Variability of Soil Properties at the Transition Zone Between Two Parent Materials in South Western Nigeria

MATERIALS AND METHODS

Description of Project Site
The project site is about 1,295 ha, located about 3 km northwest of Atuagbo town in the Esan Central local government of Edo state, Nigeria. It is about 10 km from Irua. It lies centrally within latitudes 6° 30’N and 7° 00’N and longitudes 6°00’E and 6° 30’E on the Topo sheet Ubiaja 285 SE of 1964. It is in the transition zone between rainforest and guinea savanna, generally referred to as forest-savanna mosaic. The mean monthly temperature ranges between 20.1 and 33°C. The project site is underlain by sedimentary deposition and Basement complex metamorphic rock. It has a humid climate with marked wet and dry seasons.

Soil Survey Method, Sampling and Analysis
A rigid grid method of soil survey was adopted at a scale of 1:10,000 to establish the mapping units (soil series). Auger observations were made at intervals of 200 m along each transects. Six mapping units were identified and demarcated on the base map. The area coverage of each of the mapping units was estimated (Iweke 75.85 ha, Mesan 34.42 ha, Iju 359.18 ha, Molo 315.50 ha, Agege 355.40 ha and Kulfo 155.65 ha). Surface soil (0-20 cm) samples were collected with the aid of soil auger from the mapping units. The number of sampling locations was proportionate to the areal coverage of each of the mapping units. The classification of the soils of the six mapping units is presented in Table 1.

The soil samples were air-dried, ground and sieved to separate the fine earth fraction (≤2 mm) from coarser fragments. Gravel content was determined gravimetrically as the percentage of particles with diameter >2 mm to the whole soil sample. Particle size analysis was done by the hydrometer method (Day, 1965). Soil pH was determined electrometrically in water (1:1 soil to water ratio). Organic carbon was analyzed by the dichromate wet oxidation method of Walkley-Black (1934). Total nitrogen was determined by Micro-kjedahl method. The exchangeable cations (Ca, Mg, Na and K) were extracted in 1 N neutral NH₄OAC. Na, K contents were determined by flame photometer while Ca and Mg were by atomic absorption spectrophotometer.

Table 1:	Soil classification

Statistical Analysis
The variability of the soil properties within and between the different mapping units was measured by estimating the mean (x), the standard deviation (SD) and coefficient of variation (CV). The variance ratio test was also calculated to test the between /within unit ratio of the mean square (i.e., F-ratio) for significance.

RESULTS AND DISCUSSION

The Table 1 shows the classification of the pedons and their local and international names. The local names were according to Moss (1957). Their appropriate taxonomic classifications were according to USDA soil taxonomy (Soil Survey Staff, 1994) and FAO/UNESCO (1988) soil map of the world revised legend. The areal covered of each of the mapping units were indicated.

Variation of Soil Properties Using Coefficient of Variation (CV)
The coefficient of variation of the properties for individual mapping unit as presented in Table 2 indicated that some soil properties are more variable than others. The mean sand content appears to be more homogenous with all the mapping units, having CV <15% (Wilding and Dress, 1978). The CV of sand ranges from 4.33-18.4%. High CV were recorded for silt (16.56-68.04%), clay (1.25-39.28%) and gravel (2.82-76.58%) content. With respect to gravel content, all the mapping units are extremely variable except for Iju and Kulfo soils, indicating that the soils are more heterogeneous in gravel content when mineral fractions are considered.

Chemical properties for different mapping units indicted that total N, available P, B organic C, K and Mg with CVs, range of 40.4-75.7, 52.8-88.9, 27.6-70.5, 33.8-75.2, 28.0-51.2 and 21.6-43.7%, respectively are more variable while Ca, Na, Exch. Acidity, ECEC, Base. Saturation, Zn with CVs range of 0.93-28.5, 15.8-39.8, 0.00-15.2, 14.9-26.8, 0.00-1.47 and 5.9-25.3%, respectively are less variable. These values obtained when grouped together by the method of Wilding and Drees (1978) puts Iju, Agege, Kulfo series as least variable while Iweke and Molo series are moderately variable while Mesan series is most variable.

Table 2:	Variability of soil properties among mapping units 0-20 cm depth
8E-05 = 5x10-8, 4E-04 = 4x10-4, 3E-05 = 5x10-5, 2E-04 = 4x10-2

Soils between the mapping units are more uniform in those properties which are genetically important such as textural properties, (% sand, silt, clay and gravel), pH, exch. acidity and percentage base saturation while those properties which are related to management e.g., (Ca, Mg, Na ECEC, organic C, Total N, Avail P, Mn, B and Zn) are more variable (Table 3).

The observed difference in the variability of physical properties can be attributed to rock weathering, biological action, micro relief and parent material from which the soils are formed. The high CVs for gravel in some mapping units e.g., (Iweke, Yampere, Molo and Agege) indicate that the soils are not uniform with respect to gravel content (Table 3).

The CV result obtained (Table 3) when all the mapping units are pulled together shows that some chemical properties are less variable than others, pH, Exch. Acid, B. saturation, Ca Mg, Na ECEC are less variable while exchangeable K, Organic C, Total N, Avail P, Mn, B and Zn appear to be more variable with CVs >35%. The CV values obtained in this study when grouped together by method of Wilding and Drees (1978) put Iju, Agege and Kulfo as least variable. Iweke and Molo are moderately variable while Mesan is most variable. Also the CV value obtained (Table 4) for soil properties put pH, Exch. Acidity, Base saturation as least variable having CV <15% while sand, clay, Ca, Mg, Na, ECEC. are moderately variable (CV 15-35%) and silt, gravel, K, organic C, Total N, Avail P, Mn, B, Zn are extremely variable having CV >35%.

The high level variation observed could be attributed to soil management practice employed in the area. At the time of survey, apart from some arable crops like cocoyam, cassava yam, that were grown, it was noted that oil palm trees were very common in the area. It has been noted that oil palm do have remarkable influences on soil chemical properties, Kang (1977) and Ogunkunle (1986). Zebarth et al. (2002) noted that human induced changes in soil can fundamentally alter the natural pattern of soil distribution in landscape even over relatively short time scales. Also Becket and Webster (1971) showed that in cultivated area, contrasting crops, soil amelioration and addition of fertilizer super impose differences between fields in addition to the variation already present in the native soils.

Table 3:	Estimate of variability between the mapping units
*p<0.05, **p<0.01, ns = Not significant

Table 4:	Grouping of soil properties using CV values
Adopted from Wilding and Drees (1978)

Variance Ratio Test (F-Ratio Test)
The result in Table 3 indicate that for soil properties considered, 50% of the mapping units are homogenous with reference to nine out of the eighteen properties studied showing significance difference between the mapping units. All the physical properties except gavel content significantly (p<0.01) affect the homogeneity of the soil (Table 3). The high variability within the mapping unit obtained for available P in this study was however not significant, despite the high CVs obtained. This can be explained from the view expressed by Becket and Webster (1971) that the variability in these properties which are largely affected by management introduced by differences in crops and management practice within a soil series could be as high as or higher than the variability introduced between series.

CONCLUSION

The high level of variability associated with these soils can be accounted for in term of their complex nature arising from their parent materials, landuse and management practices. The soils lie on the fringes of two-parent materials (Sedimentary and Basement complex). Mesan (highly variable), Iweke and Molo series all differed significantly in most of their chemical properties, Hence they should be managed differently. They cannot be representative enough or uniform for the whole land area for agricultural purpose.

In order to achieve optimum crop production, there is need to embark on the use of recommended cultural practices, rational use of fertilizers accompanied with optimum soil and crop management inputs that will take into cognizance, the variability of the soils. Since agronomic trials are measured by yield produced and such yields are highly dependent on soil fertility, the result will be affected by high soil variability as demonstrated in this study. Minimum tillage operation is recommended here, in order to effectively control water erosion and hence land degradation.