|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 STATISTICAL AND GEOSTATISTICAL MODELING OF LUNNYU SOILS IN THE LAKE VICTORIA BASIN, UGANDA By BERNARD FUNGO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIR EMENTS FOR THE DEGREE OF MASTER OF SCINCE UNIVERSITY OF FLORIDA 2008
2 2008 Bernard Fungo
3 To my family, who have provided enduring support at all the times I needed it.
4 ACKNOWLEDGMENTS F inancial, material and m oral support were made possible from various individuals and institutions to whom an indebted with no reservations. Funding was made available through the Strengthening Agricultural and Environmental Capacity through Distance Education (SAEC DE) a Univer sity of Florida/CIAT joint USAID Africa pilot project. My nomination to this project was courtesy of the coordinator of this project at Makerere University who is and continues to be my academic Godfather. Only God knows how much his reward is. I owe spec ial thanks to my advisors at the University of Florida, who se hasty and meticulous review process es and guidance are what made my study at University of Florida both successful and enjoyable. The co advisor from TSBF CIAT was an obliging member of the supe rvisory team and his contributions made immense transcend of this thesis. Field assistants from Bisanje village in Masaka were valuable, especially for allowing me collect the soil sample from their garden. Colleagues in the Soil Science Department at Make rere University where the soil samples were analyzed did s great job My colleagues at the Faculty of Forestry and nature Conservation, Makerere University extend encouragement during times when the going was tough and their contribution needs mention. Las t but (by any standards) not least, I deprived my family of the time I should have had with them, but realising the noble work that I was undertaking, neither of them complained. To you all, may the Almighty God bless you abundantly.
5 TABLE OF CONTENTS p age ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 7 LIST OF FIGURES .............................................................................................................................. 8 ABSTRACT .......................................................................................................................................... 9 CHAPTER 1 GENERAL INTRODUCTION .................................................................................................. 11 1.1 Background ........................................................................................................................ 11 1.2 Problem Statement ............................................................................................................ 13 1.3 Objectives .......................................................................................................................... 13 1.3.1 General Objective ................................................................................................. 13 1.3.2 Specific Objectives ............................................................................................... 14 1.3.3 Hypotheses ............................................................................................................ 14 1.4 Significance of the Study .................................................................................................. 14 2 LITERATURE REVIEW ........................................................................................................... 15 2.1 Characteristics of Lunnyu Soils ........................................................................................ 15 2.2 Soil Variabil ity .................................................................................................................. 18 2.3 Soil Landscape Relationships .......................................................................................... 21 2.4 Methods of Studying Soil Variability .............................................................................. 26 3 DESCRIPTION OF STUDY AREA ......................................................................................... 29 4 FIELD -LEVEL VARIABILITY OF A LUNNYU -AFFECTED SOIL .................................... 33 4.1 Introduction ....................................................................................................................... 33 4.2 Materials and Methods ...................................................................................................... 33 4.3 Results ................................................................................................................................ 36 4.3.1 Statistical A nalysis ................................................................................................ 36 4.3.2 Geostatistical Analysis ......................................................................................... 37 4.4 Discussion .......................................................................................................................... 38 4. 5 Conclusions and Recommendations ................................................................................ 41 5 INFLUENCE OF SLOPE POSITION AND SOIL TYPE ON LUNNYU SOILS .................. 49 5.1 Introduction ....................................................................................................................... 49 5.2 Material and Methods ....................................................................................................... 50
6 5.3 Results ................................................................................................................................ 51 5.3.1 Statistical Analysis ................................................................................................ 51 5.3.2 Influence of Soil Type on Soil Properties ........................................................... 51 5.3.3 Influence of Slope Position on Soil Properties ................................................... 52 5.4 Discussion .......................................................................................................................... 52 5.4.1 Descriptive Statistics ............................................................................................ 52 5.4.2 Influence of Soil Type on Soil Properties ........................................................... 53 5.4.3 Influence of Slope Position on Soil Properties ................................................... 54 5.4.4 Influence of Soil Depth on Soil Properties ......................................................... 55 5.5 Conclusions and Recommendations ................................................................................ 56 6 GENERAL DISCUSSION AND CONCLUSION ................................................................... 63 LIST OF REFERENCES ................................................................................................................... 65 BIOGRAPHICAL SKETCH ............................................................................................................. 73
7 LIST OF TABLES Table page 4 1 Descriptive statistics of field -level lunnyu soil in the L. Victoria Basin, Uganda ............. 43 4 2 Spearman rank correlation of field level lunnyu soil properties ......................................... 44 4 3 Descriptive statistics of transformed lunnyu soil properties, Uganda ................................. 45 5 1 Characteristics of sampled landscapes .................................................................................. 57 5 2 Descriptive statistics of lunnyu soils in the Lake Victoria Basin, Uganda ......................... 59 5 3 Spearman rank correlation of field level soil properties of lunnya soils ............................ 59 5 4 Soil properties across slope positions in the Lake Victoria basin, Uganda ........................ 61 5 5 Subsoil soil properties across slope positions i n the Lake Victoria basin, Uganda ........... 62 5 6 Soil properties of selected lunnyu soils at two depths ......................................................... 62
8 LIST OF FIGURES Figure page 2 1 Cymbopogan ssp. (inside circle) a weed commonly associated with lunnyu .................... 28 2 2 Plants growing on lunnyu soil in Kabonera sub cou nty, Lake Victo ria Basin Uganda .... 28 3 1 Location of sampled areas in the Lake Victoria Basin of Uganda Field -level (red dot) and Landscape level (blue dots ) analysis ...................................................................... 31 3 2 Road -side profiles of a Plinthic lixisol in Ssenyange village, Masaka district .................. 31 3 3 Land use and soil maps of the study area ............................................................................ 32 4 1 Point map of sampling locations relative to lunnyu soil ...................................................... 42 4 2 Digital Elevation Model for the sampled lunnyu soil in Masaka, Uganda ......................... 42 4 3 Top soil properties ................................................................................................................. 47 4 4 Sub soil properties ................................................................................................................. 48 5 1 Idealized locations of sampling positions on the slope ....................................................... 58 5 2 Topsoil properties across ....................................................................................................... 60 5 3 Subsoil properties ................................................................................................................... 61
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science STATISTICAL AND GEOSTATISTICAL MODELING OF LUNNYU SOILS IN THE LAKE VI CTORIA BASIN, UGANDA By Bernard Fungo December 2008 Chair: Sabine Grunwald Major: Soil and Water Science Our study aimed at characterizing the distribution of lunnyu soils in the Lake Victiria Basin of Uganda at field and landscape level. At field -le vel, s oil samples were collected from the center point square grids obtained by laying a 180 m x 180 m plot. At spacing of 20 m x 20 m, 81 sampling points were georeferenced and samples taken at two depths (0 -20 cm and 20 40 cm). And an additional 19 rando m locations within the plot were taken to make a total of 100 sampled locations. The soil properties analysed were texture, pH, available P, and exchangeable bases V ariograms were used to describe the spatial structure of the soil properties. All the soil properties, except silt, showed spatial dependence at both depths at the scale of study. Phosphorus Ca, Na and sand showed shorter ranges of between 42 and 58 meters, all in the top soil, but the other properties in top and subsoil have larger ranges of 149 meters. At landscape level, six existing lunnyu patches located on four different soil types were identified and a t each, the slope was divided into three parts; shoulder, back -slope and foot -slope. Five locations at each landscape position, from where soil samples were taken at two depths (020 cm and 2040 cm), were separated by a distance of 20 30 m across the slope. Lunnyu patches on Chromic lixisol and Mollic gleysols had higher pH, P, sand, clay and silt compared to those on Plinthic ferralsols an d
10 Petrifferic lixisol. Neither of the soil properties was influenced by landscape position. Soil pH, Ca, Mg, and K were higher in topsoil compared to subsoil. Neither slope position nor the type of lunnyu has showed consistent differences in all the soil p roperties. Thu s lunnyu phenomenon cannot be explained by these two factors. Overall, t he spatial distribution of the soil properties in the sampled lunnyu patches are not straightforward. In order to resolve the lunnyu problem, objective identification of lunnyu patches should be sought, followed by careful monitoring of crop performance under different soil fertility management interventions.
11 CHAPTER 1 GENERAL INTRODUCTION 1.1 Background About two thirds of people in Africa depend on agriculture for the ir livelihoods and most economies are agriculturally based (FAO, 2004). Most farmers in this region are smallholders with land size ranging between 0.5 and 2 ha, earn less than US$1 a day, face 3 5 hunger months, have large families and are malnourished (B ocar, 2003). Consequently, as agricultural productivity declines, peoples livelihoods in this part of the world are in jeopardy. Agricultural intensification on marginal and poor quality lands often using poor farming methods, results from population grow th, may soon reach the limits. According to Brady and Weil (2002), only 12% of African soils are moderately fertile compared to 33% in Asia. To further compound the ecological fragility of these regions, there is highly variable and declining rainfall patt erns observed since the 1970s (Brown, 2004). More specifically these areas are characterized by high intensity rainfall coming in short durations and variably distributed. Soil degradation is largely responsible for continuous reduction in agricultural out put that eventually leads to food insecurity (Scherr, 1999; Hartemink, 2003; Brown, 2004). Soil degradation results from such known factors as deforestation, population growth, poor land use, insecure land tenure, inappropriate land management practices an d poverty (Bocar, 2003; FAO, 2005), acting either in isolation or in permutations. Although much information is available about the causes, processes and effects of soil degradation, corrective measures are needed because the number and scale of these fact ors is highly variable. In particular, regional and site -specific recommendations are needed to manage land resources in sustainable manner reducing soil degradation while providing resources to produce food. Along with the failures in adoption of improved farming methods and technology, is the increase in various forms of land degradation.
12 Land degradation is widespread and believed to be increasing (Sanchez et al., 2005) Quantification of soil spatial variability across soil -landscapes is important in eco logical modelling, environmental prediction, precision agriculture, and natural resources management (Grunwald, 2006). Spatially averaged soil erosion data provide some information on soil erosion dynamics but temporal variation and, spatially distributed erosion data are needed to better understand erosion processes and thoroughly evaluate process -based erosion prediction models. Reduced soil productivity resulting from soil erosion on slopes has been observed in several parts of the world. This relations hip, however, has not received much attention but can be potentially useful in highlighting the spatial distribution of relatively infertile soil patches (Van Oost et al., 2000; Li and Lindstrom, 2001; Timothy et al., 2003). These have been observed to occ ur mostly at the shoulder position of sloping land strongly suggesting that erosional processes have impacted the formation of these soils. In Uganda there is a local land degradation problem ( Lunnyu) that still defies understanding (Tenywa, 2004). Lunnyu locally means salt like. It manifests in isolated unproductive patches of land in otherwise productive land (Chenery, 1954; Ssali, 1975; Zake 1986). Reports from farmers suggest that it is widespread and believed to be increasing. Originally, the lunnyu p henomenon was thought to be related to chemical degradation but recent studies (Tenywa, 2004) suggest that soils are characteristically around the rim/shoulder position of the soil -landscapes. This study builds from earlier studies conducted on lunnyu desc ribing chemical characteristics (Zake, 1986), erosion (Lufafa, 2003), physical characteristics and mapping (Tenywa, 2004) and nutrient status (Taulya, 2004). Their nature and evolut ion is still
13 not clear and this study is intended to provide additional information to enable better understanding of the lynnu phenomenon. 1.2 Problem Statement From the observations of Lufafa et al. (2003) that soil erosion was most severe at the backslope position of the landscape, followed by the fact that lunnyu patches are located characteristically around the rim/shoulder position, it seems likely that the lunnyu phenomenon, is an erosional phase and that the soil properties are most influenced by this process. The catena concept posits that: (1) topography controls landscape hydrology; and (2) landscape hydrology controls soil formation; and (3) soil properties are determined by topography, hydrology and in situ soil formation. Brown et al (2004) suggested that soil prediction can be derived from mapping of factors suc h as topography. It is hypothesized that lunnyu soils are an erosional phase that occurs at landscape positions with greater intensity of hydrological processes (e.g. runoff, water erosion) and are dependent on soil type and land use. Most previous studie s have employed classical statistics that requires validity of some basic assumptions, such as the independence between observations, due to randomness of spatial variations. In contrast, geostatistics, based on the theory of regionalized variables, enable s interpretation of results based on the structure of their natural variability, taking into consideration spatial dependence within the sample space. 1.3 Objectives 1.3.1 General O bjective To develop a quantitative soil landscape model that predicts lunny u soils in dependence of landscape factors (geology, topography and land use).
14 1.3.2 Specific O bjectives Objective 1: To characterize base properties of lunnyu soils. Objective 2: Describe the spatial distribution and variability of lunnyu soils at fiel d and landscape level within the Lake Victoria Basin of Uganda. Objective 3: Investigate relationships between soil -forming factors (landscape position, parent material and land use) and lunnyu soils to understand their formation and distribution. 1.3.3 Hypotheses H1: Characteristics of luunyu soils occur mostly at the rim/shoulder, where erosion is most severe, indicating that they are an erosional phase H2: The occurrence of lunnyu soils is dependent on the geology and toposequence H3: Lunnyu soils o ccur on pre -cambrian Buganda surface and remnants derived from granitic -gneiss of the early proterozoic era that have been subjected to continuous cultivation 1.4 Significance o f the Study To be able to implement management practices, community development agents such as extension workers, land use planners regulatory agencies and researchers need to understand the spatial patterns of soils and how they are bound to change with time. In this study, the model developed will make a possible explanation to inf orm these agents on the likely locations of lunnyu patches and how they expand with time. Management interventions and land management policies could then be designed to curb their expansion. In a broader sense, this background information makes it possibl e to attempt managing soil processes such as erosion that result in the formation of infertile soil patches. Since landholdings are small in Uganda they have to be optimized using precision agricultural techniques optimizing yield, while minimizing degrada tion of soils and preserving nutrient status.
15 CHAPTER 2 LITERATURE REVIEW 2.1 Characteristics o f Lunnyu Soils The phenomenon of lunnyu soils in Uganda was first documented in 1954 (Chenery, 1954). Their recognition was based on three major features: poor crop vigour, loss of soil consistency and poor quality crop and low crop yields. Another feature common to lunnyu is occurrence of Cymbopogan ssp. (Figure 21), a weed thought to negatively influence native plant communities by forming dense monotypic sta nds that alter ecosystem properties and lower local species diversity. A hypothesized mechanism by which Cymbopogan ssp. achieves competitive dominance is an innovative use of its below -ground vegetative disturbance of neighboring vegetation (i.e. punctur ing) (Lippincott, 2000). In other words, as the roots of Cymbopogan ssp. find their way through the soil, they pierce and destroy the root systems of neighbouring vegetation and thus survive at the expense of the latter. The general explanation given for t he formation of monotypic Cymbopogan ssp. stands is that it is a superior competitor against native vegetation. However, Holly and Ervin (2006) indicated that rhizome -mediated, below -ground vegetative penetration is a much larger intraspecific phenomenon t han an interspecific one. The data also strongly suggested spatial location as a significant factor, with most penetrations occurring in the interior of a Cymbopogan ssp. stand as opposed to the advancing border. Infertile soil units within a landscape can invariably be delineated based on indigenous knowledge of farmers. Recent interviews with farmers (Tenywa, 2004) revealed that these soils emerge suddenly and do not expand thereafter. The patches can support some crops (e.g, maize, sorghum, ground nuts, and some species of bananas. It is not possible to recognise lunnyu at first site in the field. In most cases, a crop such as beans is planted and as it grows, location of lunnyu
16 is recognised and delineated on the basis of poor performance of the crop. The most common symptoms of crop failure are stunted growth and chlorosis (yellowing of leaves). Figure 2 2 shows some of the plants growing on site said to be under lunnyu in Kabonera sub county, Lake Victoria Basin of Uganda. Farmers perceptions of soil fertility versus laboratory tests (Micheni and Irungu, 2003) showed that the farmers knowledge and information on soil quality and fertility corresponds well to the laboratory assessments of soil phosphorus (P) and organic carbon (C) status, with differen ces between fields within a farm and/or between different farms within a given location. Therefore, it seems possible to delineate infertile soil units within a landscape based on indigenous knowledge of farmers. The studies further revealed that the caus es of lunnyu are multi -factorial. The major ones that explain the occurrence of lunnyu soils that have been suggested include the nature of the parent material from which the soil is formed, overuse and erosion, and farming systems (the detrimental ones be ing sugarcane and Eucalyptus land use). It is also thought that there may be different forms of lunnyu soils, although only few farmers could distinguish these types. They suggest that one type is related to inherent soil properties, while the other one is due to the management and cropping system. Chenery (1954) noted that lunnyu soils occurred as patches of unproductive land in the midst of fairly fertile land and that they were most common in areas receiving over 1,250 mm of rainfall. Although different parts of the country were reported to have lunnyu, the worst affected areas were Ssese Islands, Lake Victoria Basin, West Kigezi, East Ankole and Busoga. Surprisingly, these were, and still constitute the most productive parts of Uganda. The farmers in th e recently surveyed area (Tenywa, 2004) believe that there is a pattern of distribution of lunnyu
17 soils and that they occur along strips just below summits and above footslopes and that there is a relationship between the landform patterns and type of lunnyu Farmers have tried to manage the problem of lunnyu soils mainly through use of manure, improved fallowing or combination of these. Information for the management interventions by farmers is obtained from several sources including fellow farmers, self -experimentation, and from agricultural extension workers. Despite these interventions, crop response remains low. Consequently farmers have abandoned the management and left these degraded soil patches to nature. Instead, they use the areas for activities such as brick -making and grazing, although some still maintain them as poorly cropped areas. Recent preliminary studies (Tenywa, 2004) showed that lunnyu patches occur mainly on the summits and back slopes of ridges. They vary in size from 5 100 m2. Consi dering the size of land that farmers hold (0.5 2 hectares), lunnyu soils substantially reduce agricultural potential of many households. Examination of the profile characteristics (Tenywa, 2004) showed that lunnyu patches occurring on backslopes of ridge s were generally deeper (>150 cm) compared to those of the summit (about 30 cm). The sub -soil structure is also very weak compared to that of adjacent productive areas. The infiltration and transmissivity rates of lunnyu patches were, however, found to be much higher (1.3 cm/min. and 1.5 cm/min., respectively) than that of adjacent fertile soils, which were 0.7 cm/min and 0.6 cm/min, respectively (Tenywa, 2004). Earlier studies (Lufafa, 2000) found that other soils located at the summits and backslopes in t he area, had an average infiltration rate of 0.35 cm/min. Meanwhile, productive land had higher sorptivity compared to the lunnyu patches. Both lunnyu and productive areas showed significant variations in earthworm populations from one part to another (Tenywa, 2004). These soil organisms were observed to be generally low in lunnyu patches and were independent of the type of lunnyu. Soil organisms have profound
18 roles they play in improving soil health and in nutrient transformation. As they burrow through th e soil, they create tunnel that ramify the soil system, providing continuous macropores. These facilitate water and air movement in the soil system. Macrophages are the first actors on organic matter before micro organisms break in further to release soil nutrients. (2006) showed that the activities of soil organisms vary spatially and are controlled by factors including the availability of food, moisture, temperature, predation and competition, and chemical characteristics ( Aislabie et al., 2006; Aranda and Oyonarte, 2006) Productive patches had relatively higher values of pH (6.9) when compared to unproductive ones (5.9). The same patterns was found for exchangeable P (10.8 mg/kg productive patches vs 5.6 mg/kg unproductive patches), while ni trogen (mean = 2.47 g/kg) and organic matter (mean = 16.4 g/kg) were not significantly different. Values of the chemical properties (OM, pH and available P) reduced with depth (Zake, 1986). 2.2 Soil Variability Soil landscape variability has confounded the understanding of many soil scientists owing to the large number of variables involved. Deterministic uncertainty is a perspective on soil spatial variability that reconciles the traditional reductionism view (variability can be explained with more and bet ter measurements) (see Phillips et al. 1996) and the emerging nonlinear dynamics view (variability may be an irresolvable outcome of complex system dynamics) (Reddy et al., 1996). In a natural landscape, soil represents a wide variety of attributes spati ally distributed within landscapes as a result of the interaction of the processes that rule soil formation (Junior et al., 2006). Jenny (1941) stressed that when attempting to understand geographic variation of soils, it is often useful to analyze each si te in terms of the five factors of soil formation ( climate, parent material, organisms, topography, and time ). The ultimate nature of the soil is then affected by the way they are influenced by human actives.
19 Depending on how the factors of soil formation interact, spatial soil variability can be recognized at three levels; small (meters to a few hectares), medium (few ha to few km2 landscape level) and large -scale (large geographic regions) (Vanlauwe et al., 2002). Small -scale variations are most ofte n due to small changes in topography and thickness of parent material layers or to the effects of organisms such as the effect of individual trees or past human management (Brady and Weil, 2002). Many attempts by earlier studies to describe variability hav e employed land surveys, maps and databases that provided relatively coarse -scale spatial information. For example, Chittleborough (1978) used profile data to assess soil variability within land systems and land units While there was a statistically signi ficant reduction in variability within the land units compared with that for the whole area, the amount of reduction was low. The variability of individual properties, as indicated by coefficients of variation, varied widely within land systems and land u nits. Least variable were B horizon texture, thickness and structure, and A horizon texture. Most variability was assessed based on visible landscape features such as land use, vegetation type and the challenge was to identify clear boundaries. The princip le that underpins this approach is that many hydrological, and hence soil processes, are related to topographic attributes such as elevation, slope, aspect, and plan and profile curvature (Gallant and Wilson, 1996; Wysocki et al ., 2000). Many studies have shown (Singh and Kar, 2001; King et al., 1999; Landi et al., 2004) that even subtle differences in topography can have a large effect on the spatial distribution of soil properties. This variability in soil fertility often reflects past soil -management pra ctices as well as differences in soil profile characteristics. The processes of nutrient depletion and soil degradation are spatially heterogeneous, as determined by the underlying parent material and geomorphology and by (current and historical) management (Smaling et al ., 1997). The variability between
20 different farm types within a village is associated within the soilscape, such as the location along catenary sequences (Deckers, 2002), and with differences in soil fertility management between poor and wealthy households (Crowley and Carter, 2000). Resource availability and the pattern of resource allocation to different activities are determined by household wealth, and depend on household priorities and production strategies. Therefore, the intensity of the processes, potentially leading to variation in soil fertility status at the farm level and their dynamics will vary between farms of different resource endowment and production orientation (Tittonell et al. 2005a). A study of diversity in soil fer tility management of smallholder farms in western Kenya (Tittonell et al ., 2005b) showed that farmers manage their fields according to their perceived land quality, varying the timing and intensity of management practices along soil fertility gradients. Fi elds classified by them as poor were planted later (up to 33.6 days of delay), with sparser crops (ca. 30% less plants m2) and had higher weed infestation levels than those classified as fertile, leading to important differences in maize yield (e.g. 0.9 ve rsus 2.4 t ha1). Furthermore, the internal heterogeneity in resource allocation varied also between farms of different social classes, according to their objectives and factor constraints. Additionally, the interaction of sub-location -specific socio -economic (population, markets) and biophysical factors (soilscape variability) determined the patterns of resource allocation to different activities. Such interactions need to be considered for the characterisation of farming system to facilitate targeting res earch and development interventions to address the problem of poor soil fertility. Presumably, where small -scale variability exists, the actual level of fertility at most spots in the field are likely to be either considerably higher or lower than the aver age soil test value for the field. Such variability may be difficult to measure and not readily apparent to an ordinary
21 observer and can, in most cases be detected only by analyzing soil samples taken from many spatially distributed borings throughout the plot area. The growth in the precision agricultural industry and developments in high-density mapping and geostatitical techniques to produce fine scale soil property maps have made it possible to recognize small -scale variations in soils (Grunwald and Lam sal, 2006). Landscape -level soil variability is related primarily to differences in a particular soil forming factor such as soil topography (drainage), climate, or parent material. Thus, relatively identical soil units can be classified considering the influence of these factors and even predict soil properties in the landscape positions occupied by similar soil units. This is the basis for lithosequences (occurring across a sequence of parent materials), chronosequences (occurring across similar parent material of different age), and toposequences (with soils arranged according to changes in relief) (Brady and Weil, 2002). 2.3 Soil -Landscape Relationships Attempts to characterize and predict soil attributes in the landscape have employed such concepts as t he factor (such as climate, vegetation, parent material and topography) and process (hydration, oxidation, solution, leaching, precipitation and mixing) models. Olson (2006) emphasized the operation of processes in combination, some having a positive and s ome a negative influence in horizon differentiation. These processes are largely dependent on their hilslope positions (Lufafa, 2000; Bamutaze, 2005). Catenary sequencing affects soil forming processes such as erosion, sedimentation, infiltration and drai nage (Webb and Burgham, 1994). Webb and Burgham (1997) collected morphological data from transects of two soil series into generalised soil landscape models. In both soil series, the thickness of topsoils, depth to reducing conditions, and depth to fragipa ns were greatest on footslopes and generally decreased to shoulder slopes. Penetration resistance
22 was lowest on footslopes and increased to shoulder slopes (Webb and Burgham, 1997). In cultivated land, footslope sites have markedly over -thickened topsoils. Relocation of topsoil material from upper to lower slopes is attributed mainly to the effects of cultivation, either directly, through mechanical movement of soil material during cultivation operations, or indirectly, through the promotion of soil erosion. However, there was no indication of the expected catenary relationship involving translocation of exchangeable bases from upper to lower slopes as no differences were observed in these aspects. These results corroborated strongly with the findings of Bru nner et al. (2004) and Mulumba (2004) in Uganda. It is conceivable that the direction of water movement, as determined by slope curvature, may have profound effects on profile differentiation. As a result, landform attributes have an effect on the process es and therefore the resultant soil properties. The catena concept has been a useful tool to predict the occurrence of soils in many landscapes as it contributes substantially to the input of active factors of soil formations (solar radiation and water). B runner et al. (2004) reported that soils at the summit position had a thick solum due to the stable soil formation on the flat surface, whereas soils at the shoulder position had shallow A -horizons due to active erosion processes. Valley and footslope soil s showed hydromorphic features and accumulation of soil material from upslope. Simulations considering a catenary soil sequence in the same site showed a clear spatial demarcation between erosion and sedimentation zones, bringing both findings to agreement Brown et al. (2004) observed that for the yellow gray soils along a catenary sequence, sand content and sand grain size increased with greater slope gradient and in converging areas evidence supporting fluvial control of surface soil composition. For red soils, texture contrast increased on lower backslopes, decreased over ferricrete rich parent material, and had no
23 significant association with either infiltration or runoff -influenced locations. Surprisingly, texture contrast was also reduced or even i nverted on fine scale convexities (<18 m in diameter). These findings were consistent with the following theoretical processes: (1) sand deposition on lower backslopes, increasing texture contrast; (2) ferricrete weathering at the soil surface, reducing te xture contrast; and (3) texture contrast inversion through termite turbation. Thus, slope steepness and curvature are potential terrain (and other) controls on surface texture and explanatory variables for texture contrast formation for well -drained red so ils, an d fluvial deposition for yellow -gray soils in the case of Uganda. Thompson et al (2006) reported terrain attribute distributions to have differed significantly among six sites, with regional terrain attributes (upslope contributing area, topographi c wetness index) being more similar among fields than local terrain attributes (slope gradient, slope curvature). Thompson et al (2006) predictive models explained from 28% to 67% of the variation in soil properties. The terrain attributes that best predi cted soil variability were similar across all three fields used for model development, with slope gradient, elevation, slope curvature, and upslope contributing area appearing in most of the models. However, applying models from one field to other fields w ithin the same physiographic region produced inconsistent results. This lends support to Officer et al. (2006) who found a logical relationship between the clay mineralogy and the position of two New Zealand steepland top soils in the landscape, as an indi cator of the extent of weathering. Analysis of five slope elements (shoulder, toeslope, backslope, footslope and summit) using remotely sensed satellite data showed a consistently wide range of variability within the different landscape positions (Sawhne y et al., 1996). The results further showed that the soils developed on toeslopes had thicker horizons, whereas shoulder slopes showed thinner horizons
24 with minimum solum thickness. Differences among slope positions were observed highest for clay and silt, OM and calcium (Ca) carbonate, but there was no significant difference in pH and Cation Exchange Capacity (CEC). Tenywa (1993) reported higher mean soil losses from the upper slope than on the middle and lower slopes. Therefore, as the soil is transported by surface runoff from the upper to the lower slopes, the latter ones build a thicker horizon at the expense of the upper slope horizon thickness. The combined effects of soil formation factors and soil processes should not be considered in isolation beca use of the close relationship between them. Kaspar et al. (2004) tested if a data set containing 20 soil and terrain variables could explain spatial yield variability better than a subset of seven more easily measured variables. It was demonstrated that th e larger dataset (20 variables) explained more of the spatial variation in yield than the subset of seven variables. To quantify the relationships between cotton lint yields and derived topographical attributes in combination with measured soil physical pr operties, Iqbal et al. (2005) indicated that cotton lint yield variability was explained by 65% soil properties and 40 and 21% by topographic and hydrologic attributes, respectively. Their combined effects explained 82% of yield variability. Elevation, flo w direction, sediment transport index, percentage sand content, and volumetric water content explained most of the lint yield variation. In Iqbal et al., (2005), the elevation range was 6373 m. It is important to note that in areas with large elevation change, topography explains much of the variability in soil properties. The opposite is true for landscapes with less relief (e.g. Florida, Grunwald et al., 2001; Grunwald et al., 2004; Rivero et al., 2007). Kravchenko and Bullock (2000) observed that the c umulative effect of the topographical features explained about 20% of the yield variability (6 54%) in corn and soybean grain. Elevation, which ranged from 2 m to 60 m, had the most influence on yield, with higher yields
25 consistently observed at lower landscape positions. Curvature, slope, and flow accumulation significantly affected yield only in certain conditions, such as extreme topographical locations (poorly drained depressions or eroded hilltops) combined with very high or low precipitation. Jiang an d Thelen (2004) also noted that t he combined effect of both soil and topography varied by year and explained 28 to 85% of the observed yield variability in corn and soybean. Analysis of a 20 variable data set (Kaspar et al., 2004) showed that soybean yiel d was affected more by closed depressions in wet years, and less by curvature in dry years. Similarly, yield was negatively affected by closed depressions and lower landscape positions in wet years, whereas these factors had either no effect or a positive effect in dry years. Alternately, curvature had a negative effect in dry years and no effect in wet years. The modifying effect of topography on parent material is associated with a variety of vegetation types and plant communities as determined by the res ultant soil quality. Thus, the effects of different plant species on soil properties also need to be highlighted. Wright and Hons (2005) reported crop species to have had significant impacts on soil organic C (SOC) and soil organic N (SON) sequestration. O n average, the authors reported a wheat monoculture to have had greater SOC (9.23 Mg C ha1) at the 0 to 5 -cm depth than sorghum (6.75 Mg C ha1) and soybean (7.05 Mg C ha1). No -tillage (NT) increased the proportion of >2 -mm and 250m to 2 -mm macro aggr egate fractions in soil compared with conservation tillage (CT). At the 0 to 5 cm depth, NT increased SOC compared with CT by 158% in macroaggregate fractions, but only 40% in <250m fractions. No tillage increased SON compared with CT by 300%, 94%, 41% and 39% for >2-mm, 250 m to 2 mm, 53 to 250m, and <53 m fractions, respectively. Longterm impacts of NT included a greater proportion of macro aggregates and increased C and N sequestration, but impacts were dependent on crop species and varied wit h soil depth.
26 Whitford and Kay (1999) also reported that biopedoturbations could cause substantial variability in the landscape. Terra et al, (2005) noted that soil nutrient responses to landscape positions were variable depending on land use type and the location of land-use types relative to the landscape. The highest levels of soil OM, total N and available N were observed at middle slope position under conservation systems (combined no surface tillage within row sub soiling and winter cover crops). At f ootslope position, similar levels of these soil properties were observed but under a under a conventional tillage systems (consisted of chisel ploughing/disking + in row sub soiling without cover crops) (Terra et al., 2005). However, an increasing trend fr om upper slope to footslope for five nutrients was found under the latter. 2.4 Methods o f Studying Soil Variability To study soil variability, two main quantitative approaches can be used, different in the way data is analysed. Classical statistics (Sokal and Rohlf, 1995) requires the validity of some basic hypotheses, such as the independence between observations, due to randomness of variations from one place to another. In contrast, geostatistics, based on the theory of regionalized variables, enables t he interpretation of results based on the structure of their natural spatial variability, taking into consideration spatial dependence within the sample space (Webster and Oliver, 2006). The analysis of dependence is based on the structure of the semivario gram, which demonstrates the existence of spatial dependence. Geostatistics is being increasingly used in the assessment of spatial variability in natural resource management (Verosene et al. 2006). Soil mapping generally requires (i) a predefined model of soil formation and (ii) data on soil properties and other variables that have significant effect on soil formation and thus on the spatial distribution of soil properties (McBratney et al., 2000; Grunwald, 2006; Rivero et al ., 2007). Both approaches nee d input data on soil and covariates characterizing the environment where soil formation takes place. Profile information is also needed to train the models and to
27 understand the soil resources of the area. The major differences, the strengths and also the limitations are derived from the way the environmental covariates are represented in the procedure (Grunwald et al. 2004). Digital soil mapping requires digital data sources as input variables for the quantitative models. The difference is the way the mo del derives the soil information from the input data, store, manage, analyze and display it (Grunwald and Lamsal, 2006). For example, Bishop and Minasny (2006) observed that the usefulness of digital terrain attributes for soil mapping is largely dependent on the landscape and the quality (expressed in terms of spatial resolution or uncertainty) of the model. The highresolution power of digital elevation models thus makes them superior to traditional topographic maps. Pennock and Corre (2001) described l a ndform segmentation, a branch of Digital Terrain Analysis (DTA) that groups similar topographic attributes into discrete spatial units, which can then be utilized as treatments for spatial analysis. Martin and Timmer (2006) used this method and their resul ts showed that landform segmentation was a useful tool in identifying spatial differences for topsoil and solum depth. Soil moisture content displayed clear spatial patterns at the scale of Landform Element Complexes, suggesting that differences in topogra phic characteristics largely influence these parameters. Landform Element Complexes were not as useful in explaining spatial patterns of topsoil organic matter probably due to the influence of downed woody debris and litter accumulation at localized points Landform segmentation also captured spatial trends for litter layer depth.
28 Figure 2 1 Cymbopogan ssp. (inside circle) a weed commonly associated with lunnyu A B Figure 2 2 Plants growing on lunnyu soil in Kabonera sub county, Lake Victoria Basin Uganda (A) Probable delineation of lunnyu (white dotted line) in a bean garden and (B) Bananas performing poorly on a site said to be lunnyu affected
29 CHAPTER 3 DESCRIPTION OF STUDY AREA The study was conducted in the microcatchment covering the districts of Masaka, Rakai and Sembabule. The districts are located on the western side of the Lake Victoria basin. The county lies approximately 310 40 E and 350 S (Figure 3 1) with a spatial covera ge of 126 km2. In this zone, agriculture is rain -fed, with average annual precipitation of 1,218 mm and slightly drier periods in June and July and December -February (Komutunga and Musiitwa, 2001). The average annual temperature is 21.50 C, with little se asonal variation. The altitude ranges from 1,200 1,260 m. a.m.s.l. It is located within the predominantly banana coffee farming system but other land use types present include annuals, banana and p asture/rangelands (KisambaMugerwa, 2001). The slope of the area ranges between 30 and 180 with most of the area falling in above 150 and thus has a relatively high erosion hazard. It is classified under the south central moist hills and valleys land resource areas. This zone is located on the Eastern African plateaus (1,150 1,400 m. a.m.s.l.) between the western and Eastern African rift on an extremely old (mid to end tertiary) Buganda surface characterized by hills and ridges that are highly dissected (dissected plateau) by streams and drainage ways (Hadoto, 2001). The solid geology of the area is undifferentiated acid and hornblende gneisses of the basement complex and the parent material is pre -weathered gneiss (Aniku, 2001). Geologically, t he area belongs to the Buganda surface, which covers the southern part of Central Uganda and consists of granites, gneisses and schists of the Precambrian age (Harrop, 1970). The Buganda surface is part of the Ugandan basement complex and a product of long term weathering processes. Brunner et al. (2004) reported that so ils at the summit positions had a thick solum due to the stable soil formation on the flat surface, whereas soils at the shoulder position had shallow A -horizons due to active erosion processes. Valley and footslope soils
30 showed hydromorphic features and a ccumulation of soil material from upslope. The soils on uplands are predominantly Plinthic ferralsols, and Plinthic and Chromic lixisol (WRB, 2006), and are developed from Precambrian schists and quartzites (Figure. 3 2). They are fine textured and have an isohyperthermic temperature regime and udic moisture regime. The soils in the lowlands and valleys (drainage ways between the ridges and hills) are Mollic Gleysols which occur in swampy and papyrus mashes and are seasonally or permanently water -logged. T he native vegetation is woodland with papyrus but this has been greatly modified by human activities (Aluma, 2001). The catenary sequence consists of shallow brown loam soils on broad crests with deep residual soil on the side slopes. Information about the land use history, soil degradation problems, soil types, evolution and management practices of lunnyu soils were collected from previous studies carried out in the area such as Lufafa (2000), Achan (2001), Taulya (2004) and Mulumba (2004). In the area, there are three broad land use types; pasture/rangelands, perennials (mainly banana -coffee) and annuals. Although lunnyu soils occur in all of them, they are more commonly found in annual and perennial cropping systems. At a scale of 1:250,000 four soil mapp ing units were identified (Figure 3.3) and these include; Chromic lixisol (CL), Plinthic lixisol (PL), Mollic Gleysol (PG) and Plinthic Ferralsol (PF).
31 Figure 3 1 Location of sampled areas in the Lake Victoria Basin of Ugan da Field level (red dot) and Landscape level (blue dots ) analysis Figure 3 2 Road -side profiles of a Plinthic lixisol in Ssenyange village, Masaka district
32 A B Figure 3 3 Land use and soil maps of the study area; A) Land uses. B) Soil map units Blue circles represent relative location of sampling points and black dots represent the previously sampled lunnyu patches. Layer Data from Soil Erosion Project, Makerere University
33 CHAPTER 4 F IELD -LEVEL VARIABILITY OF A LUNNYU -AFFECTED SOIL 4.1 Introduction Heterogeneity is an inherent quality of soil that typifies its anisotropy and in a natural landscape. It represents a wide variety of spatially varied soil attributes, and as a result of the interaction of the processes that rule soil formation (Junior et. al ., 2006). The lunnyu phenomenon has been described by local farmers in Masaka -Uganda as being a state of infertility occurring in patches ranging between 50 100 m2. Unlike the basic pri nciples of experimentation established by the classical statistical method that considers soil variability to occur entirely at random (Santos and Vasconcellos, 1987), the fact that soil attributes show strong spatial dependence (Journel and Huijbregts, 19 91) warrants a geostatistical analysis. It is hypothesized that lunnyu soils are an erosion phase and/or chemical degradation that occurs at landscape positions with greater intensity of hydrological processes (e.g. runoff, erosion). The aim of this analys is was to describe the spatial variability of lunnyu soils at field level in order to understand the relationship between soil variability and slope. 4.2 Materials and Methods The site where the study was carried out is one of the sites previously identifi ed as a lunnyu-affected area (Tenywa, 2004). The taxonomic unit of the soil is a Chromic Lixisol. With the help of a farmer, boundaries of the area perceived to be lunnyu affected drawn basing on performance history of crops judged by the farmer. The garde n was planted with a mixture of bananas ( Musa spp. ) and beans ( Phaseolus spp.). Soil samples were collected from the center point square grids obtained by laying a 180 m x 180 m plot. The point map for the sampled location is show n in the figure below (Fig ure 4 1). The plot contained 81 sampling points at spacing of 20 m x 20 m. An additional 19 random locations within the plot were taken to make a
34 total of 100 sampled locations. The size of the garden was selected go beyond the supposed boundaries of the lunnyu area. The elevation of the area ranged between 1,281 and 1,311 m above sea level and an almost uniform slope of 12% from top t o bottom of the slope (Figure 4 2). A Digital Elevation Model (DEM is valuable in describing topography across a site. The DEM was constructed by kriging interpolation of elevation measurements taken during the georeferencing process. The total area of the plot was 32,400 m2. A venture Cx Global Positioning System (GPS) unit configured for World Geodetic System (WGS 1984) dat um was used to collect in the Universal Transversal Macator Projection (UTM) coordinate system for each sample point. During the georeferencing process, the GPS unit was held vertically above the sample point. The soil attributes studied were texture, pH organic carbon, available P, exchangeable bases (Ca, K, and Mg) and Mg, at two depths (0 20 cm and 20 40 cm) representing top and subsoil, respectively. Available P was determined using Brady and Kurtz No. 1 method (Bray and Kurtz, 1945). The soil was ex tracted by Brady 1 solution and the P determined by the calorimetric procedure using a spectrophotometer. Soil pH was determined using a pH meter (Rhoades, 1982). Exchangeable K, Ca and Mg were measured by treating the soil samples with excess 1 M ammonium acetate solution. Later, the concentrations of exchangeable sodium and K in the extract were measured by flame photometer and the concentration of Ca and Mg was measured by atomic absorption spectrophotometry (Anderson and Ingram, 1989). Laboratory analys is was done in the soils science laboratory of the Faculty of Agriculture, Makerere University. Data were statistically analyzed in three phases: (1) Data were described using descriptive statistics (mean, median, standard deviation, coefficient of variati on and skewness). (2)
35 Geostatistical analysis is most efficient when done on variables that have normal distributions (Webster and Oliver, 2001) because it requires the assumption that the observed data are a realization of a random function, which is intr insically stationarity. Under stationarity, observations of a single realization of the random function at different positions can be treated as a form of replication. The data have to normal to satisfy the stationarity assumption. Therefore, the frequency distribution of each soil property was examined for outliers and the tests for normality using the AndersonDarling test; (3) Correlation coefficients between the different variables were calculated. The statistical analysis of data was carried out using GenStat Discovery Version 3 (VSN International Ltd, UK). (4) Geostatistical analyses involved three steps. For each variable y, a geostatistical analysis was done by estimating the variogram; ) ( 1 2)] ( ) ( [ ) ( 2 1 ) ( h N i i u i uh x z x z h N h where ) ( h : Sem ivariance at lag h h : Distance between data pairs ( zu(xi) zu(xi+h) ) (or : lag) N: Total number of data pairs ( zu(xi) zu(xi+h) ) zu(xi) : Variable u at location xi It included calculation of empirical variograms and determination of the best fitting model ) ( ) ( 1 0b f C C h as judged by the mean squared error, where f(b) is any permissible variogram function, c0 the nugget variance, c1 the sill value and b the range (Chils and Delfiner, 1999; Rodenburg et al ., 2003). Variograms were used in geostat istical interpolation (ordinary kriging), followed by display of predicted values as a map. The variability structure was assumed to be isotropic for all the variables over the study area. Prediction performance was evaluated using the root mean square pre diction error (RMSE). The RMSE provides a summary of the difference between the true (measured) and predicted control point coordinates The
36 program used for geostatistical analysis was VESPER (Variogram Estimation and Spatial Prediction with ERror) versio n 1.6. 4.3 Results 4.3.1 Statistical A nalysis The descriptive statistics for all measured soil properties in the top and subsoil are shown in Table 4 1. The lowest coefficient of variation obtained was 6.98% for pH in the topsoil and the highest was 396.62% for exchangeable Na in the subsoil. The variability of available P and exchangeable K were much higher for both top and subsoil depths compared to any other soil property measured. Conversely, pH was the least variable in both the top and subsoil with c oefficient of variation of 6.98 and 7.99, respectively. Exchangeable Mg varied more in the topsoil than in the subsoil, but was, on average, higher in the topsoil. The ranges in sand, silt and clay were wide in both top and sub soils. In the topsoil, mean sand, and clay were all lower than in the subsoil, but silt was higher in the topsoil than in the subsoil. The correlations among the various soil properties are shown in Table 4 2 below. For clay content, the correlation observed with other soil properti es in both the top and subsoil was negative except for Na and K in the top and subsoil, respectively. At 5% significance level, all the soil properties correlated with at least two other properties in the set, except for Na in the subsoil. The highest corr elation was observed between clay and sand in the subsoil but was also comparatively higher in the topsoil, being negative at both depths. The highest positive correlation was observed between Mg and Ca in the subsoil. Mg and Na in the top soil were almost as highly positive as was Mg and Ca but extremely different correlation in the sub soil. Only sand and clay in the top soil were correlated with elevation. Whereas the correlation between sand and elevation was negative, the one between clay and elevation was positive. This
37 implies that there was more sand at lower elevations than at higher ones while there was more clay at higher elevations. 4.3.2 Geostatistical A nalysis Phosphorus, K, Na and silt in the topsoil, and P and K in the subsoil were transfor med using appropriate sta tistical distributions (Table 4 3). To estimate the variograms for each soil property, the exponential model, using normalized data was employed. The variograms were estimated to a cutoff distance of 180 m and a lag of 20 m, assumi ng that there was no anisotropy. The sill variances of pH, P and exchangeable bases in the top and subsoil were generally smaller than those of textural properties, except silt, which sho wed pure nugget effect (Table 4 4). In the top soil, sand and clay ha d the highest sill, and nugget, while silt had a pure nugget. Phosphorus, Ca, Na and K had shorter ranges between 42 and 58 meters, while the ranges of the other soil properties ranged between 106 and 216 meters. Calcium in the topsoil had the shortest ra nge of 43 meters but also had the longest in the subsoil. The patterns of the soil properties in the subsoil were similar to those in the topsoil. Silt, like in the topsoil, had a pure nugget. Although the sill variances of the textural properties in the s ubsoil were higher than those of other properties in the same layer, they were generally lower than those in the topsoil. Except for Ca and sand, which had longer ranges of 171 and 468 meters, respectively, all the other properties had, on average, an equa l range of 149 meters. Again like in the topsoil, the prediction errors of the textural properties were poorer that those of other properties in the subsoil. Noteworthy is the fact that sand and silt had much smaller prediction errors in the subsoil than i n the topsoil. Considering the RMSE, soil textural properties were poorly predicted compared to other soil properties. Silt had pure nugget effect in both the top and subsoil.
38 In most part of the sampled site, soil chemical properties did not show any cle ar relationship with elevation and relative location of the lunnyu boundary. Noteworthy is that for both top and sub soil, the north-eastern part of the study area was high in most properties, except clay and Na. This part coincides with part of the lunnyu peninsula. The elevation in this part was moderate. 4.4 Discussion The complexity of the compositional and spatial relationships of the soil properties in this study reflects the inherently variable nature of soils (Brady and Weil, 2002). Soil redistribu tion, for example, is controlled by factors such as parent material, climate, terrain, land use and land management interventions. Land management is an important factor to consider in this study, because soils respond differently to land use management ac tivities (Heuvelink et al., 2006). The mean values of the soil properties were not different from those obtained in previous studies in the same area (Taulya, 2002; Tenywa, 2004; Mulumba, 2004). The low coefficient of variation exhibited by pH is character istic of the redistribution pattern of soil water, which tends to make the pH uniform over the area. On the other hand, Na is a microelement and its highly varied occurrence in the soil is primarily determined by the source of this element. The high varia bility of P and exchangeable ions for both top and subsoil, compared to any other soil property measured, could be due to differential vertical (leaching) and horizontal movement. The negative correlation between clay content and other soil properties is expected because the clay binds with the ions, sometimes making them unavailable in soil solution or even on the exchangeable form of soil solids. The proportion of clay and sand has an inverse relationship because the increase in one leads to a decrease i n the other. The high positive correlation observed between Mg and Ca in the subsoil is because these two ions have the same factors such as parent material, leaching and utilization by plants that control their occurrence
39 and distribution in the soil. The discrepancy between the top and subsoil Mg and Ca can be explained by the fact that the differential redistribution of these ions during illuviation/eluviation. Silt had pure nugget effect in both the topsoil and subsoil, but its textural counterparts (s and and clay) showed spatial dependence over a range of 149 meters. This implies that the spatial distribution of silt in the sampled area is random at the spatial scale at which it was obtained. Although Milne (1935, 1936) suggested soil landscape proces ses including erosion deposition and hillslope solute transport as specific mechanisms for the formation of soil landscape patterns, deviations from these important catena concept occur. Silt ranges (minimum maximum values) were much lower in topsoil (1 5%) and subsoil (18%) when compared to sand (topsoil: 54%; subsoil: 49%) and clay (topsoil: 39%; subsoil: 34%) content suggesting that most of the textural variability within the study site was due to variations in sand and clay. Silt showed much more homo genous variation across the study site when compared to heterogeneous patterns in sand and clay (compare point maps). The DEM shows highest values in the southern part of the study area and lowest in the western and northern parts. In contrast, sand showe d higher values at lower elevation and clay showed lower values at lower elevation. The spatial distribution patterns of sand and clay do not seem to be correlated to topography, which suggests that in -situ pedogenic processes (e.g. formation of secondary clays; pedoturbation) due to their dominance may mask erosion/deposition process. The low correlation coefficients between elevation and sand; and elevation and clay indicate limited influence of erosion on textual classes in the sampled area. Theoreticall y, the flat hilltops (summit landscape position) would have greater infiltration potentially leading to greater eluviation illuviation of clays -whereas the steeper backslope regions would be more prone to erosion and A horizon stripping (Brown et al ., 2004).
40 However, given that slope gradient was not so steep and uniformly sloping, coupled with the varied management practices in the sampled area, the redistribution of soil textural classes is not straightforward. Generally, spatial variability of soil chemi cal properties occurs over shorter distances compared to the physical ones, other factors being constant. The reason could be that chemical properties change rapidly with subtle changes in environmental factors compared to physical soil properties. For exa mple, a short rain may cause dissolution and subsequent horizontal and lateral flow of a chemical substance without necessarily causing physical soil movement. In the topsoil, this effect is more pronounced than in the subsoil because the topsoil is more e xposed. Calcium in the topsoil having had the shortest range of 43 meters but also having the longest in the subsoil is further evidence to this proposition. The field that was sampled extends beyond one farmers plot and each plot had different cropping pr actice. For example, whereas the south-western part of the grid was a fallow about two years old, the center part was grown with Bananas ( Musa spp. ) intercropped with common beans ( Phasiolus vulgaris ). The similarity in the patterns of the soil properties in the top and subsoil can be explained by the variability in use practices with in the field (different parts of the field had different crops e.g. bananas, sweet potatoes, beans and fallow). Range is the distance beyond which spatial dependence between soil samples ceases to exist and it can be used as indicator of the appropriate cell size for a field survey in site -specific management. Thus, range is important, both to define the different classes of spatial dependence for these soil variables, and to establish the sampling interval for future surveys. The sampling interval should be less than half of the range of the variogram as a rule of thumb (Kerry and Oliver, 2003). Except for Ca and sand, which had longer ranges of 171 and 468 meters,
41 respecti vely, all the other properties had, on average, showed an equal range of 149 meters. Future studies in this area should consider these ranges to optimize future samplings. A large RMSE means the errors are widely spread, while a small RMSE means the error s are packed tightly around the mean value (Bolstad, 2005). The RMSE of the textural properties were poorer that those of other properties probably because of varied redistribution resulting from land management practices in the sampled field. This situati on reveals that soil silt does not show any kind of spatial dependence, while sand and clay show only weak spatial dependence indicated by the nugget/sill ratios. According to Wang et al (2001) and Grunwald et al. (2007), if the nugget/sill ratio is less than 25%, a variable has strong spatial dependence; between 25 and 75%, the variable has moderate spatial dependence; greater than 75% the variable only shows weak spatial dependence. This implies that all the soil properties have weak spatial dependence, except pH at both depths. There is a possibility that the sample spacing was too coarse to delineate fine -scale spatial variability of soil properties considered in this study. The fact that was no definite pattern providing evidence of erosion/deposition processes lends support to the absence of any clear relationship of soil chemical properties with elevation and relative location of the Lunnyu boundary the in most part of the sampled site. This further suggests in -situ masking of pedogenic processes. No teworthy is that for both top and sub soil, the north -eastern part of the study area was high in most properties, except clay and Na. This part coincides with just one part of the Lunnyu peninsula. The elevation in this part was moderate. 4.5 Conclusions and Recommendations The spatial distribution of the soil properties in the sampled field are not straightforward. All the soil properties, except silt, showed spatial dependence at both depths at the scale of study. The properties P, Ca, Na and sand showed shorter ranges of between 42 and 58 meters, all in the top soil, but the other properties in top and subsoil have larger ranges of 149 meters. The spatial
42 dependence of variables was observed to be larger in subsoil than in the topsoil, which suggests the disturbing effect of tillage on the spatial structure. Figure 4 1 Point map of sampling locations relative to lunnyu soil m m 12% slope N Figure 4 2 Digital Elevation Model for the sampled lunnyu soil in Ma saka, Uganda
43 Table 4 1 Descriptive statistics of field level lunnyu soil in the L. Victoria Basin, Uganda Statistic pH P (ppm) Ca (ppm) K (ppm) Mg (ppm) Na (ppm) Sand (%) Silt (%) Clay (%) Top Soil Mean 6.46 13.96 6.4 5 0.76 2.20 0.08 48.66 15.96 35.32 Starndard error of mean 0.05 1.50 0.28 0.08 0.10 0.01 0.70 0.49 0.61 Median 6.50 8.98 6.08 0.39 2.10 0.07 50.00 16.00 34.00 Minimum 5.00 3.78 1.35 0.16 0.22 0.03 10.00 8.00 18.00 Maximum 7.50 99.59 21.00 3.82 7.65 0.4 7 64.00 46.00 56.00 Coeficient of Variation 6.98 107.35 43.95 100.19 46.39 70.56 14.43 31.00 17.21 Skewness 0.33 3.15 1.49 1.77 1.62 5.24 2.61 3.08 0.51 Sub Soil Mean 6.46 10.37 5.59 0.62 1.82 0.10 46.06 12.78 41.32 Starndard err or of mean 0.05 1.09 0.21 0.07 0.07 0.04 0.51 0.39 0.64 Median 6.60 6.68 5.33 0.31 1.96 0.05 46.00 12.00 44.00 Minimum 4.90 1.54 1.35 0.16 0.19 0.00 32.00 6.00 24.00 Maximum 7.50 94.59 12.90 2.93 3.62 4.00 60.00 24.00 58.00 Coeficient of Variation 7.99 104.74 37.81 108.73 35.80 396.62 11.00 30.25 15.39 Skewness 0.69 5.14 0.71 2.05 0.23 9.80 0.17 0.52 0.14
44 Table 4 2 Spearman rank correlation of field -level lunnyu soil properties Soil Property Elevation pH P Ca Mg K Na S and Silt Top Soil Ph 0.13 P 0.14 0.33** Ca 0.06 0.27** 0.28** Mg 0.07 0.16 0.20* 0.66** K 0.11 0.15 0.24* 0.56** 0.46** Na 0.10 0.03 0.07 0.28** 0.74** 0.37** Sand 0.22* 0.21* 0.33** 0.24* 0.22* 0.05 0. 58** Silt 0.13 0.10 0.03 0.12 0.44** 0.14 0.61** 0.53** Clay 0.19* 0.32** 0.40** 0.36** 0.10 0.05 0.17 0.73** 0.19 Subsoil pH 0.05 P 0.11 0.18 Ca 0.09 0.38** 0.45** Mg 0.08 0.31** 0.38** 0.76** K 0.1 6 0.23 0.37** 0.36** 0.28** Na 0.04 0.13 0.02 0.07 0.06 0.03 Sand 0.16 0.33** 0.27** 0.19 0.21* 0.01 0.13 Silt 0.05 0.25* 0.02 0.27** 0.17 0.06 0.04 0.03 Clay 0.19 0.42** 0.24* 0.33** 0.28** 0.03 0.13 0.81** 0.59** Values with and ** were significant correlated at 0.05 and 0.01 alpha level, respectively.
45 Table 4 3 Descriptive statistics of transformed lunnyu soil properties, Uganda Top soil Sub soil P K Na Silt P K Statistical transformatio n Log 10 Log 10 Square root Square root Log 10 Log 10 Mean 1.07 0.29 0.26 3.89 0.88 0.38 Median 1.01 0.41 0.27 4.00 0.82 0.51 Minimum 0.58 0.80 0.20 2.83 0.59 0.80 Maximum 2.00 0.58 0.33 4.69 1.46 0.47 Standard Deviation 0.30 0.36 0.79 0.38 0. 21 0.35 % Coefficient of Variation 28.13 126.54 16.22 9.89 23.30 92.26 Skewness 0.78 0.67 0.02 0.45 0.86 0.99
46 Table 4 4 Sill, nugget, range and root mean square prediction error (RMSE) lunnyu soil Soil Property Nugge t Sill Range (meters) RMSE Nugget/Sill ratio (%) Topsoil pH 0.12 0.19 149 0.29 63 log P (mg/kg) 0.05 0.06 59 0.08 83 Ca (mg/kg) 3.53 2.85 43 8.18 124 Mg (mg/kg) 0.54 0.35 216 0.78 154 log K (mg/kg) 0.09 0.09 149 0.18 100 Square root Na (mg/kg) 0.49 0.19 47 3.14 258 Sand (%) 20.92 22.81 46 87.62 92 Clay (%) 26.40 18.87 106 51.43 140 square root Silt (%) 15.29 96.17 Subsoil pH 0.15 0.23 149 0.52 65 Log P (mg/kg) 0.04 0.01 149 0.04 400 Ca (mg/kg) 4.09 1.88 468 6.03 218 Mg (mg/kg) 0.38 0.09 149 0.52 422 Log K (mg/kg) 0.10 0.04 149 0.20 250 Na (mg/kg) 0.00 0.00 149 0.01 0 Sand (%) 21.77 7.34 171 28.77 297 Clay (%) 35.48 8.65 149 46.59 410 Silt (%) 15.06 16.94
47 Figure 4 3 Top soil propert ies
48 Figure 4 4 Sub soil properties
49 CHAPTER 5 INFLUENCE OF SLOPE POSITION AND SOIL TYP E ON LUNNYU SOILS 5.1 Introduction Soil variability is a function of soil forming factors climate, parent material, organisms (inclu ding human activities), topography and time. Most tropical soils have undergone severe weathering and the resultant soil properties are mostly dependent on parent material and the influence of human activities, especially in agricultural landscapes (Brown et al., 2004). Because of i ncreased population and the introduction of annual food and cash crops that require more intensive tillage, there is increasing evidence of severe soil degradation due to nutrient mining and erosion (Lufafa et al. 2003). Agricul ture in the Lake Victoria Basin of Uganda is practiced on small holdings ranging from 0.5 1.5 ha. The lunnyu phenomenon is a form of soil infertility described by farmers in this region of the country but the scientific understanding of the situation has not been established. Studies modeling soil variability (Moore et al., 1993; Gessler et al., 2000; Chaplot et al., 2000, 2001; Park et al. 2001; Florinsky et al. 2002; Brown et al ., 2004) have used the catena concept to demonstrate that topographically associated soil profiles are repeated across certain landscapes. According to Lagacherie and Voltz (2000), however, predictive capabilities of these models are limited, especially over large areas, because relationships between soil properties and landscap e attributes are nonlinear or unknown. This is especially important if other soil -forming factors change, such as differing parent materials or variations in land use. Thompson et al. (2006) noted that this lack of transportability of models has never been fully or explicitly tested by developing and validating models for fields from similar landscapes, and therefore warrants investigation into the possible cause and subsequent provision of remedial management interventions. In precision agriculture, site -s pecific modeling of soil properties is thus inevitable.
50 The o bjective of this part of the study is to investigate the influence of slope position and soil type of lunnyu soils. 5.2 Material and Methods Six lunnyu patches were selected from an area covering approximately 50 x 50 km within th e Lake Victoria Basin (Figure 5 1). The lunnyu patches were selected in such a way that they captured different land uses and soil types. Site conditions at each site were described and are showen in Table 5 1 below. At each patch, the slope was divided into three parts; shoulder, back -slope and foot -slope (Fig. 5 3). The division of the slope was such that the points were roughly equidistant from each other at the three positions. In each part of the slope, five locatio ns, separated by a distance between 20 and 30 m across and along the slope, were taken along the contour. Soil samples were taken at two depths (0 20 cm and 2040 cm). It is worthwhile to note that lunnyu patches do not have clear -cut boundaries with some larger and others smaller. The soil samples were taken to the Soil Science laboratory at Makerere University and a similar analysis done as for the field level study (C hapter 4). The descriptive statistics for the tested soil properties are shown in table 5 2 below. Before performing ANOVA, normality tests were performed using the Andersen Darling test. In the top soil, pH, Mg, and K followed a normal distribution; P, Ca and Na were logtransformed, while sand and silt content were arcsine -transformed. In the sub-soil, pH was normally distributed, Ca, Mg were square root transformed and K was log transformed. When comparing different soil depths, data for topsoil and subsoil were combined and also tested for normality. In this case, soil P and percent silt were log transformed while Mg was square root transformed. One -way ANOVA was used to determine the effect of slope position and soil type on individual soil properties at 95% level of confidence using GenStat Discovery
51 Version 3 ( VSN International Ltd, UK ). Whereas Na was the most variable property in the top soil, P was the most variable in the sub -soil, considering the absolute value of the co efficient of variation (Table 5 2). 5.3 Results 5.3.1 Statistical A nalysis A correlation matrix (Table 5 3) showe d that sand and silt have the highest negative correlation while Ca and Mg have the highest positive correlation. Generally, most of the properties had very low correlation (< + 0.3). The highest positive correlation was observed between Ca and Mg both in t he topsoil and subsoil. Sand was highly correlated negatively with silt and clay in both the topsoil and subsoil. Sodium, sand and silt in the topsoil did not show any relationship with soil pH but sand showed a relationship in the subsoil. In the topsoil, only Na and silt did not show a significant relationship with soil P but in the subsoil, only clay showed a significant relationship. In both the topsoil and subsoil, Na showed no relationship with any textural property. 5.3.2 Influence of S oil T ype on S o il P roperties Whereas base cations (Ca, Mg, Na and K) in the topsoil did not vary across soil types, pH, P, sand, clay and silt in topsoil varied among different soil types. Similar topsoil p H values observed for Chromic Lixi sol (CL) and Mollic Glaysol (M G) and were significantly higher than those of the PF a nd PL (Figure 5 2). For soil P, MG had a much higher content than all the other soils types, which did not differ significantly from each other. All the exchangeable bases (Ca, Mg, K and Na) were not s ignificantly different among soil types. Calcium had the highest content in the soil, followed by Mg, then K and least was Na. The sand content of CL and MG was significantly higher than that of Plinthic Ferralsol (PF) and Petrofferic Lixi sol (PL). Of the four soil types, PF had the highest clay content. Clay content of MG was significantly lower than
52 that CL and PL. MG and PL had similar silt content significantly higher than that of CL and PF, which also had similar values. Comparing the textual fractions showed that there was generally higher sand, followed by clay and silt was least. In the subsoil, content of Ca, Mg, Na and K did not differ significantly across soil types. However, pH, P, sand, clay and silt in topsoil varied among different soil types. Topsoil pH values observed for Chromic L ixi sol (CL) and Mollic Gleysol (MG) and were significantly higher than those of the Plinthic F erralsol (PF) and Petrofferic Lixisol (PL) (Figure 5 3). For soil P, MG had a much higher content than all the other soil s types, which did not differ significantly from each other. Percent sand was higher in CL and MG than in PF and PL. MG had the lowest percent clay, followed by CL and PL while PF had the highest content. PL had higher silt content than all the other soil types. No exchangeable bases (Ca, Mg, K and Na) were significantly different among soil types. 5.3.3 Influence o f Slope Position o n Soil Properties Slope position did not significantly influence all the measured topsoil properties except silt, which was significantly lower at the shoulder slope position (Ta ble 5 5). The silt content of the midslope and footslope positions were similar but significantly higher than at shoulder position. In the subsoil, a similar trend as in th e topsoil was observed (Table 5 6). No soil properties showed significant variation for all the slope positions. Soil pH, Ca, Mg, and K were higher in topsoil compared to subsoil (Table 5.6). No other properties differed significantly between top and subsoil. 5.4 Discussion 5.4.1 Descr iptive S tatistics According to Obreza and Rhoads (1988), the critical levels of P, K, Mg and Ca are 10, 45, 33 and 250 mg kg1, respectively. By these standards, phosphorus is the most deficient
53 nutrient while Mg and K are about 30% deficient. Calcium is a bove the critical level and pH is within the optimum range for most crops in the area. In most soils, P tends to move less than Ca and K because of all the different types of chemical reactions that may occur, rendering it insoluble. Ironically, the lower level of P compared to other nutrients could be due to the large quantities utilized by plants. It is also possible that P fixation is high as the soils in the area are highly weathered with potential of high content of aluminum oxides. The observed high a nd negative correlation between sand and silt is expected because these two soil properties are complementary to each other. On the other hand, the high positive correlation between Ca and Mg is explained by the fact that they may have similar parent mater ial mineralogy. For example, mafic mantle -derive rocks typically weather to a smectite and iron oxiderich colloidal fraction with the simultaneous release of both Ca and Mg (Chadwik and Graham, 2000). 5.4.2 Influence of Soil T ype on S oil P roperties Accord ing to the WRB (2006) (World Reference Base for Soil classification), Ferralsols are either red and/or yellow strongly weathered tropical soils with a high content of sesquioxides; resulting in a residual concentration of resistant primary minerals (e.g. q uartz) alongside sesquioxides and kaolinite. The chemical fertility of Ferralsols is poor; weatherable minerals are scarce or absent, and cation retention by the mineral soil fraction is weak. On the other hand, Lixisols have high base status and low activ ity clays throughout the argic horizon and a high base saturation at certain depths and without marked leaching of base cations or advanced weathering of high activity clays. The fact that there was no difference in content of exchangeable base cations was expected for Plinthic ferralsol (PF) and Petroferic Lixisol (PL) and probably Chromic Lixisol (CL) because they exhibit almost similar levels of weathering. Additionally, since they occur within the same climatic zone with similar annual precipitation, le aching differences are expected not to differ significantly.
54 However, the pH did not follow the same trend as for base cations. The significantly higher pH of CL is probably because the site on which the soil occurs is a perennial cropping system (coffee) In perennial systems, there is limited change in vegetation and turning of the soil compared to annual systems. Therefore, the tendency for leaching of base cations may be less likely in annual systems where the soil is turned several times, thereby retu rning leached ions to near -surface layers. Thus, the pH is likely to remain unaffected. On the contrary, the frequency of cultivation may result in a more rapid decomposition of organic matter and weakening of soil structure, which later results in lowerin g soil pH. Steenwerth et al. (2002) found lower values of soil pH in the grassland than in cultivated soils. The low pH under grasslands was attributed to leaching. Some cropping systems may also have an acidifying effect on the soil that is related to the amount of materials removed at harvest, amount and type of fertilizers normally used and the amount of leaching that occurs (Mulumba, 2004). The latest classification of soils in this area is based on fairly old and probably outdated soil survey by Harrop (1960). Over time, land use has altered the soil significantly and the taxonomic units are not proficient in precisely explaining variation in soil properties. Noteworthy is the fact that the soil taxonomic units in the memoirs (Harrop, 1960) on the basi s of which the classification of this study relied, were obtained at a very small scale (1:1,500,000) and therefore the profiles sampled could not have adequately extort lunnyu properties. 5.4.3 Influence of Slope Position on S oil P roperties Relocation o f topsoil material from upper to lower slopes is attributed mainly to the effects of cultivation, either directly, through mechanical movement of soil material during cultivation operations, or indirectly, through the promotion of soil erosion These resul ts corroborated strongly with the findings of Brunner et al. (2004) and Mulumba (2004) in Uganda where it was observed that s oils at the summit position had a thick solum due to the stable soil
55 formation on the flat surface, and soils at the shoulder posit ion had shallow A -horizons due to active erosion processes. Mulumba (2004) also observed no significant different in pH at different slope positions in nonlunnyu soils. However, the observed pattern in soil properties is difficult to reconcile within thes e concepts. We expect that the soil redistribution and subsequent formation of distinct soil layers along a toposequence should be reflected in differences in other soil properties. Paradoxically, Webb and Burgham (1997) did not find indication of the expe cted catenary relationship involving translocation of exchangeable bases from upper to lower slopes as no differences were observed in these aspects. In the current study, t he reason could be that considering the gradient of all the sites that ranges betwe en 3 and 13%, soil redistribution may not be significant to cause distinct patterns in soil properties at different slope position. It is also possible that pedeturbation (in -situ processes along the vertical soil profiles) mask and overshadow horizontal t ransport and erosion/deposition processes. This is quite common in African soils. 5.4.4 Influence of Soil D epth on S oil P roperties The difference in salt content between top and sub soil is the net balance between leaching and upward flux due to evapotranspiration (ET). When the downward leaching flux of water exceeds upward flux due to ET, soluble salts are minimal throughout the profile. When leaching is slightly greater than ET, salts are leached from the surface to deeper soil layers. When ET exceeds l eaching, salts are carried to the evaporating surface (Chadwik and Graham, 2000). Differences in vegetation cover can influence the rate of ET and therefore the rate and direction of movement of ions in the soil. The random changes in the vegetation cover and plant species over the seasons could have introduced the unpredictable variability in soil cations at different depths. All the soil types in the present study lie in the region with the same rainfall pattern and therefore rainfall likely did not diffe rentiate leaching at different sampling sites. Soil
56 pH, Ca, Mg, and K were higher in topsoil compared to subsoil. The soil depth of 0 20 and 2040 cm was subjectively selected and may not reflect real top/sub soil profile characteristics. The study area is a generally gently sloping area with gradient of 6 15%. The influence of slope can be thought to be general insufficient to make significant contribution to microclimatic conditions that result in differences in soil depth. In future, it would be better t o compare properties of horizons rather that fixed depth as was the case in this study. 5.5 Conclusions and Recommendations Lunyu soils on Chromic Lixisols and Mollic gleysols had higher pH, P, sand, clay and silt compared to those on Plinthic ferralsols and Petrifferic Lixisols All the soil properties were not influenced by slope position. Soil pH, Ca, Mg, and K were higher in topsoil compared to subsoil. Neither slope position nor the type of lunnyu has showed consistent differences in all the soil prop erties. This means that the lunnyu phenomenon cannot be explained by these two factors and other factors such as mineralogy and soil management aspects. There may also be need to redefine the concept of lunnyu to understand the basis for the local definiti on and further explore its occurrence and description in other parts of the country. Getting a more recent classification and management history of these soils could be useful in further understanding the causes and trends. Factors affecting nutrient utili zation efficiency by crops in this area are a critical area worth of exploration. Field experimentation on lunnyu soils with some crops needs to be done to establish the relationship between crop productivity and soil quality, especially to cover longterm trends. There is a strong need for research to address management levels in more detail, encompassing weed cover, mulch cover, ground cover, and the crop stand (varieties). Poorly managed systems would be separated into different classes in order to quant ify the management effects.
57 Table 5 1 Characteristics of sampled landscapes Sub county Soil type (FAO) Land use Notes Byakabanda Mollic Gleysol (MG) Annual crops Grown with maize and sweet potatoes with an Eucalyptus plantation covering the lower quarter of the sampled area. Slope was 6%. Kabila Mollic Gleysol (MG) Annual crops Planted with bananas, and formerly maize millet. Lower part seasonally water logged but upper half with high gravel content. Slope was 3%. Kalisizo Petroferic Lixisol (PL) Pasture land Pasture land with patches of thickets ranging between 5 and 10 m2. Slope was 7%. Lwebitakuli Chronic Lixisol (CL) Perennial crops Only coffee, aged about 10 years according the farmer. Slope was 8%. Lwengo Plinthic Ferralsol (PF) Annual crops Maize and cassava intercropped and coffee occupying larger part on lower part of the slope. Age of coffee plants was about 25 years. Slope is 10%. Masaka Plinthic Ferralsol (PF) Coniferous forest Pure Eucalyptus on the lower side and maize ( Zea maize ), potatoes ( Impomea batatus ). Slope was 13%.
58 Figure 5 1 Idealized locations of sampling positions on the slope
59 Table 5 2 Descriptive statistics of lunnyu soils in the Lake Victoria Basin, Uganda Statistic pH Av. P (ppm) Ca (ppm) K (ppm) Mg (ppm) Na (ppm) Sand (%) Silt (%) Clay (%) Top Soil Mean 5.52 9.19 4.07 0.49 1.43 0.04 54.44 15.18 30.04 Standard error of mean 0.05 1.55 0.24 0.04 0.08 0.00 1.73 1.12 1.12 Median 5.40 3.71 3.23 0.33 1.10 0.04 58.00 10.00 28.00 Minimum 4.40 1.10 1.05 0.15 0.24 0.02 16.00 4.00 8.00 Maximum 6.90 92.59 12.60 1.56 4.01 0.09 76.00 56.00 52.00 Coefficient of Variation 8.35 160.07 56.44 73.54 51.77 34.32 30.19 69.84 35.25 Skew ness 0.61 3.12 1.48 1.31 1.43 0.63 0.87 1.81 0.25 Sub soil Mean 5.52 38.76 11.30 11.06 8.58 5.02 33.64 22.56 22.10 Standard error of mean 0.05 23.80 7.69 10.42 7.22 4.88 11.29 10.69 7.37 Median 5.40 3.71 3.23 0.49 1.43 0.04 30.19 10.00 28.00 Minimum 4.40 1.10 0.24 0.04 0.08 0.00 0.87 1.12 0.25 Maximum 6.90 160.07 56.44 73.54 51.77 34.32 76.00 69.84 52.00 Coefficient of Variation 8.35 162.45 179.92 249.17 222.51 257.42 88.78 125.36 88.25 Table 5 3 Sp earman rank correlation of field -level soil properties of lunnya soils pH P Ca Mg K Na Sand Silt Top Soil P 0.33** Ca 0.34** 0.36** Mg 0.32* 0.34** 0.84** K 0.33** 0.21* 0.37** 0.28** Na 0.02 0.06 0.36** 0.31** 0.06 Sand 0.14 0.23* 0.15 0.20 0.21* 0.12 Silt 0.13 0.00 0.30** 0.33** 0.28** 0.07 0.72** Clay 0.32** 0.35** 0.11 0.09 0.06 0.13 0.76** 0.13 Sub Soil P 0.39** Ca 0.34** 0.11 Mg 0.26* 0.06 0.62** K 0.50** 0.19 0.26 0.15 Na 0.11 0.07 0.26* 0.16 0.02 Sand 0.09** 0.03 0.26* 0.11 0.29** 0.10 Silt 0.15 0.23 0.33** 0.18 0.30** 0.05 0.77** Clay 0.29** 0.27** 0.08 0.01 0.16 0.09 0.79** 0.23* Values with and ** were significant correlated at 0.05 a nd 0.01 alpha level, respectively.
60 5% error bars 0 1 2 3 4 5 6 7 pH 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 ln P (mg/kg) 1.35 1.40 1.45 1.50 1.55 1.60 1.65 1.70 1.75 ln Ca (mg/kg) 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 Ln Mg (mg/kg) 0 0.5 1 1.5 2 2.5 3 3.5 Ln K (mg/kg) 0.00 0.20 0.40 0.60 0.80 1.00 1.20 Na (mg/kg) 0 10 20 30 40 50 60 70 80 Chromic Lixisol (CL) Plinthic Ferralsol (PF) Mollic Gleysol (MG) Plinthic Lixisol (PL) Sand (%) 0 5 10 15 20 25 30 35 40 45 50 Chromic Lixisol (CL) Plinthic Ferralsol (PF) Mollic Gleysol (MG) Plinthic Lixisol (PL) Clay (%) 0 0.5 1 1.5 2 2.5 3 Chromic Lixisol (CL) Plinthic Ferralsol (PF) Mollic Gleysol (MG) Plinthic Lixisol (PL) Ln Silt (%) Figure 5 2 Topsoil properties across (N for CL = 15; PF = 30; MG = 30 and PL = 15)
61 5% error bars 0 1 2 3 4 5 6 7 pH 0 1 2 3 4 5 6 Ln P (mg/kg) 0 2 4 6 8 10 12 14 Square rook Ca (mg/kg) 0 1 2 3 4 5 6 Square rrot Mg (mg/kg) 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 Ln K (mg/kg) 0 0.2 0.4 0.6 0.8 1 1.2 Na (mg/kg) 0 10 20 30 40 50 60 70 80 Chromic Lixisol (CL) Plinthic Ferralsol (PF) Mollic Gleysol (MG) Plinthic Lixisol (PL) % Sand 0 10 20 30 40 50 60 Chromic Lixisol (CL) Plinthic Ferralsol (PF) Mollic Gleysol (MG) Plinthic Lixisol (PL) % Clay 0 5 10 15 20 25 Chromic Lixisol (CL) Plinthic Ferralsol (PF) Mollic Gleysol (MG) Plinthic Lixisol (PL) % Silt Figure 5 3 Subsoil properties (N for CL = 15; PF = 30; MG = 30 and PL = 15) Table 5 4 Soil properties across slope positions in the Lake Victoria basin, Uganda Slope Position N pH Ln P Ln Ca Ln Mg Ln K Na Sand Clay Ln Silt Shoulder slope 30 5.64 4.69 1.59 3.43 2.69 0.90 52.80 30.07 2.38 a (0.47) (0.71) (0.09) (0.41) (0.69) (0.27) (17.58) (10.73) (0.61) Mid slope 30 5.7 4.83 1.57 3.36 2.87 0.96 54.73 31.07 2.54 b (0.59) (0.60) (0.09) (0.60) (0.70) (0.39) (14.99) (10.45) (0.44) Foot slope 30 5.7 4.81 1.61 3.44 2.59 0.96 5 5.80 29.00 2.51 b (0.39) (0.75) (0.10) (0.53) (0.57) (0.30) (17.04) (10.85) (0.54) F value 0.18 0.35 1.00 0.20 1.39 0.30 0.25 0.28 7.52 P Value 0.837 0.706 0.372 0.822 0.254 0.739 0.777 0.756 0.038 Values in parentheses show the Standar d Deviation
62 Table 5 5 Subsoil soil properties across slope positions in the Lake Victoria basin, Uganda Slope Position N pH Ln P Ln K Na Sand Clay Silt Midslope 30 5.5 4.52 11.18 5.03 2.46 0.84 50.27 34.70 14.70 SD (0 .58) (0.58) (2.49) (1.12) (0.43) (0.32) (16.22) (11.08) (10.14) Shoulder 30 5.5 4.50 11.31 5.31 2.28 0.94 53.00 32.03 14.97 SD (0.41) (0.66) (2.39) (0.76) (0.53) (0.24) (16.64) (11.03) (10.96) Toeslope 30 5.6 4.76 10.93 5.06 2.3 8 0.89 52.00 32.10 16.23 SD (0.37) (0.75) (2.39) (1.14) (0.41) (0.26) (19.38) (12.12) (11.60) F 1.2 1.40 0.20 0.66 1.01 1.15 0.19 0.53 0.17 P 0.32 0.25 0.82 0.52 0.37 0.32 0.83 0.59 0.85 Values in parentheses show the Standard Deviation Table 5 6 Soil properties of selected lunnyu soils at two depths Soil depth pH Ln P Ca Mg K Na Sand Clay Ln Silt Topsoil 5.68 a 4.75 a 4.96 a 5.68 a 2.72 a 0.94 a 54.44 a 30.04 a 2.54 a (0.49) (0.65) (0.51) (1.42) (0.66) (0.15 ) (16.43) (10.59) (0.56) Subsoil 5.52 b 4.64 a 4.79 b 5.26 b 2.45 b 0.93 a 51.76 a 32.94 a 2.54 a (0.46) (0.73) (0.45) (1.22) (0.56) (0.15) (17.31) (11.36) (0.59) F Value 5.42 0.93 5.61 4.64 8.58 0.17 1.14 3.14 0.01 P Value 0.021 0.335 0. 019 0.033 0.004 0.685 0.287 0.078 0.941 Within each column, means with the same superscript do not differ significantly at 5%. Values in parentheses show the Standard Deviation
63 CHAPTER 6 GENERAL DISCUSSION AND CONCLUSION Answering the following questions would be useful in diagnosing the occurrence and distribution of lunnyu soils? First, are there specific crops/vegetation types that do or do not survive on lunnyu soils? Second, are there pedological and/or other observable features unique to these soi ls? The first question relates to the genesis of name lunnyu. The name lunnyu, by its interpretation in the local language, was probably premised on the fact that these unproductive soils could be having excess quantities of salts that deter crops from performing as expected. This meant that the cause of infertility was chemical degradation rather than anything else. What one would expect is that the plants growing on a lunnyu area should have symptoms of soil chemical degradation nutrient deficiency, salinity, alkalinity or sodicity (such as scorching, chlorosis and overall stunting). It is not reported in the literature whether the farmers that coined this name of these unproductive soils had evidence of these various symptoms. Literature available do es not provide a comprehensive evaluation of any of these possibilities (see Zaake, 1986; Tenywa, 2004), implying that the description of the infertile patches was inept and requires revisiting. Information from farmers indicates that some crops could thr ive appreciably well on lunnyu soil, while others could not. There is no reliable information on the crops that grow or do not grow well on lunnyu soils. This could be a useful indicator for investigating the relationship between crop performance and soil properties in more detail in the future. From informal discussions with farmers, it was clear that they are unable to distinguish lunnyu from nonlunnyu soil on the basis of physical appearance. Those who claim to know the soil have either experienced it o f being told about the location of these soils. Chemical analysis at both field and
64 landscape levels does not show that the properties of the lunnyu soils are any different from those of other non lunnyu soil. It is agreeable (based on classification of th e soils) that soils in this area are generally less fertile and require careful management interventions in order to obtain good crop yields. Therefore, I believe that application of appropriate soil management practices such as improvement of organic matt er levels, nutrient inputs and erosion control could contribute to improved soil quality for crop production in the area. The second question elicits concerns about farmers capability to map lunnyu soil in the landscape, a fundamental premise used by Z aake (1986) to compare chemical properties of lunnyu and non -lunnyu soils. This could have been the cause of failure to detect any particular patters of soil properties by Zaake (1986) and Tenywa (2004). Similarly, modelling these soils in this study did not reveal specific properties to be influential in controlling soil quality for crop growth. Related to this, the severity of the lunnyu phenomenon has never been highlighted. It is likely that over time, the severity would either decrease or increase depe nding on the cause. For example, if it were accumulation of salts, continuous precipitation would leach much of them and reduce severity, but protracted drought would exacerbate the condition. This theory should be the basis for further investigation into the temporal variability of lunnyu soils. Neither previous studies nor this study are able to precisely identify the cause and spatial variability of lunnyu soils. There is need to return to the origin of the concept of lunnyu before embarking on further s ystematic investigations.
LIST OF REFERENCES Aluma J. R. W., 2001. Vegetation. pp. 169185. In: Mukiibi (Ed). Agriculture in Uganda General information. NARO -Uganda Anderson J. M. and Ingram J. S. I., 1989. Tropical soil biology and fertility pp 383 9. In: Anderson J. M. and Irama J. S. I. (Eds). A handbook of methods of analysis. CAB International. Aniku J. F. R., 2001. Soil classification and pedology. pp. 66 103. In: Mukiibi (Ed). A griculture in Uganda General information. NARO -Uganda. Aranda V and Oyonarte C., 2006. Characteristics of organic matter in soil surface horizons derived from calcareous and metamorphic rocks and different vegetation types from the Mediterranean high -mountains in SE Spain. European Journal of Soil Biology 42(4): 247 258 al variability of soil enzyme activities in pasture topsoils using geostatistics European Journal of Soil Biology 42(4):230 -237 Bamutaze Y., 2005. Impact of land-use on soil loss from Wanale micro-catchment, Mt. Elgon. M.Sc. thesis, Makerere University. Bishop T. F. A and Minasny B., 2006. Digital soil -terra in modelling: predictive potential and uncertainty. pp 185 214. In S. Grunwald (ed.) Environmental Soil -Landscape Modeling: Geographic Information Technologies and Pedometrics CRC Press, New York. Bocar D., 2003. Land degradation in SubSaharan Africa: What explains the wide spread adoption of unsustainable farming practices? Working paper, Department of Agricultural Economics and Economics, Montana State University, Bozeman, MT, USA. Bouyocos G. J., 1962. Hydrometer method improved for making particle si ze analyses of soils. Journal of Agronomy 53: 464-465. Brady and Weil, 2002. The Nature and Properties of Soils. 13th Edition. Prentice Hall. Brown D. J., Clayton M. K. and McSweeney K., 2004. Potential terrain controls on soil color, texture contrast and grain -size deposition for the original catena landscape in Uganda. Geoderma, 122(1):51-72. Brown L. R., 2004. Outgrowing the Earth: the food security challenge in the age of falling water tables and raising temperatures. Earth Policy Institute, USA. pp 23 9. Brunner A. C., S. J. Park, G. R. Ruecker, R. Dikau and P. L. G. Vlek 2004. Catenary soil development influencing erosion susceptibility along a hillslope in Uganda CATENA, 58(1):1-22
Chaplot, V., Bernoux, M.,Walter, C., Curmi, P., Herpin, U., 2001. Soil carbon storage prediction in temperate hydromorphic soils using a morphologic index and digital elevation model. Soil Science 166, 48 60. Chaplot, V., Walter, C., Curmi, P., 2000. Improving soil hydromorphy prediction according to DEM resolution and avai lable pedological data. Geoderma 97, 405 422. Chenery E. M. 1954. Luunyu soils in Uganda. Dept. of Agriculture records of investigation, 19481954. pp. 3253. Chittleborough D. J., 1978. Soil variability within land systems and land units at Monarto, Sout h Australia. Australian Journal of Soil Research 16(2) 137-155 Crowley, E.L., Carter, S.E., 2000. Agrarian change and the changing relationships between toil and soil in Maragoli, western Kenya (1900 1994). Hum. Ecol. 28, 383 414. Deckers, J., 2002. A sys tem approach to target balanced nutrient management in soilscapes of sub -Saharan Africa. In: Vanlauwe, B., Diels, J., Sanginga, N., Merckx, R. (Eds.), Integrated Plant Nutrient Management in Sub -Saharan Africa. From Concepts to Practice. CAB International, Wallingford, Oxon, UK, pp. 47 61. Duffera M., Whiteand J. G. and Weisz R., 2007. Spatial variability of Southeastern U.S. coastal plain soil physical properties: Implications for site -specific management Geoderma 137 (3 -4 ) 327-339 FAO, 2005. Management of Degraded Soils in Southern and East Africa (MADSSE A Network) Online at: http://www.fao.org/ag/agl/agll/madssea/intro.htm Florinsky, I. V., Eilers, R. G., Manning, G. R., Fuller, L. G., 2002. Prediction of soil properties by digital terrain modelling. Environmental Modelling Software 17, 295 311. Gallant, J.C., Wilson, J.P., 1996. TAPES G: a grid based terrain analysis program for the environmental sciences. Computer Geoscience 22, 713 722. Gessler, P. E., Chadwick, O. A., Chamran, F., Alth ouse, L. D., Holmes, K. W., 2000. Modeling soil landscape and ecosystem properties using terrain attributes. Soil Science Society of America Journal 64, 2046 2056. Grunwald S. 2006. What do we really know about the space time continuum of soil landscapes? pp. 3 36. In S. Grunwald (ed.) Environmental Soil -Landscape Modeling: Geographic Information Technologies and Pedometrics CRC Press, New York. Grunwald S. and Lamsal S., 2006. The impact of emerging geographic information technologies on soil landscape m odelling. pp. 127 154. In S. Grunwald (ed.) Environmental Soil Landscape Modeling: Geographic Information Technologies and Pedometrics CRC Press, New York.
Grunwald S., K. R. Reddy, J.P. Prenger and M.M. Fisher. 2007. Modelling of the spatial variability of biogeochemical soil properties in a freshwater ecosystem. Ecological Modeling, 210: 521-535 Grunwald S., K. R. Reddy, S. Newman and W.B. DeBusk. 2004. Spatial variability, distribution, and uncertainty assessment of soil phosphorus in a south Florida we tland. Journal of Environmetrics, 15: 811-825. Hadoto D. P. M., 2001. Geology. pp. 33 65. In: Mukiibi (Ed). Agriculture in Uganda General information. NARO -Uganda. Hartemink A. E. 2003. Soil fertility decline in the tropics with case studies on plantati ons. ISRIC -CABI, Wallingford. 360 pp. Hillel D., 1998. Environmental Soil Physics Academic Press. pp 771 Holly, D. C. & Ervin, G. N. 2006. Characterization and quantitative assessment of inter -specific and intraspecific penetration of below -ground vegetat ion by cogongrass Imperata cylindrica (L.) Beauv.) rhizomes. Weed Biology and Management 6 (2), 120 123. Hulugalle N. R., Lal R. and Te Kiula H. H. C., 1984. Soil physical changes and crop root growth following different methods of land clearing in Wester n Nigeria. Soil Science 138:172179 Iqbal J., Read J. J., Thomasson A. J. and Jenkins J. N., 2005. Relationships between s oil landscape and dryland cotton lint yield. Soil Science Society of America Journal 69(3):872-882 Jenny H., 1941. Factors of soil for mation. McGraw Hill. New York. Jiang P. and Thelen K. D., 2004. Effect of soil and topographic properties on crop yield in a North Central corn soybean cropping system. Agronomy Journal 96:252-258 Journel, A.G., Huijbregts, C.J., 1991. Mining Geostatistic s 5th edition. Academic Press, London. Junior V. V., Cavalho M. P., Dafonte J., Freddi O. S., Vazquez V. E., and Ingaramo O. E., 2006. Spatial variability of soil water content and mechanical resistance of Brazilian ferralsol. Soil and Tillage Research 85: 166-177. Kaspar T. C., Pulido D. J., Fenton T. E., Colvin T. S., Karlen D. L, Jaynes D. B. and Meek D. W., 2004. Relationship of corn and cow bean yield to soil and terrain properties. Agronomy Journal. 96:700-709. King, D., Bourennane, H., Isambert, M. Macaire, J.J., 1999. Relationship of the presence of noncalcareous clay loam horizon to DEM attributes in a gently sloping area. Geoderma 89, 99 111.
Kisamba Mugerwa W., 2001. Social Background. pp.186199. In: Mukiibi (Ed). Agriculture in Uganda Gene ral information. NARO -Uganda. Komutunga E. T and Musiitwa F., 2001. Climate. pp. 21 32. In: Mukiibi (Ed). Agriculture in Uganda General information. NARO -Uganda. Kravchenko A. N. and Bullock D. G., 2000. Correlation of corn and soybean grain yield with topography and soil properties. Agronomy Journal 92: 75-83 Landi, A., Mermut, A.R., Anderson, D.W., 2004. Carbon distribution in a hummocky landscape from Saskatchewan, Canada. Soil Science Society of American Journal 68, 175 184 Li Y., and Lindstrom M. J., 2001. Evaluating Soil Quality Soil Redistribution Relationship on Terraces and Steep Hillslope. Soil Science Society of America Journal 65:15001508 Lippincott C. L. 2000. Effects of Imperata cylindrica (cogongrass) invasions on fire regimes in Florid a sandhill. Natural. Areas Journal. 20, 140 149. Lufafa A., 2000. Validation of pedotransfer functions in a banana-based micro-catchment of the Lake Victoria basin, Uganda. MSc. Thesis, Makerere University -Kampala, Uganda. Lufafa A., Tenywa, M. M. Isabiry e, M. Majaliwa M. J. G. & Woomer. P. L., 2003. Prediction of soil erosion in a Lake Victoria basin catchment using a GIS -based Universal Soil Loss Model. Agricultural Systems 76:883-894. Martin W.K. E. and Timmer V. R., 2006. Capturing spatial variability of soil and litter properties in a forest stand by landform segmentation procedures. Geoderma 132: 169 181 McBratney A. B., Odeh, I. O. A., Bishop T. F. A., Dunbar, M. S. and Shatar, T. M. 2000. An overview of pedometric techniques for use in soil survey. Geoderma 97: 293-327. Micheni A. and Irungu J., 2003. Farmers perceptions versus chemical analysis of soil fertility status in semi arid areas: A case study of Tharaka district. Proceedings of the 21st annual conference of the soil science society of East Africa. pp. 523 526 Moore, I. D., Gessler, P. E., Nielsen, G. A., Peterson, G. A., 1993. Soil attribute prediction using terrain analysis. Soil Science Society of America Journal 57, 443 452. Mulumba L. N., 2004. land us effect on soil quality and productivity in the Lake Victorial basin of Uganda. PhD Thesis, The Ohio State University, USA. Officer, S.J., Tillman, R.W. Palmer A.S. and Whitton J.S., 2006. Variability of clay mineralogy in two New Zealand steep land topsoils under pasture. Geoderma 132 (3-4): 427-440. Olson C. G., 2006. Geomorphological Soil -Landscape models. Pp 105124. In: In S. Grunwald (ed.) Environmental Soil -Landscape Modeling: Geographic Information Technologies and Pedometrics CRC Press, New York.
Park S. J. and van de Giesen N., 2004. Soil landscape delineation to define spatial sampling domains for hillslope hydrology. Journal of Hydrology 295 (1-4): 28 -46 Park, S. J., McSweeney, K., Lowery, B., 2001. Identification of the spatial distribution of soils using a process based terra in characterization. Geoderma 103, 249 272. Pennock, D. J., Corre, M. D., 2001. Development and application of landform segmentation procedures. Soil Tillage Research 58, 151 162 Phillips J. D. Perry D., Garbee A. R., Carey K., Stein D., Morde M. B., She ehy J. A., 1996. Deterministic uncertainty and complex pedogenesis in some Pleistocene dune soils. Geoderma. 73 (3/4). 147-164 Reddy B. S .R., Balasubramanian S., Phillips J. D., Perry D., Garbee A. R., Carey K., Stein D., Morde M. B., Sheehy J. A., 1996. Deterministic uncertainty and complex pedogenesis in some Pleistocene dune soils. Geoderma 73 (3): 147-164 Rhoades J. D., 1982. Soil testing. pp 16 21. In: Page A. l., Miller R. M. and Keeney D. R. (Eds). Methods of soil analysis Part 2 2nd edition. American society of Agronomy. Rivero R. G., Grunwald S., Osborne T. Z., Reddy K. R. and Newman S. 2007. Characterization of the spatial distributio n of soil properties in Water Conservation Area 2A, Everglades, Florida. Soil Science 172(2): 149-166 Sanchez, P.A., Swaminathan, M.S., Dobie, P., Yuksel, N. 2005. Halving hunger: it can be done. Summey version. UN Millennium Project. Task Force on Hunger, pp 1 3. Sawhney J. S., Deka B., Sharma B. D. and Sidhu P. S., 1996. Magnitude of soil variability in morphological and other properties across different landscapes in the Siwalik hills of Punjab. Journal of the Indian Society of Soil Science. 1996. 44 (3) 465-469 Scherr, S. J., 1999. Soil Degradation: A Threat to DevelopingCountry Food Security by 2020? Brief No. 58 February 1999. Soil Use and Management 21 (1): 5 8 64 Singh, N., Kar, A., 2001. Characteristics of major soils of Banni mudflat in arid western India and their relationship with topography. Journal of Arid Environments 48, 509 520. Smaling, E.M.A., Nandwa, S.M., Janssen, B.H., 1997. Soil fertility in Africa is at stake. In: Buresh, R.J., Sanchez, P.A. (Eds.), Replenishing Soil Fertility in Africa. ASA, CSSA, SSSA, Madison, WI, pp. 47 61. Sokal R. R. and Rohlf J. F., 1995. Biometry: The principles and practice of Statistics in Biological Research 3rd Edition. Freeman and Company. p. 887. Ssali H., 1975. Unpublished PhD dissertation. Ohio State University.
Steenwerth K. L., L. E. Jackson, F. J. Calderon, M. R. Stromberg, and K. M. Scow, 2002. Soil community composition and land use history in cultivated and grassland ecosystems of coastal California. Soil Biology & Biochemistry 34 (11): 1599-1611 Taulya G. 2004. Topsoil depth-banana yield relationships on a chromic luvisol in a Lake Victoria basin microcatchment. MSc. thesis. Makerere University, Uganda. 72 pp. Tenywa M. M., 1993. Soil erosion and overflow processes on spatially variable soils PhD Thesis. The Ohio State University. Tenywa M. M., 2004. Characterization of unproductive Luunyu soils for improved agricultural productivity in the Lake Victoria Basin, IFS research project grant C/33921 report. Uganda. Terra J. A., Joey N. Shaw, D. W. Reeves, R. L. Raper, E. van Santen, E. B. Schwab, and P. L. Mask 2005: Soil management and landscape variability affects Field -Scale cotton productivity: Soil Scie nce Society of America Journal 70: 98-107. Thompson, J. A., Bell, J. C., Butler, C. A., 1997. Quantitative soil landscape modeling for estimating the areal extent of hydromorphic soils. Soil Science Society of America Journal 61, 971 980. Thompson, J. A., Bell, J. C., Butler, C. A., 2001. Digital elevation model resolution: effects on terrain attribute calculation and quantitative soil landscape modeling. Geoderma 100, 67 89. Thompson, J. A., Eugenia M. Pena Yewtukhiw and John H. Grove, 2006. Soil landscape modeling across a physiographic region: Topographic patterns and model transportability. Geoderma 133 (1-2): 57 -70. Timothy A. Q Les R. B., and Andrew P. N., 2003. Tillage erosion intensity in the South Cant erbury Downlands, New Zealand. Australian Journal of Soil Research 41 (4): 78980 Tittonell P., Vanlauwe B., Leffelaar P.A., Rowe E.C., Giller K.E.. 2005a. Exploring diversity in soil fertility management of smallholder farms in western Kenya I: Heterogen eity at region and farm scale. Agriculture, Ecosystems and Environment 110 149 165 Tittonell P., Vanlauwe B., Leffelaar P.A. Shepherd K.D., Giller K.E., 2005b. Exploring diversity in soil fertility management of smallholder farms in western Kenya II. Within farm variability in resource allocation, nutrient flows and soil fertility status. Agriculture, Ecosystems and Environment 110 (2005) 166 184
Vamocil, J. A., 1965. Soil porosity calculation from particle density. Agricultural Engineering. 25:447-500. Van Oost, K., Govers, G. Van Muysen W. and Quine T. A., 2000. Modeling Translocation and Dispersion of Soil Constituents by Tillage on Sloping Land. Soil Science Society of America Journal 64:1733-1739. Vanlauwe, B., Diels, J., Lyasse, O., Aihou, K., Iwuaf or, E.N.O., Sanginga, N., Merckx, R., Deckers, J., 2002. Fertility status of soils of the derived savanna and northern guinea savanna and response to major plant nutrients, as influenced by soil type and land use management. Nutr. Cycl. Agroecosyst. 62, 139 150. Veronese V.J., Carvalho M. P., Dafone J., Freddi O. S., Vidal E. V. and Ingaramo O. E., 2006. Spatial variability of soil water content and mechanical resistance of Brazilian ferralsol. Soil and Tillage Research 85: 166-177. Walkley A. and Black I. A., 1934. An examination of the Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Science 37: 29 -38. Wang, J., Bojie, F., Qiu, Y., Liding, C., Wang, Z., 2001. Geostatistical analys is of soil moisture variability on Da Nangou catchment of the Loess Plateau, China. Environmental Geology 41, 113 120 Webb T. H. and Burgham S. J., 1994. Catenary relationships of downland soils derived from loess, South Canterbury, New Zealand Australian Journal of Soil Research 32(1) 1 11 Webb T. H. and Burgham S. J., 1997. Soil landscape relationships of downlands soils formed from loess, eastern South Island, New Zealand. Australian Journal of Soil Research, 35 (4) 827 842. Webster R. and Oliver M. A., 2001. Geostatistics for environmental scientists. Wiley, 271 pp., ISBN 0471965537 Webster R. and Oliver M. A., 2006. Modelling spatial variation of soil as random functions. pp. 241287. In: Grunwald S. (Ed.); Environmental Soil -Landscape Modeling; Ge ographic Information Technologies and Pedometrics CRC Press, New York. Whitford W. G. and Kay F. R., 1999. Biopedturbation by mammals in deserts: a review. Journal of Arid Environments. 41(2): 203-230 Wright A. L. and Hons F. M, 2005. Soil carbon and ni trogen storage in aggregates from different tillage and crop regimes Soil Science Society of America Journal 69: 141-147. Wysocki, D. A., Schoeneberger, P. J., LeGarry, H. E., 2000. Geomorphology of soil Landscapes. In: Sumner, M.E. (Ed.), Handbook of Soi l Science. CRC Press, Boca Raton, FL, pp. 5 40.
Zake J. K. Y., 1986. Preliminary findings on the investigations of the causes of infertility (Luunyu) in some Uganda soils. 115120. In: Proceedings of 8th Annual meeting of the Soil Science Society of East Africa. December 1986 Kampala -Uganda. Zhang, G. L. Zhang, H. 2005. Landscape -scale soil quality change under different farming systems of a tropical farm in Hainan, China. Soil Use and Management 21 (1): 58-64.
BIOGRAPHICAL SKETCH Bernard Lukoye Fungo w as born to Mr. Francis Lukoye and Mrs. Petronilla Nambuya, both of Manafa district, Uganda in 1979. He lived in Kamwokya, Kampala until 1986. Thereafter, his family moved to Makerere University Farm in Kababyolo, Uganda, where he lived until the year 2000 before moving to Wandegeya in Kampala. His seven of primary school were spent at City Primary School in Kololo, four years of Ordinary level at Bukedi College Kachonga and two years of Advanced Level at St. Peters College, Tororo. In October 2004, he was awarded a Bachelor of Science degree in forestry from Makerere University. While living in Kabanyolo, his frequent interaction with agricultural researchers created a liking for agricultural sciences and this led him to study Forestry. His interest in this profession culminated into applying for a Master of Science in Soil Science, this thesis being part of the requirements for the award on the MS degree in Soil Science of the University of Florida.