 0.14ab
Mean (soil depth x slope) 0.15ab 0.13ab

 Lower slope

Land use 0-0.15 m 0.15-0.30 m Mean

Cultivated 0.14 0.20 0.14a
Fallowed 0.28 0.17 0.17a
Mean (soil depth)
Mean (slope) 0.20a
Mean (soil depth x slope) 0.21a 0.19a
COPYRIGHT 2007 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Salako, F.K.; Dada, P.O.; Adesodun, J.K.; Olowokere, F.A.; Adekunle, I.O.
Publication:Australian Journal of Soil Research
Article Type:Report
Geographic Code:6NIGR
Date:Dec 1, 2007
Words:7781
Previous Article:Nitrogen contributions in a windmill grass (Chloris truncata)--wheat (Triticum aestivum L.) system in south-western Australia.
Next Article:Using a legacy soil sample to develop a mid-IR spectral library.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters