Evaluating the Impacts of Land Degradation on the Quality of Soils and their Variations between Different Clusters in Mosiro Irrigation Scheme, Narok County, Kenya

Aims: To study the impact of land degradation on soil quality Study design: Comparative analysis of soil quality indicators in year 2002 and 2013 in Mosiro Irrigation Scheme. Methodology: Changes in soil indicators determined in 2002 and 2013 were analyzed using ANOVA at 95% confidence level through Genstat Computer Software. The same indicators were applied to identify and characterize different clusters whose degree of variations were analysed using the same statistical method. Results: The results showed that all the soil quality attributes changed in the range of 21.4 and 79.1%. The greatest change was recorded in potassium which decreased by 79.1%, followed by phosphorous (60% decrease). The increase in sodium by 47% had a negative implication in terms of its increased potential to cause soil structural deterioration, while the increase of soil pH from 6.74 to 8.18 implied increased


INTRODUCTION
Delineation of soils into clusters which vary in productivity to an extent that they would respond differently to management inputs is a preamble to precision agriculture, aimed at improved soil quality and water use efficiency (Muya et al., 2011). Soil is a key natural resource and its quality is as a result of integrated effects of all soil forming factors and management (Sharma et al., 2004). As a result of human activities on land over time, the capacity of a given soil to produce goods and services is often decreased at much more rate than the formation of topsoil (Rijsberman, 1984). A cluster of soils or production systems are said to be biophysically sustainable if the compounded sufficiency of land quality attributes (productivity index) does not deteriorate over a realistic time horizon (Driessen, 1997). Indexing and monitoring soil productivity under crop and soil specific production system is the means to judge whether or not the system is sustainable in biophysical sense (Hurni, 1997). According to Muchangi et al. (2005), agricultural production from most irrigation schemes in Kenya has either reduced significantly over time or is far much below the level expected per unit of water supplied, and with advancement of agriculture under irrigation, soils are being degraded, thereby causing tremendous decline in their quality and productivity. Maintaining soil quality at a desirable level is a complex issue due to involvement of the interactive actions of the soil forming factors on the parent materials to form soil properties that determine its quality and productivity (Prasad and Biswas, 1999;Arshad and Gilley, 1999). The results of these interactions are the vertical, spatial and temporal variations in soil properties or soil quality indicators within and between clusters of soils under different landforms. The assessment and evaluation of soil quality indicators and their variations are required, not only to determine the type and quantity of agricultural inputs to apply, but also soil and water management strategies for sustained ecosystem services under irrigated agriculture. In this regards, there is an urgent need to adopt appropriate soil and plant management practices to reduce land degradation and maintain soil quality at a desirable level, based on the evaluation of homogeneity and heterogeneity of the delineated clusters of soils under different landforms or slopes (Smith and Elliot, 1990). Low irrigation and fertilizer use efficiency as well as declining quality and productivity of soil in most of the schemes are exacerbated by the blanket fertilizer recommendations that ignore natural and biophysical diversity of production systems (Muya et al., 2012). For example, fertilizer recommendations in Kenya by Kenya Agricultural Research Institute were based on trials without detailed characterization and clustering of production systems, based on relevant soil quality indicators (Sanginga and Woomer, 2009). Soil quality assessment, indexing and clustering of production systems under different landforms and slopes is a new paradigm in soil science research that is applied in vertical, spatial and temporal evaluation of land degradation, aggregation and management effects on soil quality and productivity. The degree of land aggregation that caters for soil structural formation and stabilization are influenced by the factors of soil formation including geology (parent material), climate, topography, time and human activities.
The degree of variations in the resulting soil properties depends on the scale. At farm level, the effects of climate and geology are generally insignificant, while the effects of topography and human activities on soils under different clusters over specified time becomes an important issue to examine at site specific scale. In this context, the objective of this research was to examine soil conditions within different topographical facets to evaluate the homogeneity of soils and the impacts of land degradation on their quality as a basis of formulating the cluster-specific interventions for improved nutrient and water use efficiency in Mosiro Irrigation Scheme. The homogeneity of different clusters of biophysical domains was assessed, based on the physical parameters and chemical characteristics which influence the availability of water and plant nutrients, while the impacts of land degradation on soil productivity was examined by indexing the dynamic and management dependant soil quality indicators.

MATERIALS AND METHODS
Mosiro Irrigation Scheme is situated South-West of Nairobi in Narok County, about 12 km South of Mosiro trading centre, and about 82 km from Ntulele Market Centre. The intersection of the grid line 36 o 04' E and 1 o 28' S marks approximately the centre of the scheme, at an elevation of approximately 1265 m above the sea level. The area lying to the North of the scheme is the main source of sediment, causing increased land degradation and siltation of irrigation channels (Muya et al., 2013). Therefore, water use efficiency and biophysical sustainability of the scheme depends on the extent to which the excess flows are checked and soil aggregate stabilization is maintained (Muya et al., 2013). The Physiography of the scheme comprises two major landforms, namely flat to gently undulating erosion plain and undulating old river alluvial plain. The soils of erosion plain are developed on volcanic tuff while those of river alluvial plain are derived from alluvium and volcanic ash mixtures respectively (Waruru et al., 2002). The study area was divided into five sampling areas, with slopes of 0-0.5%; 0.6-1.5%; 1.6-2.5%; 2.6-4.0% and 4.1-4.5%. Systematic soil investigation and mapping was done on transects across all the sampling areas from the upper erosion plain into the bottomlands of the river alluvial plain in the same area soil survey and mapping were carried in the year 2002. Clustering of the production systems was done, based on the visual assessment of observable soil parameters (soil colour, depth, texture, consistence and structure) across the five sampling areas. In each of the five sampling area, five representative soil profiles were identified for detailed description of vertical and spatial characteristics of soils. For the analysis of the impacts of land degradation on soil quality and productivity, composite soil samples were collected using a river auger at the depth of 0-20 cm around each of the representative soil profiles. The soil quality indicators (Table 1) selected for this purpose are those whose quantitative relationships with maize yield have been developed by Aune and Lal (1997).
For the assessment of homogeneity of different clusters, calcium (Ca), magnesium (Mg), bulk density (BD) and particle size distribution (texture) were used. They were selected because they form an important measure of soil productivity as is explained by the merits indicated in Table 2   Prior to laboratory analysis of the selected soil quality attributes, the soil samples were air dried and sieved through a 2-mm sieve. Soil pH was measured in 1:2.5 soil to water mixture, using the relevant electrodes according to Hinga et al. (1980). Organic carbon was oxidized with concentrated H 2 SO 4 and K 2 CrO 7 and determined calorimetrically (Anderson and Ingram, 1993). Total N was determined using the method provided by Okaleb et al. (2002). Cation exchange capacity (CEC) and exchangeable cations were extracted using 1N ammonium acetate at pH 7.0, followed by flame photometry for the determination Na, K, Mg and Ca, using flow analyzer (Okaleb, 2002). Soil texture was determined using hydrometer method (Hinga et al., 1980). Indexing of soil productivity was done, using semi-quantitative land evaluation methods (Driessen and Konijn, 1992). In this case, a range of numerical values of the selected soil quality indicators were rated and assigned fractions in percentage, being guided by the critical limits of the indicators. The critical limit of an indicator is defined as the numerical value of the soil property where crop yield is 80% of the maximum yield (Aune and Lal, 1997).

Calcium and magnesium
The levels of these elements and their ratios determine the availability of potassium (one of the macro-nutrients). Bulk density It is one of the most important soil structure attributes that controls the circulation of air, water and nutrients between the soil media and plants, hence a principle determinant of soil health. Particle size distribution Determines the retention and availability of soil moisture to plants.
Source: Amacher et al. (2007) Productivity index (PI) was determined using parametric methods of land suitability assessment provided by Driessen and Konijn (1992). This method involves: assigning ranges of numerical values and percentage fractions to each soil property selected as key soil quality indicators, ranking (Table 3) and combining all the single factor valuations in one mathematical equation that produces a numerical expression of the system performance or a relative index of performance (compounding) as follows: PI=(SQ1/100) X (SQ2/100) X (SQ3/100) X (SQn/100) Where: PI=Productivity index in % and SQ1, SQ2, SQ3, SQn are percentage ratings of soil quality indicator number 1, 2, and number n.
To assess the homogeneity of the identified clusters the variations in the selected soil quality attributes and PI within and between the clusters were evaluated by subjecting the data obtained from laboratory determinations (in the 2002 and 2012) to analysis of variance (ANOVA) at 95% confidence level where those soil quality indicators with significant levels were separated using Genstat Computer Software.

Clusters Identified and their Characteristics
Five clusters were identified and described as shown in Table 4.

The Impacts of land Degradation on Soil Quality and Productivity
All the soil quality attributes changed in the range of 21.4 and 79.1% between the year 2002 and 2012 (Fig. 1). According to Arshad and Martin (2002), changes in the soil attributes could be used as measure of the impacts of land degradation on soil quality. The increase in soil pH from 6.74 to 8.18 implied an increased tendency of soils to fix phosphorous and render it unavailable to plants (Silveria, 2012). At soil pH above 8.0, the increased alkalinity triggers the release of aluminum, manganese and molybdenum to a toxic level that impairs the root uptake of micro-nutrients such as copper, zinc and iron (Whiting and Reeder, 2009). Soil organic carbon, nitrogen, phosphorous and potassium decreased by over 20% which is much higher than the permissible threshold decrease of 5% in ten years (Amacher et al., 2007;Steer, 1998). The greatest percentage change was recorded in potassium which decreased by 79.1%, followed closely by phosphorous (60% decrease).
The increase in ESP by over 47% had a negative implication in terms of increased deterioration of soil structure by the increased accumulation of sodium (Waruru et al., 2002).

Ratings Remarks
Soil pH 4.8-5.5 or 5.6-6.8 or 4.8-5.5 or 6.9-7. The permissible environmental threshold is 6 while maize yield is 80% (Waruru et al., 2002).  Highly degraded, gently sloping structural plains Dominantly shallow sandy loam to loam soils, on relatively steep slopes towards the low-lying area with low water uptake and retention capacity,being extremely calcareous, saline, sodic and highly degraded with relatively low productive capacity. C5 On alluvial flood plains Extremely deep stratified sandy clay loam to loam.

Characteristics and Homogeneity of Different Clusters
The mean values of soil quality indicators for different clusters are presented in Table 5. The soil characteristics with the highest variation between the clusters is the clay (P<0.003) followed by CEC (P=0.002). Significant variations occur also in other characteristics, except calcium and sodium.  As is indicated in Fig. 2, the textural characteristics of soils, namely sand, silt and clay content vary significantly between different clusters, and this has an important implication in irrigation water management. Muya (1996) showed significant relationships between these attributes and available soil moisture holding capacity, which may be applied in designing the irrigation water scheduling for different clusters. For example, cluster five has the highest sand content, implying the need for lighter and more frequent irrigation, while cluster one has the highest clay content, meaning more and less frequent water application than any other clusters.

Variations in Soil Micro-nutrients and Their Management Implications
The variations of the micro-nutrients between different clusters were found to be significant for all the micro-nutrients except sodium ( Table 6). The variation of manganese (P<0.001), zinc (P=0.003), copper (P=0.008) and iron (P=0.031) between different clusters was found to be significant. This had an important implication on the prescription of the quantity of these micronutrients for different clusters during implementation of the envisaged management strategies. According to White and Zasoski (1999), precision agriculture, aimed at improved nutrient use efficiency should be based on the understanding of the variations in the levels of micro-nutrients as a basis of prescribing the soil and site specific types and quantity of micronutrients required as is normally done for macronutrients.

CONCLUSION
The study demonstrated the impacts of land degradation on soil quality in terms of the change in soil quality attributes between the year 2002 and 2013. The change in all the soil quality attributes examined was found to be in the range of 21.4 to 79.1%. The increase in soil pH from 6.74 to 8.18 implied an increased tendency of soil colloids to fix nutrients and render them unavailable to plants, which could explain the decreased availability of nutrients. The decrease in phosphorous, potassium, nitrogen and soil organic carbon by over 20% was found to be much higher than 5%, which is the value permissible (threshold) within 10 years. The five clusters identified were described as: C1: very deep clay with crusts that caused water stagnation on the surface; C2: shallow to moderately deep sandy clay loam to clay in places gravelly over murrum; C3: moderately structured and deep clay loam to clay with high carbonate concentration; C4: Shallow and highly degraded sandy loam to loam; and C5: extremely deep, stratified sandy clay loam to clay loam. The variations in soil characteristics between the five clusters were found to be significant for cation exchange capacity (P=0.002), magnesium (P=0.003) and bulk density (P=0.01). There was no significant difference for calcium (P=0.147).
All the textural characteristics of soil varied significantly with the highest being clay content (P=0.001), followed by silt (P=0.0018) and sand (P=0.008). This meant that the design of irrigation scheduling for different clusters should be based on their textural differences, which have an important bearing on their water uptake and retention capacities.
The variation of manganese (P<0.001), zinc (P=0.003), copper (P=0.008) and iron (P=0.031) was found to be significant, meaning that the quantities of the fertilizers applied to supply these nutrients should be different for different clusters as opposed to the current blanket recommendations that ignore these differences. This had an important implication on the prescription of the quantity of these micronutrients for different clusters during implementation of the envisaged management strategies.

ACKNOWLEDGEMENT
Funding from African Development Bank to Small Scale Horticultural Development Project and timely facilitation for the implementation of the project activities by the Project Coordination Unit in the Ministry of Agriculture, Livestock and Fisheries and Kenya Agricultural and Livestock Research Organization are highly appreciated.