SPATIO-TEMPORAL DYNAMICS OF LAND USE CHANGE ON RIVERS IN TROPICAL WATERSHEDS: A CASE STUDY OF RUIRU AND NDARUGU BASINS, KIAMBU COUNTY, KENYA.
BY
GEOFFREY MWANGI WAMBUGU
A THESIS SUBMITTED IN PARTIAL FULFILLMENT FOR THE REQUIREMENTS FOR THE AWARD OF THE DEGREE OF DOCTOR OF PHILOSOPHY IN ENVIRONMENTAL PLANNING AND MANAGEMENT IN THE DEPARTMENT OF GEOGRAPHY AND ENVIRONMENTAL STUDIES, UNIVERSITY OF NAIROBI
MAY, 2018 DECLARATION
ii
DEDICATION
To my Father, Simon Wambugu Mwangi, and my Mother, Sarah Wanjiku Mwangi, for their sacrifices and belief, even when everyone else was in doubt.
iii
ACKNOWLEDGEMENTS
I sincerely thank the Department of Geography and Environmental Studies, University of
Nairobi, for the support accorded to me during this study. Sincere gratitude also goes to the staff of Water Resources Management Authority, Kiambu Regional Office, for allowing me to access and use their long-term water quality monitoring data from the gauging stations of Ruiru and Ndarugu Rivers. I also thank the Ruiru-Juja Water and
Sanitation Company for allowing me access to their plant and providing me with valuable input during proposal development for this study.
This study was financed from two sources: first, from my employer, Karatina University
(KarU) through the “Research Grant for Academic Staff Programme”; and second, from the National Commission for Science, Technology and Innovation (NACOSTI) Grant for
PhD Studies. Much gratitude goes to the two institutions for providing the financial resources to facilitate the study.
I gratefully acknowledge my supervisors, Dr. Isaiah Nyandega and Dr. Shadrack Kithiia for their unequaled support since the conception of the idea, throughout the proposal development, their support while I sought funding, and during the thesis write-up. My wife,
Peris Wanjiku Mwangi, and my daughter, Claire Wanjiku Mwangi, endured long periods of my absence during data collection and gave me hope and strength during thesis write- up. The chairman of the Department of Geography and Environmental Studies, University of Nairobi, Dr. Samuel Owuor, has been very supportive to me. I would also like to thank the Kenyatta University Alumni Network for arranging access to Kenyatta University library.
iv
Dr. Chris George of Texas A & M University provided valuable technical advice with regard to the input files of the Soil and Water Assessment Tool (SWAT) model. My GIS advisor, Mr. John Musau was instrumental in introducing me to the workings of the SWAT model and debugging the errors encountered while running the model. I thank my research assistant, Mr. Ernest Ng’ang’a for enduring long days with me in the field. Special thanks to my family, who supported and encouraged me throughout the study.
v
TABLE OF CONTENTS
DECLARATION...... ii
DEDICATION...... iii
ACKNOWLEDGEMENTS ...... iv
LIST OF FIGURES ...... xi
LIST OF PLATES ...... xv
LIST OF TABLES ...... xvi
ABBREVIATIONS AND ACRONYMS ...... xvii
ABSTRACT ...... xix
CHAPTER ONE: INTRODUCTION ...... 1
1.0 Introduction ...... 1
1.1 Study Background ...... 1
1.2 Statement of the Problem ...... 6
1.3 Research Questions ...... 7
1.4 Objectives of the Study ...... 7
1.4.1 Specific Objectives ...... 7
1.5 Justification of the Study ...... 8
1.6 Operational Definitions ...... 10
CHAPTER TWO:LITERATURE REVIEW ...... 11
2.0 Literature Review...... 11
2.1 Introduction ...... 11
2.2 Theoretical Literature ...... 11
2.2.1 Land use change theories ...... 11
vi
2.2.2 The River Continuum Concept ...... 12
2.3 Empirical Literature ...... 21
2.3.1 Water as an Ecosystem Service ...... 21
2.3.2 Population Impacts on Watersheds ...... 23
2.3.3 Urbanization and Land use Change Dynamics ...... 26
2.3.4 Hydroclimatic Conditions of Tropical East Africa ...... 29
2.3.5 Water Quality impacts on Ecological Integrity ...... 37
2.3.6 Relationship between Land Use Patterns and Rivers’ Ecological Health ...... 42
2.3.7 Hydrologic modeling of Watersheds ...... 44
2.3.8 Comparison of GIS-based Hydrological Models ...... 48
2.3.8 Application of the SWAT Model in Kenya ...... 53
2.3.9 Practical Application of SWAT model in Kenya ...... 61
2.4 Conceptual Framework ...... 62
2.5 Research Hypotheses ...... 64
CHAPTER THREE: STUDY AREA ...... 65
3.0 Study Area ...... 65
3.1 Location ...... 65
3.2 Geology and Soils ...... 66
3.3 Topography ...... 69
3.4 Climate and Hydrology ...... 69
3.5 Flora and Fauna...... 71
3.5.1. Flora ...... 71
3.5.2. Fauna ...... 72
vii
3.6 Socio-economic Activities ...... 73
3.6.1. Population ...... 73
3.6.2 Economic Activities ...... 74
3.7 Land use and land Tenure ...... 75
CHAPTER FOUR: RESEARCH METHODOLOGY ...... 79
4.0 research methodology ...... 79
4.1 Study Design ...... 79
4.2 Data Types and Sources ...... 82
4.3 Pilot Survey ...... 84
4.4 Assessing Spatio-temporal Land Use Dynamics ...... 85
4.4.1 Classification Scheme ...... 87
4.5 Water Quality Assessment and Macro Invertebrate sampling ...... 88
4.5 Impacts of land use on the Hydrology of Ruiru and Ndarugu Rivers ...... 92
4.5.1 The Soil and Water Assessment Tool (SWAT) Model ...... 92
4.6 Data Preparation...... 97
4.6.1 Land cover ...... 97
4.6.2 Land topography data ...... 97
4.6.7 Soil data ...... 98
4.6.8 Climate Data ...... 100
4.6.9 River Discharge Data ...... 100
4.6.10 Model Evaluation, Calibration, Validation and Parameter Definition ...... 101
4.7. Data processing and Analysis ...... 102
4.8 Scope and Limitations of the Study ...... 103
viii
CHAPTER FIVE: RESULTS AND DISCUSSION ...... 105
5.0 result and discussion ...... 105
5.1 Land Use Types, Distribution and Changes ...... 105
5.2 Land Use Impacts and seasonal Variations on Surface Water Quality ...... 122
5.3 Watershed Modelling ...... 142
5.3.1 Precipitation and Stream flow ...... 142
5.3.2 Water Temperature ...... 145
5.3.3 Dissolved Oxygen ...... 146
5.3.4 Sediment ...... 146
4.3.5 Nitrite and Nitrate ...... 150
4.3.6 Land use impacts on Simulated Sediment, Phosphate and Nitrate ...... 152
CHAPTER SIX: SUMMARY OF FINDINGS, CONCLUSION AND
RECOMMENDATIONS ...... 156
6.0 summary of findings, conclusion and recommendations ...... 157
6.1 Research Findings ...... 156
6.2 Conclusion ...... 158
6.3 Recommendations ...... 160
REFERENCES ...... 162
ix
LIST OF APPENDICES ...... 191
Appendix I: MiniSASS Table for Interpretation of Ecological Condition Based On
Composition Of Macro Invertebrates In Rivers ...... 191
Appendix II: Long Term On-Site (In Situ) Water Quality Monitoring Data From Water
Resources Management Authority (Kiambu Regional Office) ...... 192
Appendix III: Long Term Water Quality Monitoring Lab Data From Water Resources
Management Authority (Kiambu Regional Office) ...... 194
Appendix IV: NEMA Water Quality Standards For Domestic Use ...... 196
Appendix V: SWAT Tables ...... 198
x
LIST OF FIGURES
Figure 2 1: Diagram of the river continuum concept depicting a river channel and riparian
vegetation as the river grows from a headwater stream to an eleventh-order
stream...... 14
Figure 2-2: Challenges Facing Water Resources Management...... 16
Figure 2-3: The Hydrologic Pathways Connecting the Landscape to Streams and Rivers.
...... 19
Figure 2-4: Relationship between population growth, economic growth and water
resources ...... 20
Figure 2-5: The Köppen–Geiger climate classification system updated with CRU TS 2.1
temperature and VASClimO v1.1 precipitation data for 1951 to 2000 ...... 33
Figure 2-6: Classification of the tropics based on the seasonal distribution of rainfall .... 35
Figure 2-7: Climatology of tropical Africa for four representative months is plotted on the
left, including precipitation rates (mm/month) ...... 36
Figure 2-8: Relationship between Natural State of Ecosystems and Altered State of
Freshwater Systems...... 40
Figure 2-9: A framework for ecologically sustainable water management...... 41
Figure 2-10: Location of SWAT-applied basins in Kenya...... 61
Figure 3-1: Map of the study area showing study basins relative to geographical features,
urban centers and infrastructure in Kiambu County...... 78
xi
Figure 4-1: Schematic Diagram showing the overall study design. Sampling sites were
selected from a multi-stage process described above...... 80
Figure 4-2: Map showing sampling sites in Ruiru and Ndarugu Watersheds, Kiambu
County...... 81
Figure 4-3: SWAT Model flow chart showing the inputs, outputs and process ...... 94
Figure 4-4: Altitude map generated from SRTM DEM used in the SWAT model ...... 98
Figure 4-5: Soil map showing extent of soils in the watersheds ...... 99
Figure 4-6: Location of weather and Gauging Stations ...... 101
Figure 5-1: Random points generated for use in combination with ground truth data for
accuracy assessment ...... 106
Figure 5-2: Land use change in Ruiru and Ndarugu Watersheds (2005 to 2015) ...... 109
Figure 5-3: Comparison of Urban and settlement in Ruiru and Ndarugu watersheds,
showing a steady rise in this land use type between 2005 and 2015...... 110
Figure 5-4: Dendrogram based on the hierarchical clustering algorithm, showing the
different clusters from the three time epochs (2005, 2010 and 2015). A=Large-
scale Agriculture; B= Small-scale Agriculture; C= Forest; D= Grassland; E=
Urban and Settlement; F= Water ...... 111
Figure 5-5: Land use change between 2005 and 2015 showing increasing urban areas and
reducing agricultural areas in Ruiru and Ndarugu Watersheds ...... 115
Figure 5-6: Land use and land cover characteristics of Ruiru and Ndarugu watersheds in
2005, 2010 and 2015 ...... 119
xii
Figure 5-7: Area under forest cover showing a decline between 2005 and 2015 in both
Ruiru and Ndarugu watersheds ...... 120
Figure 5-8: Sharp decline in areas under water, indicative of pressure on water resources
in the watersheds of Ruiru and Ndarugu ...... 121
Figure 5-9: Projections of land use changes for year 2020 for Ruiru and Ndarugu
watersheds ...... 122
Figure 5-10: Distribution of sampling sites along the altitudinal gradient (FD-Forest
Dominated, AD-Agriculture Dominated, UD-Urban Dominated)...... 124
Figure 5-11: Mean ± standard error values for the water quality parameters among three
land use-based site Groups in Ruiru and Ndarugu Watersheds ...... 125
Table 5-3: Summary of surface water quality between the dry and wet seasons in the
Ruiru and Ndarugu Rivers, Kiambu County ...... 126
Figure 5-12: PCA analysis of water quality parameters across land use systems in Ruiru
and Ndarugu basins, Central Kenya. The first two axis components 58.5% of
the variation: Component 1 (32.5%) and Component 2 (26%)...... 127
Figure 5-13: Discriminant analysis on measured water quality parameters across land use
types on two canonical axes showing water quality parameters and their
associated land use types...... 128
Figure 5-14: Performance of SWAT Model under recorded (2002 to 2012) and simulated
(2005 to 2014) ...... 142
Figure 5-15: Variation in OutFlow regime on Ruiru Basin as simulated by SWAT model
...... 143
xiii
Figure 5-16: Variation in OutFlow regime on Ndarugu Basin as simulated by SWAT
model ...... 144
Figure 5-16: Observed and Simulated Temperature records for Ruiru and Ndarugu Rivers
...... 1466
Figure 5-17: Simulated mean monthly dissolved oxygen for Ruiru (top) and Ndarugu
(bottom) Rivers ...... 146
Figure 5-18: Mean monthly sediment inflow in Ruiru River ...... 147
Figure 5-19: Mean monthly sediment inflow in Ndarugu River ...... 148
Figure 5-20: Simulated Nitrite and Nitrate levels in Ruiru and Ndarugu River ...... 150
Figure 5-21: Levels of organic phosphates, nitrates and sediment loading in agriculture-
dominated, forest-dominated and urban/settlement dominated sub basins. ... 154
xiv
LIST OF PLATES
Plate 4-1: Random Points Generated for Performing Accuracy Assessment for Image
Classification...... 86
Plate 4-2: Ground Verification (Ground Truthing) using a Global Positioning System .. 87
Plate 4-3: Measuring Water Quality Parameters (top) and Water Meters (Turbidity Meter
and DO Meter) ...... 91
Plate 5-1: High Resolution Imagery of 2003, showing large parts previously under sisal in
Juja (A) and stone quarrying B ...... 116
Plate 5-2: High resolution imagery showing the shift of urban land use replacing areas
previously under sisal in Juja (A)- changing to settlement, and (B) Stone quarry
expansion ...... 116
Plate 5-3: Coffee Plantations in Ruiru Watershed ...... 117
Plate 5-4: Small-scale farming in Ndarugu watershed ...... 117
Plate 5-5: The Ruiru River at Thika Road Bridge, showing urban land use and
agricultural land use (small scale) adjacent to the river...... 129
Plate 5-6: Streambank cultivation (foreground) and development (background) of the
Ndarugu River ...... 134
xv
LIST OF TABLES
Table 2-1: A comparison of four commonly used hydrological Models, their processes and process representation ...... 51
Table 2-2: Applications of the SWAT Model in Kenya ...... 54
Table 3-1: Population sizes in different Urban Centres in Kiambu County, showing a steady population increase since 2009 ...... 74
Table 4-1: Data types used in SWAT and their sources ...... 83
Table 4-2: Classification scheme for the classification and change detection procedures 88
Table 4-3: Description of sampling sites for water quality assessment ...... 90
Table 4-4: Soil Characteristics in Ruiru and Ndarugu watersheds ...... 99
Table 4-5: Parameters selected for water quality simulation in the SWAT model ...... 102
Table 5-1: Land cover classification accuracies computed from ground truth reference points over the 2015 maximum likelihood-classified image ...... 107
Table 5-2: Land use characteristics in Ruiru and Ndarugu Watersheds, showing areas in square kilometers and percentage in brackets ...... 108
Table 5-3: Summary of surface water quality between the dry and wet seasons in the Ruiru and Ndarugu Rivers, Kiambu County ...... 126
Table 5-4: Macro invertebrate richness at sampled sites ...... 131
Table 5-5: MiniSASS Sensitivity Scores for Ruiru and Ndarugu Rivers ...... 132
Table 5-6: Mean and maximum values of water quality parameters recorded in Ruiru and Ndarugu monitoring stations ...... 140
Table 5-7: Simulated vs recommended water quality parameters on Ruiru and Ndarugu Watersheds, Kiambu County ...... 155
xvi
ABBREVIATIONS AND ACRONYMS
AOI - Area of Interest
BOD- Biological Oxygen Demand
CFSR- Climate Forecast System Reanalysis
DEM- Digital Elevation Model
ETM- Enhanced Thematic Mapper
FAO- Food and Agricultural Organization
GIS- Geographical Information System
HRU- Hydrological Response Unit
ILRI- International Livestock Research Institute
ISODATA- Iterative Self-Organizing Data Analysis Technique
LULC- Land Use Land Cover
LULCCS- Land Use/Land Cover Classification System
IWRM- Integrated Water Resources Management
MSS- Multi Spectral Scanner
TM- Thematic Mapper
NACOSTI- National Commission for Science, Technology and Innovation
RDA- Redundancy Analysis
xvii
RCC- River Continuum Concept
SRTM- Shuttle Radar Topography Mission
UNEP- United Nations Environmental Programme
UNU-INWEH- United Nations University - Institute for Water Environment and
Health
USGS- United States Geological Survey
WRA- Water Resources Authority
WGS- World Geodetic System
SWAT - Soil and Water Assessment Tool
UTM- Universal Traverse Mercator
xviii
ABSTRACT
Land use dynamics are known to cause considerable modifications to the environment which can cause severe impacts to aquatic natural resources. This study addressed the impacts of land use change in the Ruiru and Ndarugu river basins in Kiambu County of Kenya. The objectives of the study were to determine the land use and land cover trends, implications of land use practices on physico-chemical and macro-invertebrate richness and to create an impact model for the assessment of land use effects on hydrological parameters in the Ruiru and Ndarugu river basins of Kenya. The hypotheses used to guide the study were that no change occurred in land use practices between 2005 and 2015 in the study basin; land use and land cover systems in the basins do not have significant impact on water quality and macro invertebrate composition and; simulated water quality parameters do not significantly differ from measured water quality parameters in the basins. The data used in the study included spatial datasets made up of satellite imagery for the study area sourced from Landsat 7 and 8 (TM and ETM), administrative boundaries sourced from Survey of Kenya, Soil data from Soil Survey of Kenya, digital elevation model sourced from United States Geological Survey (USGS), climate data from world weather database, all of which were secondary data. Other secondary datasets used in the study included rainfall and temperature data from Kenya Meteorological Department and long term water volume flow data from Water Resources management Authority (WRMA). The study also used primary data collected in situ within the basins on the physical-chemical water quality parameters (Water Temperature, Dissolved Oxygen, pH, Electrical Conductivity and Turbidity). The resulting datasets were used to assess the land use change dynamics impacts on water quality using the Soil and Water Assessment Tool (SWAT) model which is a spatial distributed GIS-based hydrological model. The land use change assessment indicated that a significant change in land use and land cover had occurred between 2005 to 2015, with both watersheds changing to a more urbanized landscape; land use and land cover systems had significant influence in water quality parameters and macro invertebrate composition in the two basins and; the SWAT outputs closely matched with the in situ water quality measurements. The study concluded that the changes in land cover and land use were mainly due to increasing population pressure, and land cover is projected to continue to change towards an urbanized landscape, land use and land cover changes contribute to changes in water quality parameters and; the SWAT model with local modifications is a suitable tool for monitoring and management of land use, water quality and macro invertebrate assemblages in the Ruiru and Ndarugu basins. It is recommended that the use of Intergrated Watershed Management (IWM) based on scientific approaches to be adopted by both the County and National governments; the use of ecosystem management techniques in basin management to be incorporated in land use policies (such as the adoption of blue-green infrastructure) and; the use of hydrological models such as SWAT in data-scarce regions must allow the modification that suits local conditions to avoid over or under-estimation of hydrological parameters.
xix
CHAPTER ONE 1.0 INTRODUCTION 1.1 Study Background
In 1972, the Stockholm Conference recognized the need for global environmental planning, during which the United Nations Environment Programme (UNEP) was born (United
Nations, 1994). The conference recognized that unsustainable pressures on water resources was causing concerns for long term sustainability and human life in the participating countries. The concept of sustainable development was born at this gathering, and has been adopted as a basis for the Human Environment (United Nations, 2005). Sustainable development is defined as “Development that ensures that the use of resources and the environment today does not restrict their use by future generations” (United Nations, 1994;
United Nations, 2005). In many parts of the world, water resources management has become a critical aspect with population and economies growing, which increases the demand for water (Global Water Partnership, 2005). This culminated in the need for allocation and conflict resolution for water resources, causing concerns for long term sustainability and human life observed in the participating countries.
Rivers are largely susceptible to land use change and ubiquitous exploitation (Withers and
Jarvie, 2008; Vörösmarty et al., 2010). Deterioration of rivers in terms of water quality as a result of unsustainable human activities has become a major environmental concern
(Chen and Lu, 2014), which are directly reflected in land use and land cover characteristics
(Kang, et al., 2010). To make meaningful decisions for effective water quality management, there is need to understand the relationship between land use and water
1 quality as this relationship can be used to target critical land use areas and to institute appropriate measures to minimize pollutant loading in water resources (Abler et al., 2002).
Human well-being is fundamentally dependent on ecosystems for the provisioning, cultural and regulating services that they provide (World Bank, 2007). Clean water is one of the critical resources provided to man by ecosystems. As such, the ecosystem concept has been elevated as a fundamental attribute for human development. Maintenance and access to ecosystem services has consistently been associated with better health and economic outcomes.
Water resources degradation result from both ‘traditional’ forms of pollution and broader pressures on ecosystems, ranging from depletion and degradation of freshwater resources, to the impacts of natural disasters, agricultural production and climate change (Corvalan et al., 2005). The potential for emergence of new livelihood impacts is much greater with ecosystem pressures, including impacts on fresh water supply, local food yield, impaired vitality of ecosystems, and loss of livelihoods.
Population expansion in Kenya is closely associated with a massive increase in demand for land, which is highly related to urban growth and increased agricultural activities. Kenya is classified as one of the water-scarce countries in the world (Global Water Partnership,
2015), and land use implications on water systems have been shown to cause far-reaching consequences, both ecologically and economically. The demand for land in the region is largely fueled by agriculture and settlement. The high population expansion and intense land utilization in the catchment are attributed to increased land degradation, leading to degradation of water resources, one of the critical ecosystem services. This ultimately leads to compromised water quality in rivers. Additionally, increased demand for food leads to
2 intensive farming practice and increased destruction of forest cover to open up areas for cultivation.
The central region of Kenya contains two of the five ‘water towers’ in Kenya, namely
Aberdare Ranges and Mt. Kenya (others are Cherangani Hills, Mt. Elgon and Mau Forest complex). These two catchment systems contribute to the Athi River Drainage System, one of the most important river systems in the country (WRI et al., 2007).
Measured patterns in surface water parameters can provide unique insights into the hydrological functioning of catchments (Frohlich et al., 2008). It has been shown that interactions between terrestrial water and the catchment play an important role in regulating the physical, chemical and biological composition of the water (Stumm and
Morgan, 1996). Measuring water quality is extremely important in determining a watershed's health. Water-quality experts determine healthy levels of chemicals and microbes in stream water for a particular use, including drinking, recreating, irrigating, or supporting fish. These safe levels have been adopted by governments as water-quality standards. If water quality in a stream is measured and compared to water quality standards, the health of the stream as it relates to that particular use can be determined. For example, drinking water must be free of pathogens; and other harmful chemicals should be below the set standards. Temperature, metals, dissolved oxygen and chemical contaminants must be at levels that do not harm fish and other stream-dwelling organisms. Stream health assessment addresses six main water quality components: sediment, dissolved oxygen, nutrients, temperature, bacteria, and toxic chemicals.
Sediment is delivered to a stream through the erosion of upland areas as well as from the stream's banks. The implications of too much sediment are that it could cover gravel on a
3 streambed, which may suppress eggs deposited by spawning fish. It could also burry gravelly substrates needed by aquatic insects. The amount and size of sediment in a stream helps to determine whether there are excess levels, and therefore determine where it originates from (Allaby and Allaby, 1999).
Nutrients are typically at low levels in natural fresh water. Excessive nutrient concentration can lead to too much growth of aquatic plants. This may lead to occurrence of decaying vegetation which may lead to reduction of dissolved oxygen needed by fish to survive.
Further, excessive plant material can interfere with human uses through obstruction of water supply courses, or by obstructing waterways used for recreation. Nitrogen and
Phosphorous are the major nutrients that cause this plant growth in water and they should be monitored to monitor whether their levels might stimulate the excessive plant growth
(Alexander et. al., 2000).
Bacteria load in water determines the sanitary condition of the water. The presence of
Coliform bacteria in stream water is used as an indicator of the sanitary quality of the water for drinking and swimming, while fecal coliform bacteria are good indicators of human- caused pollution. This is because they originate from the gut of warm-blooded animals, including humans. They can originate from sewage, animal feedlots, pastureland, and cities. The presence of fecal coliform bacteria in streams is an indication that disease- causing organisms may be present.
Temperature determines the amount of dissolved oxygen in the water. If the water is too warm or the oxygen level is too low, most fish will not survive. Cold temperatures and high dissolved oxygen levels are critically important to fish (Russell, 1999).
4
Chemicals are used in the day to day activities, pest control, including manufacturing, normal agriculture, household use, forestry, and city operations. Although they are intended to be safely handled and controlled, they sometimes find their way into streams.
Chemical can cause catastrophic impacts to streams, particularly to biodiversity. Although low levels of toxic chemicals can have fewer catastrophic effects to biodiversity, they can be equally dangerous to humans who consume the water or fish from the water body
(Stumm and Morgan, 1996).
Aquatic organisms respond to deviations in stream quality and water quantity. Both chemical and biological parameters can be used to evaluate the water quality in a stream and its ability to support a thriving aquatic community. The presence and density of vegetation near a stream has a strong influence its health; and the type and size of this riparian vegetation is one of the most important components of a stream health assessment.
Riparian vegetation influences fish habitat quality by acting as a filter to help keep sediment and other pollutants out of streams (Ritcher et al., 1997a). Anthropogenic disturbance of ecosystems and their services are assessed through land-use changes ecological processes (Xu et al., 2008).
5
1.2 Statement of the Problem
Land use practices develop over a long period of time under different environmental, political, demographic, and socio-economic conditions (Jackson and Jackson., 1998). The interaction of land use systems in a watershed and their implications on water quality is a complex phenomenon that encompasses a wide range of anthropogenic and natural processes. These conditions often vary and have a direct impact on ecological health of water resources and therefore have implications on the human population that relies on the water for consumption. Growing human populations exert increasing pressure on the landscapes as demand for resources such as food, water, shelter and fuel multiply (DeFries and Eshleman, 2004). Land use and land cover changes are therefore regarded as a central component in determining the approaches for managing natural resources (including rivers) and monitoring environmental changes globally. They also play a fundamental role in local, regional and national socio-economic development plans. Land-use and land cover change has been recognized as a major driving force of global environmental change and is central to the sustainable development issues (Thuo, 2013). As a result, monitoring and evaluating of sustainability indicators, such as change in land use and settlement, urbanization, loss of natural vegetation cover, pressure on land resources, ecosystems and their products (e.g. water) are required (Wunder et al., 2008).
Rapid human population growth coupled and associated with land use practices has implications on resource use and environment change (Liverman et al., 1998). Integration between landscape processes and water quality is vital for better understanding of critical water problems that might change drastically against the backdrop of emerging issues, such as Climate Change. Thus, this study sought to fill this gap and estimate the implications of
6 watershed degradation to water quality in the catchments of Ruiru and Ndarugu Rivers.
Further, the increase of large agricultural activities, including horticulture with the existing small-scale farms and settlements within the watershed has implications on the demand for water resource at the streams. The capacity of these basins to provide environmental services is seriously at risk because of increasing environmental degradation and over-use.
1.3 Research Questions
1. What are the land use and land cover trends in Ruiru and Ndarugu Watersheds
between 2005 and 2015?
2. Do land use and land cover systems have an effect on physical-chemical and
macro-invertebrate composition of Ruiru and Ndarugu Rivers?
3. What are the effects of land use systems on the hydrological processes in Ruiru
and Ndarugu watersheds?
1.4 Objectives of the Study
This study was aimed at revealing the linkages between land use patterns and the environmental integrity of water resources of a significant part of the Upper Athi River
Basin in Kenya with the goal of establishing the impacts of land use and land cover dynamics on the quality of water resources in the watersheds of Ruiru and Ndarugu Rivers.
1.4.1 Specific Objectives
The specific of this study were to determine:
i. Land use and land cover trends in Ruiru and Ndarugu Watersheds between 2005
and 2015.
7
ii. The impact of land use and land cover systems on physical-chemical and macro-
invertebrate composition of Ruiru and Ndarugu Rivers and. iii. To develop a land use and land cover simulation model for assessing the impacts
of land use on hydrological processes in Ruiru and Ndarugu watersheds.
1.5 Justification of the Study
As Kenya positions itself to become a middle-income economy as entrenched in its Vision
2030, there is need to put in place sustainability indicators to show the state of human, environmental and economic conditions. This will not only help to assess the trend of changes in these attributes but will also play an important role in achieving sustainable development. These indicators will also provide a framework under which crucial decisions to national and international policy will be made.
With the human population growing rapidly globally, the Earth’s land cover has been transformed tremendously, especially in developing countries. Kiambu County has experienced a steady human population growth, making it the third most populous county in Kenya (KNBS, 2016). As a result, environmental degradation and decline in both quality and quantity of environmental services has been experienced, leading to serious implications on human well-being. As such, issues such as sustainable development, pollution control, environmental change and other related matters of human-environment interaction have been a major concern by the scientific community and policy makers
(Codjoe, 2007).
According to Torey (1998), rapid land use degradation, especially in developing countries, presents one of the crucial factors that must be considered in the human dimension of the
8
21st century. Because of lack of basic knowledge of the landscape-level attributes to water quality problems as well as their impacts on water quality and ecological aspects, it has been impossible to assess, manage and restore limited water resources in Kenya. Therefore, timely and accurate estimation of implications of population systems to land use attributes is of considerable significance for decision makers in watershed planning and for a better understanding of the relationships between population growth, economic and environmental conditions (Yu and Changshan, 2006). Despite these dire needs to monitor ecosystems, the ecologic and health values of the catchment of Ndarugu and Ruiru Rivers, particularly within the settlement and agricultural areas are not yet entirely addressed.
Since ecosystems have been degraded worldwide leading to loss of valuable environmental services that they provide, there has been a growing search for practical solutions. Results of this study will provide vital information on threats facing these river systems, including nutrient enrichment and pollution from agricultural systems. The research sought to address the major driving forces for catchment change and degradation and associated land use and land cover changes in the upper catchment of these rivers. The study also sought to reveal the implications of these problems to the state of water resources that most of the population derives water from. Further, the study sought to identify processes for a multitude of water quality problems, including pollution by domestic sewage, industrial effluent, agro-chemicals and associated water quality aspects. This study was intended to provide information that can be used in developing policies for conservation and related natural resources by relevant public sectors such as Water and Irrigation, Agriculture,
Livestock and Fisheries, Planning, Development, Natural Resources and Environment.
9
1.6 Operational Definitions
Land use: The way in which humans use land, involving modification of natural
environment into built environment such as agricultural fields, pastures, and
settlements. It is the arrangements, activities and inputs people undertake in a
certain land cover type to produce, change or maintain it.
Land cover: the surface cover on the ground, including vegetation, urban infrastructure,
water, bare soil or other types of cover.
Water Quality: a set of conditions that define the state of a water body.
Watershed: an area of land draining all the rainfall and streams to a common outlet
Hydrologic Modeling: The use of use of physical models, mathematical analogues, and
computer simulations to characterize the likely behavior of real hydrologic
features and systems.
Spatio-temporal Dynamics: changes in land use or land cover in both space and time
10
CHAPTER TWO 2.0 LITERATURE REVIEW 2.1 Introduction
Theories and concepts related to land use and land cover dynamics, hydrological processes, water quality and watershed modelling are reviewed. Literature on topics related to spatial- temporal dynamics of watersheds is also presented.
The first part focuses on a theory describing river systems in terms of their physical, chemical and biological characteristics. The second part covers current trends in land use and watershed research in thematic areas as follows: Population impacts on watersheds,
Urbanization and Environmental Change, Land use Impacts on Watersheds, Macro invertebrates as indicators of rivers’ health and hydrological modelling. The third part presents the conceptual framework of the study.
2.2 Theoretical Literature
2.2.1 Land use change theories There are three main perspectives of understanding for the study of land use and land cover change analysis (LUCC, 1999). These are: the narrative, the agent-based, and the systems approach. The narrative perspective seeks to explain the LULC story with depth of understanding through historical detail and interpretation by providing an empirical and interpretative baseline from which the validity and accuracy of other visions are evaluated.
This method is suitable in identifying stochastic and random events that significantly affect land-use/land-cover changes but might be missed in approaches employing less expansive time horizons or temporal sampling procedures.
11
The agent-based perspective seeks to distill the general nature and rules of individual agents’ behavior in their daily decision making. There are many forms of such distillations, and they range from rational decision making of the average or typical actor in neo-classical economics to household, gender, class among others. In this perspective, significant attention is given to human agents in determining land-use decisions.
The systems/structures perspective, on the other hand, seeks to understand the organization and institutions of society that establish the opportunities and constraints on decision making (Ostrom 1990). These structures operate interactively at different spatial and temporal scales, linking local conditions to global processes and vice versa (Morán, Ostrom and Randolph, 1998). The systems or structures may manifest themselves in unforeseen and unintended ways. Some institutions are direct drivers of change, while others (such as markets) are intricately linked to individual decisions and affect these decisions and at the same time are the aggregate result of these decisions (LUCC, 1999). This study integrates all the tree approaches as recommended by LUCC (1999), where the described epistemological traditions are incorporated. First, observations and descriptions to understand land use change are adopted (i.e. inductive approach). Secondly, the use a model to understand land use impacts on water resources (i.e. deductive approach) is used.
Thirdly, a critical evaluation of land use change aspects is done to understand the drivers of land use change (i.e. dialectic approach) (LUCC 1999).
2.2.2 The River Continuum Concept Rivers are commonly understood as complex, living systems (Cosgrove and Petts, 1990;
Gregory, 2006) that are heterogeneous and dynamic across multiple scales of space and time. One of the theories that attempts to describe rivers’ functions across a landscape is
12 the River Continuum Concept (RCC) (Vannote et al., 1980). RCC attempts to explain the relationship between several river characteristics such as fluvial geomorphic processes, the physical structure, as well as hydrologic cycle to patterns of community structure and function, organic matter loading, transport, utilization and storage along the length of a river (Vannote et al., 1980, as indicated in Figure 2-1). According to this concept, stream order is related to an expression of the physical component of the river (Leopold et al.,
1964) and has influence of riparian vegetation, the status of trophy, load, transport, and the relative importance of functional feeding groups (shredders, collectors, etc.). Stream order, however, is not usually regarded as an important description of the physical environment
(Gregory and Walling 1973), and should be viewed only as an indication of the relative position of a stream reach within the entire running water system.
Because the concept has been developed for natural, unperturbed stream ecosystems, the
RCC does not qualify in all environmental scenarios, and this has been the subject of its criticisms by several authors. Some of the limitations of this theory include its application in (i) human-dominated landscapes with anthropogenic disturbances (ii) streams at high elevations and latitudes, xeric regions, and deeply incised valleys, (iii) tributaries entering the main stream have localized effects of varying magnitude depending on the volume and the nature of the input (Statzner and Higler, 1985).
The RCC is used today during environmental assessment of rivers by comparing the variation between an ideal river as described by the concept with observations from the field (Stout and Ben, 2003). Alteration of the watersheds through land use practices and other associated activities may lead to alteration of the biophysical characteristics of rivers.
This study investigates how such alterations affect water quality in rivers and alter
13 biophysical characteristics, which could ultimately alter the ideal river system as described by the RCC.
Figure 2 1: Diagram of the river continuum concept depicting a river channel and riparian vegetation as the river grows from a headwater stream to an eleventh-order stream.
Source: Johnson et al., (1995) adapted from Vannote et al., (1980).
14
According to Lee (1995), a river basin system is comprised of multiple components, including precipitation, floodplains, lakes and swamps etc. Because these components are interdependent, piecemeal approaches to river basin development and management have often failed to optimize the management outcome, resulting in ineffective water and land resource use, economic losses and environmental degradation. In addition, water is said to be a special commodity to manage because: a) all life and sectors of the economy depend on it; b) humans live in and with the hydrological cycle (that is, water is constantly being recharged, used, returned and reused) and c); both depend on one another (Global Water
Partnership, 2000). A delicate balance exists between water for livelihood and water as a resource as shown in Figure 2-2 (Global Water Partnership, 2000). The scenario described above is typical of Ruiru and Ndarugu basins, where water demand is fuelled by a multitude of needs (domestic, agricultural and industrial). Understanding how these needs influence available water resources in Ruiru and Ndarugu watersheds will provide insights on the impact of water resources status on the economy, the interaction between human activities and water resources, the mechanisms required to sustainably manage the water resources and the hydrological processes that may be altered.
15
Figure 2-2: Challenges Facing Water Resources Management.
Source: Global Water Partnership (2000)
Due to this interdependence, there is need for integration of the various components of water resource use (Global Water Partnership, 2000). Water experts have recognized the importance of integration of the natural system between land and water use; between surface water and groundwater; between water quantity and quality; between the freshwater system and coastal waters and between upstream and downstream on one hand and, on the other hand, integration in our management of the natural system and calls for
Integrated Water Resources Management (Global Water Partnership, 2000).
16
Every important aspect of the river ecosystem, the river geomorphic system, and the river chemical system begins in headwater streams (Winter, 2007). Headwater streams clearly dominate surface water drainage networks (Freeman et al., 2007). Definitions of headwater streams vary, but they can generally be defined as all first- and second-order streams.
Stream order was originated by Horton (1945) and slightly modified by Strahler (1964).
First and second order streams compose over two-thirds of the total stream length in a river network (Leopold et al., 1964). A first-order stream is an intermittent or perennial stream with no temporary or perennial tributaries, while a second-order stream is created by the confluence of two first-order streams; both typically feeding larger rivers (Leopold et al.,
1964). Stream order defines stream networks, which is used to profile the geomorphology of a watershed. The streams in the Ruiru and Ndarugu basins are characterized as first and second order streams, and therefore their physic-chemical characteristics are expected to be highly influenced by seasonal variations.
Longitudinal connections within riverine ecosystems have long been recognized by both aquatic and terrestrial ecologists, as illustrated by the widespread use of the term river corridor in the literature; but “connectivity” did not emerge in the freshwater literature until the early 1990s (Amoros and Roux, 1988). Freshwater ecologists frequently use the term connectivity to describe spatial linkages within rivers (Ward and Stanford 1982; Ward,
1997; Amoros and Bornette, 1999; Wiens, 2002). Ward (1997) defines riverine connectivity as energy transfer across the riverine landscape.
In contrast to riverine connectivity, hydrologic connectivity (Pringle, 2003) encompasses broader hydrologic connections, beyond the watershed, on regional and global scales.
Hydrologic connectivity refers to the water-mediated transport of matter, energy, and
17 organisms within or between elements of the hydrologic cycle (Pringle, 2001), in essence combining the hydrologic cycle with riverine connectivity. Hydrological connectivity is essential in maintaining ecological integrity of ecosystems, where ecological integrity is defined as the undiminished ability of an ecosystem to continue its natural path of evolution, its normal transition over time, and its successional recovery from perturbations
(Westra et al., 2000). Conversely, hydrologic connectivity also directs and facilitates the flow of exotic species, human-derived nutrients, and toxic wastes in the landscape.
Hydrologic connectivity at large scales is a formidable concept because of the inherent complexity of water movement within and between the atmosphere, surface-subsurface systems and the ocean (e.g., Winter et al., 1998); and the extent and intensity of human alterations (Pringle and Triska, 2000).
Scientific concepts of connectivity differ from legal definitions. Hydrologists view connectivity as a continuum because the entire landscape is hydrologically connected
(Figure 2-3). Moreover, biological connections among water bodies are not restricted to pathways of water flow; e.g., migratory birds, amphibians, and winged aquatic insects travel across watershed boundaries.
Environmental effects of altered nutrient transport in rivers due to watershed changes have emerged are important and suggest that the current extent and magnitude of hydrologic alterations and pollutant loading will result in new, perhaps unexpected, environmental problems, and raise questions of the larger scale effects of other alterations in hydrologic systems (Pringle, 2003). These interactions are have been described by Global Water
Partnership (2000) in Figure 2-4.
18
Figure 2-3: The Hydrologic Pathways Connecting the Landscape to Streams and Rivers.
Source: Freeman et al., (2007)
According to Global Water partnership (2000), Integrated Water Resources Management
(IWRM) is defined as a “process that promotes the coordinated development and management of water, land and related resources in order to maximize the resultant economic and social welfare in an equitable manner without compromising the sustainability”. It recognizes four aspects of successful water resources management:
These are: a) mainstreaming water in the national economy; b) ensuring co-ordination between sectors; c) ensuring partnership between public and private sector management;
19 and d) involving all stakeholders. The IWRM framework emphasizes using of water resources in a manner that does not compromise its sustainability. In this respect, analyzing land use dynamics in Ruiru and Ndarugu basins will inform information for successful implementation of IWRM by providing the link between land use systems and their influence on water resources.
Figure 2-4: Relationship between population growth, economic growth and water resources
Source: Global Water Partnership (2000).
20
Figure 2-4 depicts the relationship between human population growth, the economy of an area and water resources. While the water resources remain relatively constant, the human population is steadily growing, and this is coupled by the growth of the economy.
Population and economic growth are also associated with increased pollution due to an increase in anthropogenic activities in the watersheds (agriculture, increased deforestation, industrialization and urbanization). This situation is characteristic of many watersheds in
Kiambu County, including Ruiru and Ndarugu watersheds. The end result is an increase in competition for water resources, which may lead to conflict due to water scarcity.
2.3 Empirical Literature
2.3.1 Water as an Ecosystem Service
It is estimated that humanity consumes 1000–1700 m3 of the world’s surface and groundwater resources per year; constituting between 22% and 150% of the annual global supply of fresh water; making sustained fresh water supply one of the most critical ecosystem services in the world (Hoekstra and Wiedmann, 2014). The notion of ecosystem services began to be developed in the late 1960s (King 1966). The concept stems from the recognition that human life depends on the existence of a finite natural resource base, and that nature contributes to the fulfillment of human needs (Meadows et al., 1972). In the year 2000, the Secretary-General of the United Nations called for a worldwide initiative aimed at assessing the consequences of ecosystem change for human wellbeing and the scientific basis for action needed to enhance the conservation and sustainable use of those systems, dubbed the “Millennium Ecosystem Assessment” (Millennium Ecosystem
Assessment, 2003). Ecosystem services are defined as ‘the benefits that people obtain from ecosystems’, and it has been emphasized that there is need to incorporate the value of
21 ecosystem services into decision-making to reverse increasing degradation of ecosystems
(Millennium Ecosystem Assessment, 2005). The publication of the Millennium Ecosystem
Assessment in 2005 was followed by other global and national level publications aimed at understanding the management of natural resources based on the concept of ecosystem services (Ojea et al., 2012). For example, there have been ongoing discussions about how to incorporate ecosystem services in river basin planning cycles within the Common
Implementation Strategy of the European Water Framework Directive (Martin-Ortega,
2012). Water ecosystem services derive directly from the world’s freshwater ecosystems, the suite of which includes rivers, lakes, floodplains, and an enormous range of wetlands as well as their adjoining riparian areas (Millennium Ecosystem Assessment 2005).
Water provisioning has been identified as one of the ecosystems services threatened by land use systems because they modify or degrade watersheds (Millennium Ecosystem
Assessment, 2005). One way to curb the negative effects of land use systems on the ability of ecosystems to provide water is to adopt the “Ecosystem Approach”. This approach has been adopted by the Convention on Biological Diversity (2000) as a framework for action and ‘a strategy for the integrated management of land, water and living resources that promotes conservation and sustainable use in an equitable way’. The framework is based on the application of Malawi Principles, which are explicit and prescriptive characteristics of this framework for action. The Ecosystem Approach considers humans as an integral part of ecosystems, and recognizes that the way they use land affect the integrity of water resources in rivers, streams and other surface water sources. Both the Ecosystem Approach and ecosystem services-based approaches advocate for deliberate efforts on various forms
22 of knowledge in natural resource management, with a strong emphasis on the effects of land use systems on rivers (Scott et al., 2014).
2.3.2 Population Impacts on Watersheds Finite land and water resources are experiencing increasing human population for food production, urban expansion and the need for infrastructure facilities, giving rise to concerns on sustainability due to conflicting demands (Biswas et al., 2002). In addition, anthropogenic land use changes tend to result in various geomorphic and hydrologic changes (Grey et al., 2014). A river’s ecological health is of paramount importance because it reflects the status of the land surrounding it and indicates the potential impact of practices within the watershed (particularly upper watershed management areas) (Ferreyra and
Beard, 2007). These include changes in the spatial and temporal aspects of flood peaks, and in the extent and type of soil erosion (Magilligan and Stamp, 1997).
Environmental, ecological and human well-being are interdependent (Wilson, 1998). The interconnection among the environment, ecosystems, and human activity is a complex and poorly understood phenomenon and has been the subject of numerous publications and meetings (Aron and Patz, 2001). Local, regional, and global environmental changes, traceable to population growth and development patterns, as well as human activities that move populations within geographic regions, have led to increasing use of natural resources and degradation of ecosystems. This is especially problematic when those ecosystems perform natural services (Daily, 1997), such as purification of environmental media. Most emerging human diseases are driven by human activities that modify the ecosystems or otherwise spread pathogens into new ecological niches (Taylor et al., 2001).
Human activities through land-use practices generate both ‘positive’ benefits (increases in
23 food and fiber production) and ‘negative’ costs (species' extinction, soil erosion, land degradation, water pollution, and global warming). The pace, magnitude, and spatial reach of land-cover and land-use change have increased over the past three centuries, particularly over the last three decades, as a result of human activities, and may go beyond the ecosystem's recovery capacity (Lambin and Geist, 2006).
Land use affects human well-being directly and indirectly. It affects fauna and flora, contributes to local, regional, and global climate changes and is the primary source of soil, water and land degradation (Sthiannopkao et al., 2007). Converting forest land to agriculture, for example, may result in loss of biodiversity and changes in climatic patterns at the local and regional scales. On the other hand, mining and industrial activities that emit greenhouse gases may contribute to climate change at the global level. Altering ecosystem services-i.e., the provisions people obtain from ecosystems (e.g., food, water), regulating services (e.g., predator-prey relationships, flood and disease control), cultural services
(e.g., spiritual and recreational benefits), and support services (e.g., pollination, nutrient cycling, productivity) that maintain the conditions for life on Earth affects the ability of biological systems to support human needs (Vitousek et al., 1997). Alterations lead to large scale land degradation, changing the ecology of diseases that influence human health and making it more vulnerable to infections (Collins, 2001).
Risks to human health are increased also by toxicological risks resulting from bioaccumulation of toxic substances through global and regional environmental degradation (Darnerud, 2003) and disease outbreaks resulting from disruption of species' dynamics in disease control (Berrang, 2006). Such changes in part determine the vulnerability of coupled human-environment systems to climatic, economic or
24 sociopolitical perturbations (Turner et al., 2003). Land-use decisions are therefore, human health decisions (Turner et al., 2003).
In developing countries, urbanization presents an important dimension of economic, social and physical change (UNCHS, 2001). The urban population in Africa is expected to double by 2025 (Hall and Pfeifer, 2000), thereby driving the demand for urban land for housing and other urban uses. The outcome of this demand is likely to extend to the rural-urban fringe (Tacoli, 2002). With the expansion of the city, the rural-urban fringe presents challenges and opportunities while dealing with the byproducts of land use changes.
Urbanization of fringe areas provide a number of opportunities, including employment, education, better housing, knowledge and technology transfer, and ready markets for agricultural products. However, a human population influx exerts enormous pressure on natural resources, existing social services and infrastructure (Rees and Wackernagel,
1994).
Changes in land use have potentially large impacts on water resources (Stonestrom et al.,
2009). Due to rapid socio-economic development, land use changes occur, and may include changes of land use classes, e.g., conversion of cropland to urban area due to urbanization, as well as changes within classes such as a change of crops or crop rotations (Wagner et al., 2013). Land use changes could result in an increase of water shortage and therefore contribute to a deterioration of living conditions in water-scarce countries (DeFries and
Eshleman 2004). Land use and land-cover changes have impacts on a wide range of environmental and landscape attributes including the quality of water, land and air resources as well as ecosystem processes and functions (Lambin et al., 2006). Traditional catchment-scale water quality assessments are based on monitoring chemical and
25 biological indicators at selected sites to observe long-term trends and develop causal relationships with land-use practices. Rivers in a watershed play a major role in assimilating or carrying off runoff from agricultural landscapes, as well as municipal and industrial wastewater (Singh et al., 2005). When investigating catchment impacts to water quality, landscape heterogeneity has to be considered. This is important because spatial heterogeneity of catchment characteristics results into an increased number of environmental influences, such as human impacts within the catchment (Meybeck, 2002) and leads to the potential for large variations in runoff chemistry, both in time and in space
(Shand et al., 2005). Recording these hydro-chemical patterns can reveal information on the catchment wide controls on stream water chemistry and can help to conceptualize catchment understanding at the mesoscale, and this approach is used in this study to gauge the health of Ruiru and Ndarugu watersheds.
2.3.3 Urbanization and Land use Change Dynamics
Globally, urbanization has been an important social and economic phenomenon (Deng et al., 2009). Urbanization has been identified as one the most widespread anthropogenic causes of the loss of arable land (Lopez et al., 2001), habitat destruction (Alphan, 2003), and the decline in natural vegetation cover. The urbanization process is showing no signs of slowing down and has been identified as the most powerful and visible anthropogenic force that has brought about fundamental changes in land use and landscape pattern around the globe (Deng et al., 2009). The transition of rural areas into urban areas through development is occurring at an unprecedented rate in recent human history, leading to a marked effect on the natural functioning of ecosystems (Turner, 1994). According to Yu and Changshan (2004), rapid urbanization, especially in the developing world, will
26 continue to be one of the crucial issues of global change in the 21st century affecting the human dimensions. Urban areas cover only 3% of the Earth’s land surface currently, but they have noticeable effects on environmental conditions at both local and global scales
(Liu and Lathrop, 2002), including climate change (Grimm, et al., 2000). In China, for example, urbanization has resulted in an unprecedented scale and rate of urban expansion over the last two decades, with urbanization levels predicted to reach 50% and an urban population of 1.5 billion by the end of 2020 (Yu and Changshan, 2004). Due to rapid urbanization and exploitation of natural resources, there has been a significant impact to ecosystem structure, function and dynamics, leading to fragility of the urban regions of the world (Chen et al., 2005).
Ecosystems in the urban areas are greatly influenced by anthropogenic activities, and therefore considerably more attention is currently being directed towards monitoring changes in urban land use and land cover (Stow and Chen, 2002). In the economic- developed regions, this scenario is even more serious because dramatic urban expansion and land use change have induced serious environmental issues that threaten sustainable development of urban areas (Liu et al., 2007). The World Bank (2007) states that in developing countries, LULC change due to human activities is currently proceeding quicker than in the developed world, with most of the world’s mega cities projected to be in developing countries by the year 2020. The effect of population expansion is particularly of concern because the global urban population is projected to almost double by the year
2050 (Liu et al., 2007). In developing cities, increasing urban activities has caused rapid changes in LULC, which has led to increased environmental degradation (Holdgate, 1993).
The lack of accurate and timely information on the urbanization process has attracted
27 increasing attention from local communities and policy decision makers alike, especially due to the lack of basic knowledge and timely information of the urbanization process and its long term ecological impacts (Chen et al., 2005). Scientists have not been able to assess and analyze consistently, much less to manage and restore, the urban ecosystems in both urban cores and suburban fringes (Donnay et al., 2001). This could be due to the complexity of the urban cores-urban fringe interface, the dynamic nature of socio- economic systems in the interface as well as lack of reliable information sources that can provide meaningful results.
Remote sensing represents a major, though still under-utilized, source of urban information by providing spatially consistent coverage of large areas with both high spatial detail and temporal frequency, including historical time series (Jensen and Cowen, 1999). Increased availability and improved quality of multi-spatial and multi-temporal remote sensing data and new analytical techniques makes it possible to monitor and analyze urban expansion and land use change in a timely and cost-effective manner (Yang et al., 2003). There are however some technical challenges in retrieving accurate information of urban expansion and land use changes. These include the quality of remotely sensed data, the spatial and temporal resolution of data, computational and analysis procedures and human and logistical capacity. In this study, these challenges are overcome by using data from Landsat programme with less than 5% atmospheric noise. The data is gathered at 30m spatial resolution and 16-day temporal resolution, making it reliable to analyze land use changes over a 15-year period. Several image processing tools are available for its analysis, and its configuration makes it easy to process it in regular computers.
28
A major difficulty in urban remote sensing analysis is caused by the high heterogeneity and complexity of the urban environment in terms of its spatial and spectral characteristics
(Herold et al., 2005). Successful implementation of remote sensing requires adequate consideration and understanding of these specific urban landscape characteristics in order to fully explore the capabilities and limitation of remote sensing data and to evaluate appropriate image analysis techniques (Xiang, 1995; Herold et al., 2005). The most important technical concern with regard to provision of adequate remote sensing information is the pursuit of finer spatial resolutions of image pixels (Lu et al., 2004).
Besides, image analysis techniques are also very important and sometime present technical barriers for its urban applications, in addition to land use change detection methods and choice of data to be used (Lu et al., 2004). The most commonly used change detection methods are image differencing, principal component analysis (PCA) and post classification as they demonstrate better performance compared with other methods
(Herold et al., 2005). Other image analysis techniques, albeit less known and utilized, include the application of spatial metrics, which can be a useful tool for objectively quantifying the structure and pattern of an urban environment directly from thematic maps
(Herold et al., 2005). Spatial metrics are popularly used in landscape ecology due to their capability to detect changes of landscape pattern while quantifying, categorizing and revealing complex landscapes into identifiable patterns and ecosystem properties that are not directly observable (Weng, 2007).
2.3.4 Hydroclimatic Conditions of Tropical East Africa
The hydroclimate in tropical Africa is associated with ocean-atmosphere processes from both the Indian and Atlantic Oceans (Camberlin et al., 2001). However, in the easternmost
29 zone of tropical Africa, hydroclimate is insulated from the direct influence of the Atlantic
Ocean by orography (Sepulchre et al., 2006), and is largely responsive to Indo-Pacific climate dynamics. Specifically, the El Niño-Southern Oscillation (ENSO) and the Indian
Ocean Zonal Mode (IOZM) have a strong influence on modern-day rainfall in easternmost parts of Africa during the October-November- December rainy season (Black et al., 2003).
Despite being climatically diverse, the hydrology of tropical regions of the world is relatively little studied (Giertz and Diekkruger, 2003). McGregor and Nieuwolt (1998) note that this region should be subdivided into a number of sub regions, East Africa being part of it. The term tropics is defined as “a word derived from the Tropics of Cancer and
Capricorn, the parallels 23.5o, which indicate the outer limits of the area where the sun can be in a zenith position” (McGregor and Nieuwolt, 1998) . McGregor and Nieuwolt (1998) observe that these boundaries are too rigid because latitude is not the only factor in climate and major climates frequently deviate from parallels.
The climate of East Africa is typical of equatorial regions, influenced by a combination of the region's high altitude and the rain shadow of the westerly monsoon winds created by the Rwenzori Mountains and Ethiopian Highlands (Dewar and Wallis, 1999). Some parts of East Africa experience prolonged drought periods, such as the coast of Somaliland and
Puntland which may experience many years without any rain (Dewar and Wallis, 1999).
In other parts, rainfall generally increases towards the south and with altitude, being around
400 millimetres at Mogadishu and 1,200 millimetres) at Mombasa on the coast. Towards the inland, rainfall increases from around 130 millimetres at Garoowe to over 1,100 millimetres at Moshi near Kilimanjaro. Most of the rain falls in two distinct wet seasons, one centred on April and the other in October or November. This distinctive rainfall pattern
30 is attributed to the passage of the Intertropical Convergence Zone across the region in those months, but it may also be analogous to the autumn monsoon rains of parts of Sri Lanka,
Vietnam and the Brazilian Nordeste (Dewar and Wallis, 1999).
To the west of the Rwenzoris and Ethiopian highlands, the rainfall pattern is typically more tropical. Rainfall in this part occurs throughout the year near the equator and a single wet season in most of the Ethiopian Highlands (June to September), and contracts towards July and August around Asmara. This area experiences higher rainfall, ranging from 1,600 millimetres on the western slopes to around 1,250 millimetres at Addis Ababa and 550 millimetres at Asmara. Higher parts can experience over 2,500 millimetres (Dewar and
Wallis, 1999).
In East Africa, rainfall is also influenced by El Niño events, which tend to increase rainfall except in the northern and western parts of the Ethiopian and Eritrean highlands, where they produce drought and poor Nile floods (Davies, 2002). There are moderate Maxima
Temperatures in East Africa except on the hot and humid coastal belt which records 25 °C and minima of 15 °C at an altitude of 1,500 metres.
The tropics are characterized with a diverse range of environments and climates, showing climatic extremes between the tropics of Cancer and Capricorn (Savage et al., 1982). Some of the wettest regions on Earth are found here (e.g. the rainforests of central Congo basins and western Amazon) as well as some of the driest (e.g. the Atacama Desert of northern
Chile and Peru). The tropics however show limited seasonal fluctuations in insolation and temperature, with differences occurring with regard to quantity and temporal distribution of available moisture for regional and seasonal variability (Savage et al., 1982).
31
Several authors have attempted to classify tropical climates, but most classifications are based on meteorological parameters, mainly rainfall and temperature. For example,
Reading et al., (1995) provided an overview of the various attempts to define the climates of the tropics. The classic Köppen–Geiger system (Figure 2-5), for example, centres on the concept that natural vegetation is the best expression of climate, with climate zone boundaries described in relation to vegetation distribution. The Köppen–Geiger system combines average annual and monthly temperatures and precipitation, and the seasonality of precipitation. In this system, tropical climates are described as those exhibiting a constant high temperature (at sea level and low elevations), with all 12 months of the year having average temperatures of 18 °C or higher (Köppen, 1936). This classification, excludes cooler highland regions (defined as areas above 900 m elevation), which comprise around 25% of the total land area within the tropics (Reading et al., 1995). These regions still receive high amounts of solar radiation and do not have a pronounced winter season, but temperatures may be sufficiently depressed to affect biological activity (Reading et al.,
1995).
32
Figure 2-5: The Köppen–Geiger climate classification system updated with CRU TS 2.1 temperature and VASClimO v1.1 precipitation data for 1951 to 2000
Source: Kottek et al., (2006)
33
Köppen (1936) used rainfall levels and the seasonal distribution of precipitation to subdivide tropical climates into tropical rainforest (Af), tropical monsoon (Am), and tropical savanna climates (Aw), and then includes a range of other climate types within the tropics sensu stricto.
These include tropical and subtropical steppe (BSh), tropical desert (BWh) and humid subtropical climates (Cfa, Cwa). Some highland areas within the tropics also exhibit a temperate climate with dry winters (Cwb). The Köppen–Geiger classification systems provides predictions for various types of native vegetation that might be expected in expected in a given area, and therefore informs about land cover systems in human-unmodified areas.
From an Agricultural perspective, Jackson (1989) split the tropics into three zones (Humid,
Wet and Dry, and Dry) according to the level and seasonal distribution of rainfall (Figure 2-6).
This classification recognizes the importance of seasonality for agricultural productivity, and is less focused on natural vegetation zones compared to the Köppen–Geiger scheme.
Agriculture is a major contributor to land use and land cover change in the tropics, including the study area. This classification is therefore a better predictor of areas that are more likely to have changed in terms of land use and land cover as a result of high agricultural potential.
In addition to the classification of climates explained above, there are other attempts to classify climates within the tropics. These include classifications based on hydro-meteorology, with climate types defined according to the balance of precipitation inputs and evapotranspiration outputs. For example, Garnier (1958) differentiated humid tropical climates according to the number of months in which actual evapotranspiration equals potential evapotranspiration. In another classification, an aridity index for categorizing dry areas in tropical climates has been employed, which uses the ratio of precipitation to potential evapotranspiration (Thornthwaite,
1948).
34
Figure 2-6: Classification of the tropics based on the seasonal distribution of rainfall
Source: Jackson (1989).
35
Figure 2-7: Climatology of tropical Africa for four representative months is plotted on the left, including precipitation rates (mm/month)
Source: Modified from Kalnay et al., (1996).
36
2.3.5 Water Quality impacts on Ecological Integrity
The ecological health of rivers and streams is a fundamental and increasingly important aspect of water management globally (Schofield and Davies, 1996). Although the definitions of river health are subject to considerable debate (Wicklum and Davies, 1995), biologists use biodiversity (e.g. Magurran, 1988) and the structure of benthic communities
(Wright, 1995) as a measure of rivers’ ecological health. Others advocate a focus on the distribution and abundance of specific taxonomic groups such as diatoms (Reid et al.,
1995) and aquatic plants and fish (Whitton and Kelly, 1995). Species distribution and abundance are important elements in understanding river health but are often not sufficient to an understanding of how a system works (Whitton and Kelly, 1995), and therefore should not be used as the sole consideration. Patterns of species richness and abundance are used as surrogate measures of these fundamental processes (Bunn, 1995). In this study, the structure of macro invertebrate communities (Wright, 1995) was used as a measure of rivers’ ecological health using sensitivity indices as it was deemed the most appropriate for this type of study as explained in section 4.5.
Ecosystem-level processes are ideal measures of stream health because they provide an integrated response to a broad range of catchment disturbances, including increased nutrient concentrations from diffuse or point sources, changes to the quantity and composition of organic carbon inputs, alterations to the light regime (e.g. riparian shading and turbidity), and sedimentation (Minshall et al., 1985). In addition, ecosystem measurements integrate these factors at a large spatial scale and allow the health of streams and rivers to be viewed in a catchment context (Frissell et al., 1986).
37
In most countries, point and non-point source pollution are the major environmental problems affecting water quality. Poor agricultural practices, particularly on watersheds, exacerbated the situation, along with lack of or scarcity of treatment for domestic wastes
(Dudgeon, 1992). In East Africa, land use changes are caused by clearance of forests to create room for agriculture and rapid urbanization and have emerged as major stressors of streams and rivers (Kibichii et al., 2007). Surface water quality strongly depends on the hydrologic and hydro-chemical conditions in the drainage basin (Faure, 1998). A variety of conditions have been used to analyze these relationships, including chemical modelling on the watershed scale focusing on catchment response to acidification and mass balance studies of chemical weathering (Bassett, 1997).
In Kenya, degraded water quality, losses of biodiversity and altered hydrography have been recorded among streams and rivers (Ndaruga et al., 2004). On the other hand, deforestation and cultivation have been found to cause an increase in water temperature, conductivity, total suspended and dissolved solids and turbidity (Kibichii et al., 2007). Animals along riparian areas overgraze, trample on vegetation, defecate and urinate near or in the streams, which has been found to increase ammonia and nitrite as a consequence of increased run- off of animal wastes into streams (Kibichii et al., 2007). Near stream human activities like sand mining, bathing, laundry and row crop agriculture have been reported to cause the greatest influence on stream habitat and biotic characteristics (Raburu et al., 2009).
In the upper catchments of Athi River, streams and rivers serve as the major source of freshwater to the riparian communities and their livestock. Increased intensity of agriculture and deforestation coupled with the rapid growth of urban centers and industrial activities pose a potential threat in degrading streams and rivers in the catchment. In
38
Kiambu County, growing human populations is rapidly depleting available freshwater supplies as a result of increasing human population leading to an increase of water withdrawn from natural freshwater ecosystems.
According to Gleick (1998), this situation is also characteristic of many other areas all over the world, as the human population has increased eightfold during this century. Depletion of freshwater ecosystems can be shown to have occurred if these systems exhibit a number of systems as described by Ritcher et al., (2003) in Figure 2-8. Facing an ominous specter of increasingly severe water-supply shortages in many areas of the world, social planners and government leaders are exploring strategies for managing water resources sustainably
(IUCN, 2000). Ritcher et al (2003) proposed a framework for ecologically sustainable water management shown in Figure 2-9. Because the Ruiru and Ndarugu watersheds face similar scenarios, the framework described below provides a suitable approach to a sustainable water management strategy in the study area.
39
Figure 2-8: Relationship between Natural State of Ecosystems and Altered State of
Freshwater Systems.
Source: Ritcher et al., (2003)
40
Figure 2-9: A framework for ecologically sustainable water management.
Source: Ritcher et al., (2003)
41
Ecological degradation has generally been an unintended consequence of water management, stemming from a lack of understanding of water flows necessary to sustain freshwater ecosystems (Stanford et al. 1996). Natural freshwater ecosystems are strongly influenced by specific facets of natural hydrologic variability (Poff et al., 1997). Of particular importance are seasonal high and low flows, and occasional floods and droughts
(Richter et al., 1997). A river’s flow regime is now recognized as a ‘‘master variable’’ that drives variation in many other components of a river ecosystem, e.g., fish populations, floodplain forest composition, nutrient cycling, in both direct and indirect ways (Sparks
1995). Species richness and productivity characteristic of freshwater ecosystems is strongly dependent upon, and attributable to, the natural variability of their hydrologic conditions (Poff et al., 1997).
2.3.6 Relationship between Land Use Patterns and Rivers’ Ecological Health
Stream biodiversity maintenance in the face of encroaching human development has received much attention in recent years (Sponseller et al. 2001). Studies have shown that land use conversion (e.g. forests to pastures and/or residential areas) may influence in- stream habitat and macroinvertebrate communities in a number of ways, including loss of terrestrial vegetation (Swank, et.al., 1988) and increased area of impervious surfaces
(Changnon and Demissie, 1996). This can influence evapotranspiration and infiltration, besides altering natural flow regimes (Poff et al., 1997). A variety of land-use practices increase sediment inputs into streams, which alter substratum characteristics and channel morphology, leading to a reduction in stream macroinvertebrate diversity (Waters, 1995).
Disturbance and removal of streamside vegetation and subsequent increased solar radiation reaching the stream channel can increase stream water temperature (Quinn et al., 1997)
42 which may alter stream thermal profiles that are critical to the life history and ecology of macroinvertebrates (Ward and Stanford, 1982). In addition, modified land use may alter the catchment and hydrology of streams by adding inputs of inorganic nutrients from terrestrial sources (Johnson et al., 1997), and this has been shown to interact with increased light availability and stream temperature to enhance in-stream primary production
(Webster et al., 1983), leading to changes in the trophic structure of benthic communities
(e.g. Gurtz and Wallace, 1984).
The magnitude and intensity of land-use effects may depend on the spatial distribution of land use in different catchments (Allan and Johnson, 1997). Catchment hydrology characteristics, coupled with the availability of inorganic nutrients in streams, is often attributed to processes that occur across the terrestrial landscape (Hunsaker and Levine,
1995). In contrast, the availability of light and organic carbon in streams is related to processes restricted to the scale of streamside vegetation (Gregory et al., 1991). Several studies have shown that, depending on the spatial scale, in-stream physico-chemical and biotic aspects may be constrained by catchment properties, particularly basin geology and the distribution of arable agriculture (Richards, et al, 1996). In addition, studies have shown that land-cover in the riparian corridor has a strong influence on bank erosion and in-stream sedimentation.
However, Richards et al., (1997) showed that despite a close relationship between catchment-scale properties and channel structure and hydrology, macro invertebrate species characteristics can be correlated with local habitat characteristics. Other studies
(e.g. Johnson et al., 1997) show that the strength of relationship between stream-water chemistry and catchment scale versus riparian scale land-cover patterns may be influenced
43 by seasons. Macro invertebrate assemblages have been used as bio indicators of stream biological integrity (Collins et al., 2008). The use of a multimetric approach that utilizes the index of biotic integrity (IBI) (Karr, 1981) has gained interest in biological assessment of rivers and streams in urban and suburban catchments (e.g. Collins et al., 2008). Studies have demonstrated the usefulness of this index in assessing the biological integrity of studied rivers and streams (e.g. Masese et al., 2009). For example, Raburu et al., (2009) demonstrated that in the lower catchments of Lake Victoria basin, urban and suburban developments are replacing agricultural land use in most of the catchments. Studies that have compared the two land use types indicate that a low level of watershed urbanization triggers a biological response of greater strength and magnitude than watershed agriculture due to fewer pollutants into streams and rivers (Wang and Lyons, 2003).
2.3.7 Hydrologic modeling of Watersheds
Human activities contaminate surface waters in two ways: (1) through point sources, such as sewage treatment discharge; and (2) by non-point sources such as runoff from urban and agricultural areas (Sliva and Williams, 2002). Non-point sources are especially difficult to detect since they generally encompass large areas in drainage basins and involve complex biotic and abiotic interactions (Tong and Chen, 2002). Natural catchment characteristics such as topography and surficial geology and the biochemical processes in the terrestrial environment can influence the hydrochemical response of rivers (Moldan and Cerny,
1994).
Surface runoff, especially under the first flush phenomena, is an important source of non- point source pollution, and may be enriched with different types of contaminants under different land use (Tong and Chen, 2002). For example, runoff from agricultural lands may
44 be enriched with nutrients and sediments, while runoff from urban areas may be enriched with rubber fragments, heavy metals, sodium and sulphate from road networks (Tong and
Chen, 2002). Several hydrological processes, such as evapotranspiration, interception, infiltration, percolation and absorption, coupled with different types and extent of vegetative surfaces can modify the land surface characteristics, water balance, hydrologic cycle, and the surface water temperature (LeBlanc et al., 1997). These processes affect the quantity of water available for runoff, streamflow and ground water flow, as well as the physical, chemical and biological processes in the receiving water bodies (Tong and Chen,
2002). Therefore, there is a strong relationship between land use types and the quantity and quality of water (Gburek and Folmar, 1999). Several studies have shown that there is a strong relationship between land use types and water quantity, quality. For example, in a study of the effects of forested, agricultural and urban areas on water quality and aquatic biota in the Piedmont ecoregion of North Carolina, Lenat and Crawford (1994) found that the agricultural lands produced the highest nutrient concentrations.
Fisher et al., (2000) also noted a higher amount of nitrogen, phosphorus and Fecal coliform bacteria in the poultry production areas in the Upper Oconee Watershed in Georgia. In another study of Coweeta Creek in western North Carolina, Bolstad and Swank (1997) observed that there were consistent changes in water quality variables, concomitant with land-use change. Similarly, in an earlier study of the Little Miami River Basin, Tong (1990) found that urban development in the watershed had caused substantial modification on flood runoff and water quality. Therefore, changing land use and land management practices are regarded as one of the main factors in altering the hydrological system, causing changes in runoff (Mander et al., 1998), surface water supply yields (Wu and
45
Haith, 1993), as well as the quality of receiving water (Changnon and Demissie, 1996).
Although there have been some studies on the impacts of land use on water flows and quality, the complex intrinsic relationships of land use, water yields and water quality in different geographical areas under different scales are yet to be elucidated.
Although watershed management and catchment scale studies have become increasingly more important in determining the impact of human development on water quality both within the watershed as well as that of the receiving waters, they still leave many questions unanswered. For example, there is an ongoing dispute on whether the land use of the entire catchment or that of the riparian zone is more important in influencing the water quality, if all other factors remain constant (Osborne and Wiley, 1988). These uncertainties remain partly because different catchments have unique combinations of characteristics that influence water quality, and partly because thorough investigations at the watershed scale are extremely time and resource consuming and therefore limited (Johnson et al., 1997).
Several methods have been developed to study the implications of LULC changes on hydrological processes such as the paired catchments approach, time series analysis
(statistical method), and hydrological modelling (Li et al., 2009). Among these approaches, hydrological modelling has widely been applied throughout the world as it requires fewer resources and provides more flexibility in simulating and comparing watershed processes in actual versus ideal scenarios (Li et al., 2009). Such flexibility was shown by Fohrer et al. (2001) by assessing the hydrologic response to LULC changes in four meso-scale watersheds in Germany with different LULC distributions, and tested the model performance for changing LULC in an artificial watershed with one crop at a time and one under-lying soil type to eliminate the complex interactions of natural watersheds.
46
Effective analytical tools, such as geographical information systems (GIS) and multivariate statistics, are able to deal with spatial data and complex interactions, and are coming into common usage in watershed management (Cao et al., 1999). However, their effectiveness depends on the quality and quantity of data collected in the field, which tend to be sparse, especially when dealing with entire watersheds (Sliva and Williams, 2001). Watershed models simulate hydrologic processes in a holistic approach by incorporating of the watershed area (Oogathoo, 2006), compared to other models which primarily focus on individual processes or multiple processes within a water body without full incorporation of watershed area.
Watershed models were developed in the period between 1960 and 1990. These include:
HEC-1, developed in 1967 at the Hydrologic Engineering Center in Davis, CA (USACE
1981); the Hydrologic Simulation Program in Fortran (HSPF) in the 1960’s (Bicknell et al., 1993); Topography based hydrological MODEL (TOPMODEL) in 1974 (Kirkby and
Weyman, 1974); and Soil and Water Assessment Tool SWAT (Arnold et al., 1998). The release of these models was followed by improvement on data management and utilization
(Edsel et al., 2010) through graphical user interfaces (GUIs) with geographic information systems (GIS) and the use of remotely sensed data (Borah and Bera, 2004).
The use of GUIs allows for visualization of watershed processes, while GIS enables acquisition, storage, retrieval, analysis and manipulation of hydrological information in a spatial setup. The trend in watershed modelling has been influenced by advancements in
GIS and remote sensing techniques (Edsel et al., 2010). These include remote sensing techniques such as radar and satellite imaging, which are used to obtain spatial information on land use and soil type at regular grid intervals with repetitive coverage. GIS technology
47 has provided hydrologists with further capabilities in reducing computation times, efficiency in handling and analyzing large databases that describe heterogeneity in land surface characteristics, and improving display of model results (Jain et al., 2004). Early
GIS software packages existed as isolated software programs with minimal functionality, unsophisticated user interfaces and limited processing capabilities (Clarke, 2003). This was influenced by fairly simple operating systems that possessed minimal flexibility, making these systems difficult to manipulate (Edsel et al., 2010). Advances in GIS software, coupled with availability of better hardware, have contributed to the development of hydrological modeling in GIS platform. In addition, both commercial and free software resources are available, giving users a wide range of models to choose from. However, the models require advanced technical understanding of the background processes in order to work, which limits the number of users who can effectively apply them to hydrological processes.
2.3.8 Comparison of GIS-based Hydrological Models
Hydrological models vary in terms of spatial resolution, complexity and process representation. These models are the Water Balance-Simulation Model (WaSiM), SWAT,
The UHP-HRU and the GR4J model (Cornelissen et al., 2013). The WaSiM was developed by Schulla (1997) to evaluate the influence of climate change on water balance in low and high mountain ranges. The model is deterministic, spatially distributed and incorporates physical and conceptual approaches to describe relevant hydrological processes.
The SWAT model is described in section 4.6.1 above as a time-continuous, semi- distributed catchment model developed by the US Agricultural Research Service (Arnold
48 et al., 1998) and it is used to evaluate the influence of climate, land use and agricultural cultivation techniques on water quality and sediment yield.
The UHP-HRU model is also a semi-distributed conceptual model originally developed by
Bormann and Diekkrüger (2004) and extended by Giertz et al., (2006) using the hydrological response unit concept and by Giertz et al., (2010) using a routing routine.
The GR4J model (Génie Rural à 4 Paramètres Journaliers; Perrin et al., 2003) is a daily lumped rainfall-runoff model with only four parameters, all of which must be calibrated.
Although all the models are time continuous, they have several fundamental differences in their operation as shown in as shown in Table 2-1. For example, the WaSiM model uses a grid based spatial discretization, which is in contrast to SWAT and UHP-HRU (they divide the catchment into sub-basins, which are further divided into Hydrological Response Units
(HRUs). Another difference is that SWAT creates fixed HRUs based on super-imposed land use and soil maps, while UHP-HRU allows for yearly changing land use. While
SWAT and WaSiM divide the soil into numerical layers, UHP-HRU divides the soil into two storage components whose recession constants need to be calibrated (the root zone and unsaturated zones). The root zone storage is filled by the amount of rainfall not captured by interception and runoff, while the unsaturated zone storage is filled by percolation. This is also dependent on water level in root-zone storage, its maximum water storage capacity
(field capacity) and a recession constant. In terms of interflow representation, all the models differ fundamentally. In WaSiM, lateral flow is a function of slope length and inclination, saturated water conductivity, drainable porosity and the amount of drainable water stored in the saturated zone. UHP-HRU on the other hand determines the amount of interflow as a ratio between the actual and maximum water storage of the unsaturated zone.
49
In SWAT, interflow is calculated using a linear function which consists of slope length and angle, saturated conductivity, drainable porosity and the amount of water that is stored in the saturated zone. All models use a linear storage approach to calculate base flow, but they differ in their representation of the groundwater layer. SWAT and UHP-HRU simulate one deep aquifer. The shallow aquifer is recharged by percolation and reduced by deep percolation to the deep aquifer and capillary rise to the unsaturated zone.
50
Table 2-1: A comparison of four commonly used hydrological Models, their processes and process representation
Process WaSiM (Schulla and SWAT (Arnold et UHP-HRU GR4J (Perrin et al., 2003) Jasper, 2007) al., 1998) (Giertz et al., 2010) Interception Storage approach- function of Storage approach- Storage Not considered LAI function of LAI approach- function of LAI Potential Penman–Monteith (Monteith, Penman–Monteith Penman (1956) Penman (1956) Evapotranspiration 1975) (Monteith, 1975)
Actual Separate calculation of Calculated separately Depends on PET If PET- Precipitation > 0, Evapotranspiration evaporation from vegetated for evaporation and and water actual evapotranspiration is soils considering all soil transpiration, availability in taken from soil water layers and from bare soil for reduction of PET by root storage storage using parabolic the first soil layer; both soil water content zone equation reduced by soil water content of first soil layer Soil module Richards equation Tipping bucket, soil Linear storage, Production and routing store is divided into soil is divided numerous numerical into root and layers
51
unsaturated zone Infiltration Based on Peschke (1977) and SCS curve number SCS curve Calculation of net Green and Ampt (1911) (SCS, 1972) number (SCS, precipitation with parabolic 1972) function
Percolation Function based on soil Storage routing; Storage routing; Power function saturation and saturated water content must be water content conductivity above field capacity must be above field capacity Interflow Storage approach; comparing Kinematic storage Linear storage No distinction maximum and actual rate model approach
Overland flow Horton overland flow SCS curve number SCS curve No distinction (SCS, 1972) number (SCS, 1972) Baseflow Linear storage approach Linear storage Linear storage No distinction approach approach Routing Kinematic wave approach Continuity equation Continuity Water is staggered into a considering retention and using Manning’s equation using a number of inputs for the two translation equation simplified unit hydrographs storage approach Source: Cornelissen et al., 2013.
52
The SWAT model is designed to utilize alternative data (land use change, land management practices and climate) to model watersheds (Neitsch et al., 2005; Arnold et al., 1998). Its operation in a GIS makes it convenient for definition of watershed features, storage, organization and manipulation of the related spatial and tabular data (Di Luzio et al., 2002). The model’s capability to run with minimum data inputs makes it ideal to be employed in data-scarce watersheds, such as Ruiru and Ndarugu basins. These characteristics of the SWAT model, along with its strong computational efficiency, were considered in its selection as the model of choice in this study.
2.3.8 Application of the SWAT Model in Kenya
Table 2-2 and Figure 2-10 summarizes main publications and locations on SWAT applications in Kenya respectively. These studies focus on calibration uncertainty, runoff and sedimentation, land use, climate change, water quality, SWAT development and environmental policy.
53
Table 2-2: Applications of the SWAT Model in Kenya
Calibration Runoff and Land Climate Water SWAT Environmental
uncertainty sedimentation use change Quality development policy
Le and x x
Pricope
(2017)
Omwoyo et x
al (2017)
Musau et al., x
(2015)
Musau et al., x
(2014)
Baker and x
Miller
(2013)
Dessu and x
Melesse
(2012)
Mango et al., x x
(2011)
Odira et al., x
(2010)
Githui et al., x
(2009a)
Githui et x x
al.,(2009b)
Jacobs et x x
al.,(2007).
Jayakrishnan
et al., (2005)
54
Land use
Most of the SWAT studies in Kenya have focused on the impact of land use practices on hydrological processes. Amongst the earliest SWAT applications in Kenya include a study by Jayakrishnan et al., (2005) which applied the model to study effect of land use change associated with dairy farming on the stream flow and sediment transport of the Sondu River basin which drains 3050 km2 of land to Lake Victoria in Kenya. The study indicated that the monthly simulated discharge of existing land use “compares well” to the observed value and reported a Nash–Sutcliff efficiency (NSE) of 0.1, which is regarded as a poor performance. This performance was attributed to poor input datasets, and therefore the study emphasized the need for development of better model input data sets in Africa which are critical for detailed water resources studies (Nash et al., 1970). Githui et al. (2009a) used SWAT in the Nzoia watershed, western Kenya, to examine the impacts on base flow and streamflow under prevailing land use change trends (e.g., forest conversion to smallholder agriculture) versus afforestation, and found out that flood risks were exacerbated if existing land use change trends were to continue.
The study revealed a strong relationship between the impacts of changing land use
(especially increasing in agricultural land use) the hydrological regime of the Nzoia River catchment in Kenya. Increased runoff in Nzoia catchment was attributed to an increase in agricultural land use and a corresponding decrease in forest cover between 1973–2001, with SWAT simulations reporting increased runoff by about 119% between 1970 and 1985
(Githui et al., 2001). In another study, Mango et al. (2011) used SWAT coupled with satellite-based estimated rainfall to support water resources management efforts in the
Mara River Basin, demonstrating that in data scarce regions such as Kenya, it is possible
55 to approach water resources challenges using modeling methods (Mango et al., 2002).
Although the study emphasized that sustainable water resources management can be a challenging undertaking in data-scarce regions, it demonstrated that the SWAT model can provide fair results that can be used to explore land use impacts and inform watershed interventions. The study, however, cautioned that such models may be impeded by uncertainties, including processes unknown to the modeler, processes not captured by the model and simplification of the processes by the model (Abbaspour et al., 2007). The study concluded that any further forest conversion would reduce dry-season flows and intensify peak flows in the watershed, further exacerbating already serious problems related of water scarcity in dry periods and hillslope erosion during the wet season.
Baker and Miller used the SWAT model to identify the spatial and temporal dynamics of magnitude and direction of land use change in the River Njoro watershed in Kenya. This study showed that land use changes in River Njoro Watershed led to a shift to increased surface runoff in the uplands coupled with decreased groundwater recharge. The study attributed deforestation of the Mau Forest to increased erosion and sedimentation as a result of flashier flows and increased streamflow. The importance of a healthy watershed was highlighted by this study, because upstream conditions have a direct impact on downstream ecology (e.g. River Njoro feeds into Lake Nakuru, an important National Park in Kenya and a world renown Ramsar Site supporting diverse wildlife populations and birds). The study identified a potential increase in conflict over dwindling water resources, especially between agricultural and pastoral communities within the watershed (Baker et al., 2013).
Odira et al. (2010) simulated streamflow changes as a result of the land use land cover changes using the SWAT model in Nzoia watershed in Kenya and reported increased
56 discharge during wet months and a decreased discharge in the dry periods (Odira et al.,
2010).
Runoff and sedimentation
Dessu and Melesse (2012) used the SWAT model for long-term rainfall–runoff simulation in the Mara River Basin on the border of Kenya and Tanzania. Like most watersheds in
Kenya, the Mara River Basin is highly threatened by multiple watershed-level threats, including overgrazing, agricultural expansion, deforestation, human settlement, erosion and sedimentation, flooding and low flow. As such, this study utilized the SWAT model to understand the interaction among the natural processes and human activities in the basin.
Although the study was limited by scarcity of observed data, the study showed that in the absence/limitation of rainfall data, alternative sources of data, such as satellite rainfall estimates (RFEs) can be used to run the model and produce acceptable results. Previous studies in the same basin applied SWAT on two tributaries in the Mara River Basin
(Nyangores and Amala) combined rain gauge data and satellite RFEs to assess the effect of land use change using SWAT and reported that RFE performed better than rain gauge rainfall records. However, the authors indicated that the quality of rain gauge data may have contributed to these observations, which further justifies the use of alternative data sources such as RFEs (Le and Pricope, 2017).
Environmental policy
In recognition of the critical need to address the impacts of deforestation on watershed processes, Jacobs et al used the SWAT model to predict the impacts of land use on the
Masinga Reservoir in Kenya (Jacob et al., 2007). The Masinga Reservoir serves as a
57 storage reservoir on Masinga Dam, which is one of the so called “seven forks” along River
Tana, the longest river in Kenya. Masinga Dam and its reservoir are used for power generation, and it’s one of the most important dams in Kenya (Watermeyer et al., 1976).
Because of sedimentation, it is estimated that complete siltation of the Masinga reservoir will occur within 65 years unless some type of intervention is undertaken, which is expected to drastically reduce the lifespan of Masinga dam (earlier estimated to reach upwards of 500 years prior to its construction). Jacobs et al., (2007) therefore focused on the land use interventions on reforestation of the upper reaches of the catchment as a basis to provide improved catchment hydrology and water quality by reducing sediment and runoff into the lower portions of the Tana River basin. The authors integrated an economic model with SWAT to develop to analyze the economic benefits and associated costs with establishing a green payment program in the Tana River basin. Further details of the study are discussed in section 3.4 on practical applications of SWAT model in Kenya.
Climate Change
Musau et al., (2015) forced the SWAT model with monthly temperature and precipitation change scenarios for the periods 2011–2040 (2020s), 2041–2070 (2050s) and 2071–2100
(2080s) to simulate the impacts of climate change on hydrological process in the upper
Nzoia basin, Kenya. This study efficiently captured the historical hydrological processes in the upper Nzoia Basin based on the observed meteorological data and can therefore be applied in understanding of the dynamic water balance processes in this area. The study reported that there are expected to be large uncertainties in the future precipitation, temperature and streamflow with significant implications on development and ecosystems in the watersheds and downstream areas. Under the current climate change uncertainties,
58 this study provides useful insights for long-term basin-wide strategic planning and implementation of development projects, disaster preparedness and water resources management in this important basin (Musau et al., 2015).
Omwoyo et al. (2017) used SWAT to simulate stream flow response under changing climate for the Upper Ewaso Ngiro Catchment in Kenya. The study generated temperature and rainfall climate change scenarios for representative concentration pathway 4.5 and 8.5 from 2021-2080 relative to the baseline period 1976-2005. Simulated stream flow varied between periods in different scenarios, with March-May showing a decrease (-26 to -10%) and June-February an increase (9-114%). Generally, the study reported stream flow response to be sensitive to changes in rainfall, and placed emphasis on water conservation and catchment management strategies such as agroforestry, afforestation and reforestation
(Omwayo et al., 2017).
Calibration Uncertainty
As with most hydrological models, efficient estimation of optimum parameters is inevitable to accurately estimate hydrological processes. Musau et al., (2014) applied the
HydroPSO R package to SWAT model in R software to assess parameter identification and calibration in Nzoia Basin, Kenya. In this study, fourteen parameters representing the surface flow, subsurface flow and channel routing components of the water balance were selected for model optimization. The study demonstrated that the SWAT model can effectively simulate stream flow and can be successfully combined with R software to harness the combined benefits of a distributed hydrological model and flexible computing capability of the open source R software. However, the authors acknowledged that model
59 error and input data are major sources of uncertainty in hydrological modeling (Musau et al., 2014).
SWAT development
Le and Pricope recognized that Hydrologic models are an increasingly important tool for water resource managers as water availability dwindles and water security concerns become more pertinent in data-scarce regions. The study piloted the incorporation of the
Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) dataset into
SWAT as an alternative to conventionally available climate datasets to assess its applicability in Nzoia basin of Kenya, a data scarce region. The CHIRPS dataset provides quasi-global high-resolution precipitation information derived from a blend of in situ and active and passive remote sensing data sources. The study concluded that the CHIRPS dataset is only suitable for relatively flat, poorly gauged, small-scale watersheds and with an understanding of its limitations, but can also be used in SWAT to inform water resource management strategies in data scarce regions (Le and Pricope, 2017).
60
Figure 2-10: Location of SWAT-applied basins in Kenya. Source: Modified from Survey of Kenya (2016)
Figure 2-10 indicates that most SWAT studies in Kenya have been conducted in the upper catchments. Limitation of long term data to verify and calibrate the model has been a major limitation on the use of the model countrywide, which has also contributed to the limited spread of the use of the model in other parts of the country. In contrast to most studies, this study applied the model in data scarce basins.
2.3.9 Practical Application of SWAT model in Kenya Two studies showed evidence of practical applications of the model in water resources management. Mango et al., (2011) used SWAT to support water resources management
61 efforts in the Mara River Basin. This study demonstrates that it is possible to use models to approach water resource challenges in data scarce regions in Kenya. In another study,
Jacobs et al., (2007) used SWAT to evaluate alternative reforestation scenarios in the upper
Tana basin, one of the most important basins in Kenya. The authors used an economic model to determine the opportunity costs associated with reforestation and the economic incentives, including green payments, which would be required to induce upper catchment users to engage in reforestation activities. The study found that reforestation activities would decrease sediment loading in the Masinga Reservoir (used for electricity generation) by 7 percent. This study provided practical scenarios for green payments (including specific amounts of money that users in the upper catchment would be paid for each ton of sediment retained in their farms), but also indicated that it was beyond the capacity of downstream users to sponsor green payments.
2.4 Conceptual Framework
Human disturbance of ecosystems and their services through land-use changes affect hydrological mechanisms and ecological processes as indicated in Figure 2-11. Ecosystem management through land use practices can mediate the influence of anthropogenic activities in changing the hydrological processes in the landscape, and these processes can be identified by the Soil and Water Assessment tool.
The quality of water in rivers is dependent on access to ecosystem services through different land-use practices. The health of ecosystems is affected negatively by decreased and inferior agricultural production caused by transformation in physical and human environments; excessive use of chemical pesticides and fertilizers, salinization, contamination by heavy metals, and soil depletion (Lebel, 2003).
62
Figure 2-11: Causal linkages between land-use transition, ecosystem services and water quality in the Study Area.
63
2.5 Research Hypotheses
1. Ho1: There is no change in land use practices between 2005 and 2015 in Ruiru
and Ndarugu Watersheds.
2. Ho2: Land use and land cover systems have no impact on water quality and
macro invertebrate composition in Ruiru and Ndarugu Rivers.
3. Ho3: There is no significant difference between measured and simulated water
quality parameters in Ruiru and Ndarugu Rivers.
64
CHAPTER THREE 3.0 STUDY AREA The Ruiru and Ndarugu basins are located in Kiambu County, in the former Central
Province of Kenya. The study area was selected due to its high population growth rate occasioned by the proximity to Kenya’s capital, Nairobi, which makes it experience rapid land use changes and pressure on available surface water resources.
3.1 Location
The Ruiru River (1° 4'43.90"S, 36°50'54.24"E) and Ndarugu River (1° 0'49.80"S,
36°55'8.86"E) are major tributaries of Athi-Galana River, the second longest river in
Kenya. Both rivers are located in Kiambu County in the central part of Kenya as shown in
Figure 3-1. The rivers originate from Gatamaiyo Forest, the southern-most tip of the
Aberdare Ranges. The drainage areas of the Ruiru and Ndarugu Rivers are, approximately,
367km2 and 230km2; the lengths of main river channels stands at 40kms and 48km with average gradients of about 0.057 and 0.054 respectively.
Kiambu County is located between latitudes 0o 46’ and 1o31’ south of the Equator and longitudes 36 o30’ and 37 o 20’ east of Greenwich meridian. It covers an area of 2,543sq
Km. Kiambu County borders Nairobi County to the South, Muran’ga County to the North, and Nakuru County and Nyandarua County to the West. The county is linked to the bordering counties by tarmac and marrum roads, airstrips and modern postal and telecommunication services. Kiambu County has 8 constituencies (Gatundu South,
Gatundu North, Juja, Githunguri, Kiambaa, Kabete, Limuru and Lari). The Southern tip of the Aberdare ranges (Gatamaiyo Forest) occupies about 25% of the total land area of the county. The county is divided into twelve (12) sub-counties namely Limuru, Kikuyu, Lari,
65
Gatundu South, Gatundu North, Githunguri, Kiambu, Ruiru, Thika, Juja, Kiambaa, and
Kabete. Its headquarters are in Kiambu town, which is located at 1°10′1″S 36°49′19″E and
1°10′1″S 36°49′19″E at an elevation of 1,720 m.
3.2 Geology and Soils
The study area is mainly composed of volcanic rocks of varying ages (Saggerson, 1967).
To the northeast of Kiambu town, geology varies from Miocene to Pleistocene volcanics mainly in the Aberdare ranges. The Sattima series occurs towards the south of the Aberdare ranges, which are dominantly intermediate and basic lavas. The Pleistocene volcanics are grouped into; the upper trachytes, which include Kinari tuffs and Limuru trachytes, and the middle trachytes, which are composed of Tigoni, Karura, Kabete and Ruiru Dam trachytes.
The Upper trachyte division comprises of the Kinari tuffs and the Limuru trachytes. The
Middle trachytes are composed of the Karura, Tigoni, Kabete and Ruiru Dam trachytes.
Towards the East of Kiambu town, the geology ranges from Pliocene to the Pleistocene
Basalts and intermediate lavas of the Laikipian and Sattima series to the Miocene Basalts of the Simbara series and Pyroclastic rocks deposited on eroded surface of the Simbara basalts. The central and southern parts of the study area consist of the Simbara basalts. The
Kamiti and Kahawa are composed mainly of Tertiary volcanics, predominantly trachytic tuffs and agglomerates on the plateau surface. The Kapiti phonolites and the Simbara basalts and agglomerates are exposed in the major valleys.
To the northwest of the study area, the landscape rises into the Aberdare Mountains. The geomorphology of this mountainous landscape is characterized by long slopes that lead to narrow valleys with occasional crags and rocky hills. In the direction of Kinale, the mountainous landscape changes to low interfluves and flat bottomed valleys extending
66 back from the top of the Rift Valley Escarpment to the footslopes of the Aberdare Ranges.
In the direction of Kikuyu and further to the Southwest of Kiambu town, the landscape changes to long narrow approximately parallel ridges separated by narrow winding valleys of varying widths and local streams. To the East of Kiambu town, the landscape changes to an extensive toe slope characterized with broad long undulations and gentle depressions, which are occasionally dissected by winding, steep sided, flat-bottomed valleys. The area to the south- east of the study area is characterized by low plains with occasional low hills rising from the landscape. The geology and geomorphology of the study area has an influence on land use and land cover characteristics, with the upper, higher elevation areas characterized by areas under montane forest, tea plantations and pineapple farms. The mid elevations are characterized by small-scale coffee growing, maize, beans, cabbages, potatoes among others. The lower elevations are mainly grassland previously occupied by coffee and sisal plantations in some areas. These land use types have an influence on water quality of the rivers in the landscape, including Ruiru and Ndarugu Rivers.
Soil distribution in the study area is directly linked to the geomorphology of the area, with the lower parts in the plains mainly occupied by Ironstone soils, Lithosols and Vertisols
(Sombroek et al., 1980). The toe slopes of the volcanic foot ridges are consisted of Eutric-
Nitisols and Nito-chromic Cambisols, while the upper ridge crests mainly consist of Humic
Nitisols, Mollic Andosols and Ando-humic Nitisols. The soils are deep, well drained, strongly weathered and dark red to dark reddish brown friable clay soils (Shitakha, 1983).
The topsoils have medium, moderately strong, sub-angular and blocky structure; while the
B-horizons are slightly hard when dry, friable when moist, sticky and plastic when wet
(Shitakha, 1983). The soils are acidic with pH in the topsoils varying between 6.0 and 6.6
67 and between 5.1 and 6.6 in the B-horizons. The cationic exchange capacity of the soils varies between 9.0 and 34 cmol/kg in the topsoils and 10 to 28 cmol/kg in the B-horizons.
The soils are classified as Mollic and Humic Nitisols according to the FAO-UNESCO
(1974) Soil Map of the World.
Most of the soils in the study area are dominated by Nitisols, which are generally soils with an argic B-horizon, with clay distribution, which does not show a relative decrease from its maximum by more than 20% within 150 cm of the soil surface (FAO, 1977).
They exhibit gradual to diffuse horizon boundaries between the A and B-horizons and have nitic properties in some sub-horizons within 125 cm of the soil surface. These soils are well-drained, dark red to dark reddish brown clay soils with good structures and high infiltration rates. They vary in depth, where they are extremely deep in the flatter areas to moderately deep in the steeply sloping areas. In most parts of the study area, the soils have a high concentration of organic matter in the topsoil.
In the lower plains, the Vertisols are predominant. These soils are dark-coloured, poorly drained soils with wedge-shaped or parallel prismatic structures. These soils are shallow to moderate in depth with a few pockets that are deep. To the Northwest of Kiambu
Town, the Andosols are the predominant soils. These are low bulk density soils, with a bulk density of less than 0.85 gm/cm3. These soils developed on volcanic ash and tuffaceous deposits. They are extremely deep, well drained with pockets that are poorly drained in bottomland areas. To the south-west of the study area, the Planosols, Lithosols and rock outcrops are found, particularly along and adjacent the rims of the Rift Valley Escarpment.
68
Planosols are moderately deep to shallow, moderate to poorly drained soils that support semi- arid and grass vegetation.
The soils in the landscape influence the land use types that develop in different areas, which in turn influence the erodability of the soil, the infiltration capacity and ultimately the hydrology of the landscape. Areas under forest cover are expected to have lower erosion rates compared to areas under agriculture and urban/settlement land use.
3.3 Topography
The study area and Kiambu County in general has four main topographical zones, namely the
Upper Highlands, Lower Highlands, Upper Midlands and Lower Midland Zone. The Upper
Highland Zone is found to the north-west of the landscape, and forms an extension of the
Aberdare ranges. This area lies at an altitude of 1800 to 2550m above sea level. The area is characterized by steep, highly dissected ranges, and it is mostly found in Lari division.
The lower highland zone is found in Limuru and some parts of Gatundu, Kabete, Githunguri,
Kiambaa, Kikuyu, Lari and Limuru. This area lies between 1500 to 1800m above sea level is characterized by hills, plateaus, and high-elevation plains. The area is generally a tea and dairy zone, with some agricultural activities like maize, horticultural crops and sheep farming also practiced.
The upper midland zone lies between 1300 to 1500m above sea level and it extends into most parts of Juja. It comprises of volcanic middle level uplands, while the lower midland zone partly covers Thika Town (Gatuanyaga), Limuru and Kikuyu (Jaetzold and Schmidt, 1983).
3.4 Climate and Hydrology
Rainfall is predominantly influenced by altitude, with the mean annual rainfall ranging from 500mm in the lower parts around Thika and increasing gradually to 2000 mm in the
69 upper region. The rainfall regime is bimodal, where long rains fall in April and May. This is followed by a cool dry season in July and August, before short rains which fall from
October to December. The mean maximum temperatures range from 26oC to 28oC in the eastern and southern parts, and 18oC to 20oC in the Northwest; while the mean minimum temperatures vary between 14oC and 16oC in the eastern parts and 6oC to 8o C in the north-western parts. The coolest months are July and August, while the hottest months are
January to March. The average relative humidity ranges from 54% in the dry months and
300% in the wet months.
The Climatic Zones range from Zone V-4, which is fairly warm, semi-humid to semi-arid in the southern parts to zone 1-7, which is cool and humid in the north-western parts
(Sombroek et al., 1980). Climate, soils and rainfall determine the agro-ecological zones
(Jaetzold and Schmidt, 1983). These are coded as UM1, LH1, UH1, and UH2 respectively.
UH refers to the Upper humid zone and LH to Lower Humid Zone. UM is the Upper Moist
Zone and LM the Lower Moist Zone. Humid Zones are wetter than Moist Zones. The agro- ecological zones determine the types of land use and land cover systems, ranging from forested areas, grasslands and agriculture (tea, coffee and subsistence crops). The humid zones are wetter than the moist zones. The drier southeastern parts are characterized by
Livestock and sorghum (Sorghum vulgare) growing areas and well as the Sunflower
(Helianthus annuus)-Maize (Zea mays) growing areas. The central parts, of the study area fall in the marginal and main coffee zones respectively. The drier parts in the Southeast fall in the Wheat (Triticum aestivum)-Barley (Hordeum vulgare) and Wheat-Maize-Pyrethrum
(Chrysanthemum cinerariaefolium) zones. The north and northwestern parts fall in the
70
Coffee (Coffea arabica, Coffea canephora) -Tea (Camellia siniensis) Zone, Tea-Dairy
Zone, Sheep Dairy Zone and the Pyrethrum-Wheat (Triticum aestivum) Zone.
The hydrology of the study area consists of both surface and sub-surface water, which accounts for about 90% of the water resource. Kiambu County has four sub-catchments areas. These are: Nairobi river sub-catchment which occupies the southern part of the county with major rivers being Nairobi, Gitaru, Gitahuru, Karura, Ruirwaka, and
Gatharaini; Kamiti and Ruiru rivers sub-catchment with Riara, Kiu, Kamiti, Makuyu,
Ruiru, Bathi, Gatamaiyu and Komothai rivers; the Aberdare plateau that supports Thiririka and Ndarugu Rivers sub-catchments has mainly Mugutha, Theta, Thiririka, Ruabora,
Ndarugu and Komu streams flowing from Nairobi, Kamiti, Ruiru, Thiririka, and Ndarugu sub-catchments to form Athi River sub-catchment; the Chania River and its tributaries comprising of Thika and Kariminu Rivers which rise from the slopes of Mt. Kinangop in the Aberdares range and; the Ewaso Kedong sub catchment which runs in the North-South direction and occupies the western part of the county with many several streams that form swamps.
The land use and land cover systems in Kiambu County are expected to lead to effects on the hydrological conditions of both surface and sub-surface water. The physic-chemical water quality of water in the rivers, including Ruiru and Ndarugu, could be altered by changing land use patterns.
3.5 Flora and Fauna
3.5.1. Flora Native vegetation types remain mainly Kieni and Kinale forests which occupy an area of
426.62 Km2, although they constitute both indigenous and plantation forests. The
71 vegetation in the forested area is characterized by mixed bamboo and forest in the higher north-west section, which gives way below 2,400 m to broad-leaved forest, with species such as Ocotea sp., Podocarpus latifolius, Macaranga sp., Noxia congesta, Neoboutonia and Strombosia spp. prominent among the trees; tree ferns Cyathea manniana are also conspicuous (Blackett 1994). The Escarpment strip consists of remnant Juniperus forest and overlooks the mountainous floor the Great Rift Valley. There are also exotic trees in plantations in private farm forests. Both the gazetted and private forests are sources of a range of forest products, including electricity transmission poles, construction poles, timber, firewood and charcoal. I addition, honey and wild fruits are also harvested from natural forests.
In most other parts of the study area, natural vegetation has greatly been modified, now consisting the intensive smallholder mixed farming, large holder tea (Camellia siniensis) and coffee (Coffea arabica) farming, grazing grasslands and ranches, and built up areas.
Small scale mixed farming includes: subsistence mixed farming of annual crops and zero grazing, cash crop farming of tea, coffee and horticultural crops, which are sometimes intercropped with maize (Zea mays) and beans (Phaseolus vulgaris) for food.
3.5.2. Fauna Most fauna species occur in the forested zone which provides suitable habitat. The avifauna
(birds) of the area constitute of species characteristic of the central Kenyan highlands
(Bennun and Njoroge 1999), with over 138 species of birds in and around the study area.
Out of these, 31 bird species are forest specialists and 20 are considered rare, particularly in the forested zone. Birds include Abbott’s Starling, Jackson francolin, Hunter’s Cisticola,
72
African green ibis, Ayres hawk eagle, Crowned hawk eagle, and Red-chested owlet
(Zimmerman et al. 1996).
Mammals include African elephant, which is occasionally encountered at Kereita forest when wandering from the main Aberdare forest block. Other mammals found include
Black and White Colobus Monkey, Sykes’s Monkeys forest hogs, duikers, bush-babies, porcupines, mongoose and civets. In addition, three near-endemic butterfly species occur, namely Charaxes nandina, Neptis kikuyuensis and Neptis katama (Larsen 1991)
3.6 Socio-economic Activities
3.6.1. Population
The population density of Kiambu County has been increasing in the last 40 years. The density was 638 persons per km² in 1999, ranking 5th overall amongst the 47 counties. The sub-counties are predominantly rural, but its urban population is rapidly increasing. The poverty rate is 27.2%, making the County the 4th wealthiest countrywide. Ruiru town hosts the highest population (238,858 people), while Gatundu town has the lowest as shown in
Table 3-1.
73
Table 3-1: Population sizes in different Urban Centres in Kiambu County, showing a steady population increase since 2009
2009 Census 2012 Census 2015 Projections 2017 Projections
Town Male Female Total Male Female Total Male Female Total Male Female Total
Githunguri 4,843 5,164 10,007 5,269 5,618 10,887 5,732 6,112 11,845 6,064 6,466 12,529
Juja 20,488 19,958 40,446 22,290 21,713 44,003 24,251 23,623 47,874 25,652 24,989 50,641
Limuru 39,433 40,098 79,531 42,901 43,625 86,526 46,675 47,462 94,137 49,373 50,206 99,579
Kiambu 41,247 42,908 84,155 44,875 46,682 91,557 48,822 50,788 99,610 51,644 53,724 105,368
Karuri 53,735 53,981 107,716 58,461 58,729 117,190 63,603 63,894 127,49 67,280 67,588 134,868
Thika 68,408 68,509 136,917 74,425 74,535 148,960 80,971 81,090 162,061 85,652 85,778 171,430
Ruiru 119,147 119,711 238,858 129,627 130,240 259,867 141,028 141,696 282,723 149,181 149,887 299,067
Kikuyu 114,357 118,874 233,231 124,415 129,330 253,745 135,358 140,705 276,063 143,183 148,839 292,022
Total 464,238 472,173 936,411 505,070 513,703 1,018,773 549,494 558,886 1,108,380 581,260 591,195 1,172,454
Source: KNBS (2016).
3.6.2 Economic Activities
The main economic activities in the study area is agriculture in tea, coffee, dairy, poultry and horticulture farming, with most people in the county depending on the sub sector for their livelihoods. Mining activities include natural gas exploitation in Lari by Carbacid
Company Limited and extraction of ballast, hardcore, gravel, murram, sand and building stones in Juja, Gatundu South and Gatundu North.
The major industries in the study area are located in Thika and Ruiru Constituencies. These industries include Bidco Oil Industries, Thika Motor Vehicle dealers, Thika
Pharmaceutical Manufacturers Limited, Devki Steel Mills, Broadway Bakeries, Kenblest
74
Industry, Kel Chemicals, Thika Rubber Industries Limited, Macadamia Nuts, Campwell
Industry and Kenya Tanning Extracts Limited, Clay Works, Spinners and Spinners and
Bata Shoe Factory. The agro-processing industries include Farmers’ Choice Ltd, Kenchic
Co. Ltd, Brookside Dairies, Githunguri Dairies, Ndumberi Dairies, Limuru Milk and
Palmside Dairies, among others. Tourism activities in the study area are undertaken at several tourist attraction sites, including Kinare Forest in Lari Constituency, Chania Falls and Fourteen Falls in Juja Constituency, Paradise Lost and Mugumo Gardens in Kiambaa
Constituency, Mau Mau Caves, Gatamaiyu Fish Camp and historical sites in Gatundu and
Githunguri Constituencies. Kinare forest provides opportunities for wildlife-based tourism with its dense forest that is habitat for elephants, hyenas, bush baby, baboons, colobus monkeys, dik-dik, bush pigs, tree and ground squirrels, porcupines and many species of birds such as weavers, guinea fowls, sparrows among others.
Kiambu County has a total of 2,517 trading centres with 6,634 registered retail traders and
750 registered wholesale traders. There county also has a number of urban centres including Thika Town; Kiambu, Karuri, Kikuyu, Limuru, Gatundu and Ruiru.
3.7 Land use and land Tenure
Land use systems in the study area follow the pattern of the agro-ecological zonation and soil distribution pattern. The Swynerton land use plan of 1954 has strongly influenced the current land use activities.
Land use in the study area consists of smallholder mixed farming, large holder farming
(mainly tea and coffee), grazing, and nature conservation and built up areas. Most of the
75 smallholder farms are between three and five acres, with a few farms being less than half an acre in extreme cases.
In the large-scale farms, most farm sizes are more than ten acres, where a single crop is farmed. These farms grow coffee and tea. Individual farmers own some of the tea and coffee estates, and private companies run many of the coffee estates. These include
Machure Estate and Socfinaf Company. Privately owned large estates include Farly Dam,
Kipenda Estate, and Menengai Estate, which also own large tea plantations. Coffee and macadamia (Macadamia integrifolia) are grown as intercrops by Nando and Bob Haris estates. Anak Estate and Sukari Ranch keep grade cattle and local breeds for the production of milk and beef. They also grow forage on their large estates. The large estates with tea are located in Limuru, Tigoni, Githiga and Kambaa. Kiambu, Ruiru and Juja are areas of large-scale irrigated coffee cultivation. Small-scale tea and small-scale annual crops are found in Githunguri, Githiga and the adjoining areas. Small-scale coffee and annual crops are grown in Kiambu, Ikinu, Waruhiu, Kikuyu, Gachie, and the adjoining areas. Grasslands and ranches are mostly found in Ruai. Areas previously occupied by sisal plantations in
Juja and Thika by large scale European farms are currently being developed for settlemet.
Conservation forests are found in Uplands and Kereita in Lari Division and in Karura
Forest. A number of private estates, also have plantation forests.
Land tenure systems include freehold and leasehold for individuals, while public and state land is owned by the national government and Kiambu County government. Freehold land tenure system allows the owner to hold the land for an indefinite period of time, while leasehold confers upon the owner a limited term which can be extended upon expiry
(usually 99 years). The public or state land tenure system describes a tenure type in which
76 the government is a private landowner. Customary tenure system (where rights are based on communal ownership of land, especially to a clan or an ethnic community) is less common in Kiambu County.
Anthropogenic activities are directly reflected in land use activities, which are a key environmental concern because they may lead to deterioration of river water quality if they are not carried out in a sustainable manner. Land tenure systems, on the other hand, determine the type of land use system that may occur in an area. For example, protected areas constitute state land, and most areas under agriculture are privately owned. In this study, the relationship between land use and water quality is investigated to help in identifying primary threats to water quality for effective water quality management.
77
Figure 3-1: Map of the study area showing study basins relative to geographical features, urban centers and infrastructure in Kiambu County.
Source: Modified using Base Maps from United States Geological Survey, Survey of
Kenya and International Livestock Research Institute (2016).
78
CHAPTER FOUR 4.0 RESEARCH METHODOLOGY The methodology used to assess the spatio-temporal dynamics of land use change on water quality in the basins of Ruiru and Ndarugu, is presented. A presentation on the techniques employed, the execution of the methods, and a description of the SWAT model and how it operates (including the overall concepts) is made with an in-depth description of the procedures used.
4.1 Study Design
The study used a Multi-stage sampling survey that employed a stratified purposive sampling procedure, where sampling points were derived from three sub-basin outlets that represented three dominant land use classes in Ruiru and Ndarugu rivers. In this design, a multi-stage process was used to select sampling sites as shown in Figure 4-2. The first step involved land use categorization and categorization of the upper Athi-Galana land catchment. In the second stage, the study sub-basins were selected, which was followed by a randomization procedure of sub-basin outlets with three different dominant land use types
(Forest Dominated-FD, Agricultural Dominated-AD, and Urban/Settlement Dominated-
UD). In each sub-basin, two sampling sites per land use type were selected, resulting into a total of 12 sites (6 in each watershed) as shown in Figure 4-3. Water quality assessment was carried out in two dry seasons and two wet seasons.
79
Figure 4-1: Schematic Diagram showing the overall study design. Sampling sites were selected from a multi-stage process described above.
80
Figure 4-2: Map showing sampling sites in Ruiru and Ndarugu Watersheds, Kiambu
County.
81
4.2 Data Types and Sources
Landsat Multispectoral Scanner (MSS), Thematic Mapper (TM) and Enhanced Thematic
Mapper (ETM+) imagery for the years 2005, 2010 and 2015 were collected from Global
Land cover Facility (University of Maryland, 2016), and US Geological Survey (USGS,
2010) and employed for analyzing the spatial and temporal changes in land use- land cover in the study area. Other available reference data such as aerial photography and topographic maps, and ancillary data were also acquired for familiarization with land use systems in the study area. In addition, Food and Agricultural Organization (FAO) Africover data
(International Livestock Research Institute- ILRI data base) on land use and land cover classification for Kenya were also collected. Additionally, a review of different documents on land management; conservation legislations in forest, watershed services, wetland management and urban development plan were used to support a better understanding and get reference data on land use and land cover in the region.
The following physical-chemical parameters were measured for water quality assessment: pH, temperature, conductivity, turbidity and Dissolved Oxygen (DO). Temperature and pH were measured using a digital Hanna Instruments® HI98107 pHep pH meter directly from the field. Conductivity was measured using Hanna Instruments HI8733 conductivity meter with a four-ring conductivity probe with readings adjusted with Automatic Temperature
Compensation (ATC), assuring consistency and accurate measurements. Turbidity was measured using the Hanna Instruments® HI98703 Turbidity Portable Meter, while
Dissolved Oxygen was measured using HI2040-01 edge Multiparameter DO Meter.
Aquatic macro invertebrates which are water dwelling organisms that lack a spine and are large enough to be seen with the naked eye (e.g. snails, flatworms, crayfish, clams and
82 insects, such as dragonflies etc., including juveniles) were collected and identified.
Secondary water quality data was also collected from the Water Resources Authority
(WRA) and Ruiru-Juja Water and Sanitation Company. Data used to run the SWAT model included explicit datasets for land topography, land use and/ or land cover, soil characteristics, and climate and hydrological data on a daily time step (Schuoland
Abbaspour, 2007) as shown in Table 4-1.
Table 4-1: Data types used in SWAT and their sources
Data Type Source
Digital Elevation Modem Shuttle Radar Topography Mission (SRTM)
(DEM)
Land use/land cover map United Nations University - Institute for Water Environment
and Health (UNU-INWEH): WaterBase
Soil map Food and Agriculture Organization (FAO) Soil database
Version 3.6
Climate Climate Forecast System Reanalysis (CFSR):
http://globalweather.tamu.edu/
Measured Streamflow Water Resources Management Authority, Kiambu Regional
Office.
Source: Adapted from Arnold et al., (1998).
83
4.3 Pilot Survey
The pilot survey consisted of an initial analysis and review of existing literature and data sets for Land use, Soils, Water Resources and urban development from libraries and relevant institutions for Kiambu County in general and for the Ruiru and Ndarugu watersheds specifically. The database queries included University of Maryland’s Global
Land Cover (Land Use and Land cover), United States Geologic Survey’s Earth Explorer
(Satellite data), Survey of Kenya (Land Use and Topographical Maps), Regional Centre for Mapping of Resources for Development (Satellite data), Ruiru-Juja Water and
Sewerage Company (Water Quality), Water Resources Authority (Water Quality and
Quantity), International Livestock Research Institute (Land Use and Hydrology), World
Resources Institute (Land Use, Hydrology and Soils), Kenya Metrological Department
(Weather and Climate), National Museums of Kenya (Macro-invertebrates distribution) and National Environment Management Authority (recommended surface water quality standards). Additional existing data was reviewed when available, such as agro-ecological maps and descriptions generated by Survey of Kenya and wetlands mapping and inventory by Department of Resource Surveys and Remote Sensing.
A one week pilot survey was conducted in the study area between 7th to 13th October
2013, focusing on the areas surrounding Ruiru and Ndarugu Rivers to familiarize the researcher with the area, the people and the topography. Prior to the pilot survey, the location of sampling sites had been identified in a GIS as described in section 4.1 (study design), and their coordinates generated and loaded into DNR Garmin software which allows storage, retrieval and transfer of coordinates between GPS devices and computers.
84
The GPS coordinates of sampling sites were pre-loaded into Garmin 62 GPS. During the pilot survey, all the sampling sites were visited. The pilot survey also included verification of land use types as revealed in high-resolution satellite imagery. A mock data collection was done to ensure that both the researcher and the assistant were familiar with data collection procedure. The exercise also served to verify that all water quality parameter testing equipment was in working order and well calibrated. Areas that required special permission (especially private access roads) were also identified to ensure that permission was granted before actual fieldwork.
4.4 Assessing Spatio-temporal Land Use Dynamics
Several image processing procedures were employed for this study including image pre- processing, classification, and accuracy assessment. Landsat 7 bands 4, 5, 3- Red, Green, and Blue (RGB) were composited, as well as corresponding Landsat 8 bands (5, 6, 4, RGB).
Geometric rectification and radiometric normalization were performed for image pre- processing. The Iterative Self-Organizing Data Analysis Technique (ISODATA) algorithm was adopted to identify clusters from image data. In ISODATA analysis, unsupervised classification was first conducted to identify spatial clustering of different classes. The image was segmented into unknown classes depending on its statistical similarities by using a clustering module. Supervised classification was done, by taking sample points in stratified random sampling scheme for ground truth data and with the support of field knowledge and ancillary data for land use. Those classes were labeled to the relevant land use/land cover patterns by a posteriori analysis. This technique implies a grouping of pixels in multi-spectral space. Pixels belonging to a particular cluster are therefore spectrally similar. In order to quantify this relationship Euclidean distance was used as a similarity
85 measure. Finally based on reference data which was gathered during field work, accuracy assessment was employed to measure the reliability or the overall accuracy of the classification. The reference classes were compared with the result of classification and the ratios of correctly versus incorrectly classified pixels was calculated for each class. The accuracy assessment using randomly generated points was performed through a standard method described by Cogalton (1991). Coordinates of these points were pre-set in a GPS and the exact points visited on the ground as shown in Plate 4-2. Points which were inaccessible were evaluated using high resolution Imagery available on Google Earth as shown in Plate 4-1.
Plate 4-1: Random Points Generated for Performing Accuracy Assessment for Image
Classification.
Source: Modified from Google Earth (2017)
86
Plate 4-2: Ground Verification (Ground Truthing) using a Global Positioning System
4.4.1 Classification Scheme
For this study, a Priori identified six land use and land cover classes which were subsequently used in the classification scheme. These were: (a) Forest, (b) Small-Scale
Agriculture, (c) Large-scale Agriculture, (d) Urban and Settlement, (e) Water, and (f)
Grassland as shown in Table 4-2.
87
Table 4-2: Classification scheme for the classification and change detection procedures
Indicator Class Name Description
Vegetation Forest Areas under forest mainly on the higher altitudes to the
north-west of both Ruiru and Ndarugu watersheds.
Grassland Areas vegetated with grass and herbaceous species, with
relatively low occurrence of shrubs.
Agriculture Small-Scale Areas under subsistence farming, typically less than 5
Agriculture hectares in size.
Large-scale Areas under commercial large-scale farming.
Agriculture
Built-up Urban and Areas characteristic of highly developed town/ urban
Areas Settlement areas, residential areas at the urban fringes, roads, and
other built-up areas.
Hydrology Water Dam reservoir, lakes, rivers and streams.
4.5 Water Quality Assessment and Macro Invertebrate sampling
Water quality was assessed from a total of 12 sites (6 in each watershed). Sites were selected based on land use classes derived from image classification as indicated in Table
4-3. Fieldwork was carried out four times: twice during the dry season (mid-January 2014 and 2015) and twice during the wet season (mid-April 2014 and 2015) under stable flow
88 conditions. The following physico-chemical parameters were measured in situ using an
Ecolab Multi-parameter Water Quality Meter: pH, temperature, conductivity, turbidity and
Biological Oxygen Demand (BOD).
To ensure consistency and avoid sample over-estimation due to seasonal differences, the snapshot sampling technique was employed for macro-invertebrate sampling. Sampling for macro invertebrates was done using a pond net (0.5 mm mesh size). Samples were collected from runs, riffles and pools from each station. Sampling was done for a standard three minutes by disturbing a 1m2 area for each microhabitat. Samples were sorted live in a white plastic tray and then poured into vials and preserved in 70% ethanol. In the laboratory, samples were processed and identified to genus level according to Macan
(1977), Merritt and Cummins (1996), Nilson, (1996, 1997), Quigley (1977) and Scholtz and Holm (1985). Taxonomic lists of species known to be present in Kenya were used for identification (Johanson, 1992; Mathooko, 1998). The score was obtained by calculating the average score per taxa given the sensitivity levels for each group of macro invertebrates.
The Mini Stream Assessment Scoring System (miniSASS) was used to determine the ecological health of Ruiru and Ndarugu rivers. MiniSASS is a tool developed in South
Africa’s Department of Sanitation (2002) and used to determine the river’s ecological health based on the composition and sensitivity of macro invertebrates to environmental modification. The macro invertebrate diversity, richness and abundance were determined for each sampling station and sampling occasion using the number of taxa, total number of individuals and relative abundance of each taxon. Relative abundance (R.A) was calculated as the proportion (percentage by numbers) of each taxon in a station
89
Table 4-3: Description of sampling sites for water quality assessment
Samplin Altitude Location Dominant Land use g Site (metres) RFD1 2231 0° 58.582'S, 36° 42.026'E Natural Forest
RFD2 2226 0° 58.793'S, 36° 41.690'E Natural Forest
RAD1 1655 1° 6.474'S, 36° 53.438'E Small scale Agriculture dominated by tea and coffee bushes RAD2 1971 1° 1.866'S, 36° 46.514'E Large-scale Agriculture (flowers) RUD1 1551 1° 8.257'S, 36° 57.270'E Urban areas and settlement
RUD2 1482 1° 8.495'S, 36° 57.325'E Urban areas and settlement
NFD1 2140 0° 54.858'S, 36° 44.628'E Natural Forest
NFD2 2135 0° 54.705'S, 36° 44.696'E Natural Forest
NAD1 1608 1° 1.464'S, 36° 58.139'E Small scale Agriculture dominated by tea and coffee bushes NAD2 1630 1° 1.666'S, 36° 56.724'E Large-scale Agriculture (flowers) NUD1 1500 1° 6.113'S, 37° 2.712'E Urban areas and settlement
NUD2 1487 1° 7.257'S, 37° 4.546'E Urban areas and settlement
90
Plate 4-3: Measuring Water Quality Parameters (top) and Water Meters (Turbidity Meter and DO Meter)
91
4.5 Impacts of land use on the Hydrology of Ruiru and Ndarugu Rivers
4.5.1 The Soil and Water Assessment Tool (SWAT) Model
SWAT is a model that is semi-distributed and can be applied at the river basin scale to project the impacts of land management practices on water, sediment and agrochemical yields in watersheds with varying soils, land use and other land use conditions over extended periods of time (Arnold et al., 1998).
Model characteristics
The SWAT model is based on the hydrologic cycle, which is centered on the water balance equation:
푡 SW푡 = SW0 + ∑푖=1(푅푑푎푦 − 푄푠푢푟푓 − 퐸푎 − 푊푑푒푒푝 − 푄푑푎푦) ...... Eqn (i)
Where SWt is the final soil water content (mm H2O), SW0 is the initial soil water content on day i (mm H2O), t is the time (days), Rday is the amount of precipitation on day i (mm
H2O), Qsurf is the amount of surface runoff on day i (mm H2O), Ea is the amount of evapotranspiration on day i (mm H2O),Wdeep is the amount of water into the deep aquifer on day i (mm H2O), and Qw is the amount of return flow on day i (mm H2O).
SWAT is designed to utilize the use of alternative data such as land use change, land management practices and climate to model such watersheds (Neitsch et al., 2005; Arnold et al., 1998). The model operates in geographical information system (GIS), making it convenient for definition of watershed features, storage, organization and manipulation of the related spatial and tabular data (Di Luzio et al., 2002). The model also runs with minimum data inputs, and this is advantageous in areas where data is scanty or scattered.
SWAT has a strong computational efficiency and can model large basins with relatively
92 small computational resources and time. The model application runs in six main steps, namely (1) model installation and data preparation, (2) sub-basin delineation, (3)
Hydrological Response Unit (HRU) definition, (4) parameter sensitivity analysis, (5)
Model calibration and validation, and (6) uncertainty analysis. The process is described in
Figure 4-3, Plate 4-4 and Plate 4-5.
93
Figure 4-3: SWAT Model flow chart showing the inputs, outputs and process
Source: Adapted from Mango et al., (2011)
94
Plate 4-4: Stages in SWAT Model Setup for Ndarugu Watershed- Stream
Definition and Watershed delineation
95
Plate 4-5: Stages in SWAT Model Setup for Ndarugu Watershed- Sub basin definition and HRU definition
96
4.6 Data Preparation
4.6.1 Land cover
Land use data was obtained from classification of Landsat imagery and preparation of the associated lookup tables required for the data to be usable in the SWAT platform.
Comparison of classified land use data was done with land use maps available from
WaterBase, a project of the United Nations University hosted by United Nations University
- Institute for Water Environment and Health (UNU-INWEH). The project’s aim is to advance the practice of Integrated Water Resources Management (IWRM) in developing countries (UNU-INWEH, 2016). In this platform, land use data set for Africa comes at
400m resolution and comes with look up tables that are required to run in SWAT model.
This dataset is widely accepted for use in the SWAT model in the hydrological modellers globally and was used to run the simulation (UNU-INWEH, 2016). Land use data was reprojected to projected coordinate systems (WGS 1984, UTM Zone 37) and clipped appropriately for use in the Ruiru and Ndarugu Watersheds.
4.6.2 Land topography data
Land topography data in the form of a Digital Elevation Model (DEM) was acquired from the United States Geological Survey’s (USGS) Earth Explorer website. The data used was a 90m DEM from Shuttle Radar Topography Mission (SRTM).The STRM DEM was reprojected and clipped before being used as one of the input files in the SWAT model as shown in Figure 4-4.
97
Figure 4-4: Altitude map generated from SRTM DEM used in the SWAT model
Source: Modified from Shuttle Radar Topography Mission (2014).
4.6.7 Soil data
The soil data was downloaded from Food and Agriculture Organization (FAO) of the
United Nations using Version 3.6, 2003 and presented in Table 4-4 and Figure 4-5. The spatial resolution for the soil dataset was 1:5000000, which is equivalent to a pixel size of
500metres (FAO, 2003).
98
Table 4-4: Soil Characteristics in Ruiru and Ndarugu watersheds
Dominant Soil (>40% of Texture class Slope class Soil Code Mapping unit) (dominant soil) (dominant soil)
Fo48-2ab Orthic Ferralsols 80 50
Bc14-2bc Chromic Cambisols 80 50
Nhz-2c Humic Nitosols 80 50
Tm10-2bc Mollic Andosols 80 50
Tm9-2c Mollic Andosols 80 50
Source: Food and Agriculture Organization (FAO) of the United Nations, Version 3.6,
2016)
Figure 4-5: Soil map showing extent of soils in the watersheds Source: FAO, 2003
99
4.6.8 Climate Data
Climate data was acquired from Kenya Meteorological Department and Climate Forecast
System Reanalysis (CFSR) available from http://globalweather.tamu.edu/, which hosts
SWAT-specific weather data including Temperature (°C), Precipitation (mm), Wind (m/s),
Relative, Humidity (fraction) and Solar (MJ/m2). Data from 4 weather stations (Ngewa,
Kairi, Ikinu and Kiambaa) in the area was downloaded. All spatial data were reprojected into a projected Coordinate System (UTM Zone 37S, WGS 1984). Daily data from 1st
January 1979 to 31st July 2014 (35 years) was obtained for the aforementioned parameters.
4.6.9 River Discharge Data
Monthly river discharge data were obtained for Ruiru gauging station (3BC8), and Ndarugu gauging station (3CB5) both located near Thika Highway as shown in Figure 4-6. The available discharge data for the Ruiru River gauging station ran from the year 2006 to the year 2015, while Ndarugu River ran from 2003 to 2015. However, there were missing data values for both rivers. Ruiru River was 84% complete while Ndarugu was 83% complete.
Because of the numerous missing values and the absence of daily discharge data, the weather generator module embedded in the SWAT programme, along with weather data were deemed better than creating discharge tables from this data.
100
Figure 4-6: Location of weather and Gauging Stations
4.6.10 Model Evaluation, Calibration, Validation and Parameter Definition
Most studies that employ SWAT to model hydrological processes use daily discharge data to calibrate and validate the model. In this study, available discharge data from Water
Resources Authority in Kiambu, was collected at monthly intervals, making the use of this data for calibration and validation difficult. Available discharge data for the Ruiru River gauging station spanned from the year 2006 to the year 2015, while Ndarugu River spanned from 2003 to 2015. There were missing data values for both rivers. Even at the monthly intervals, Ruiru River was 84% complete while Ndarugu was 83% complete. Because of the numerous missing values and the absence of daily discharge data, the weather generator module embedded in the SWAT programme, along with climate data from Climate
101
Forecast System Reanalysis (CFSR), were deemed better than creating discharge tables from available data. The use of CFSR dataset in the absence of daily discharge data has been evaluated by Dile and Raghavan (2014) and performed satisfactorily compared to conventional gauges. The following parameters were defined for use in modeling of selected hydrological processes.
Table 4-5: Parameters selected for water quality simulation in the SWAT model
SWAT parameter Description
SPCON Sediment re-entrainment parameter
SPEXP Sediment re-entrainment parameter
CH_EROD Channel erodibility factor
NPERCO Nitrate percolation coefficient
ERORGN Organic N enrichment ratio for loading with sediment.
ERORGP Organic P enrichment ratio for loading with sediment
4.7. Data processing and Analysis
As described above, several image processing techniques were employed, including unsupervised classification, supervised classification, accuracy assessment and development of a change matrix. In addition, comparative descriptive statistics, such as graphs were used to compare water quality parameters such as pH, temperature, conductivity, turbidity, and Biological Oxygen Demand (BOD). Kruskal Wallis test,
Principal Component Analysis and Discriminant Factor Analysis were used to compare
102 water quality parameters across land use types and seasons, while the MiniSASS was used to profile macro invertebrates in rivers. The Soil and water assessment tool was used to predict various water quality parameters in the two basins.
4.8 Scope and Limitations of the Study
The study assessed observed land use impacts and their implications on water quality through analysis of land use changes, measurement of in-situ water quality and utilization of a hydrological model. The study provides an opportunity for people working in the land planning sector, water conservation, soil conservation, nature conservation and agriculture to employ similar techniques in research and planning. The study emphasizes land use activities as the main factor determining the quality of water in rivers, and their impacts to their ecological health. The study’s methodology is designed to identify the relationship between land cover and land use factors with water quality and macro-invertebrate assemblages within specific land uses in the landscape. This will help in determining which land use practices pose significant risks to water quality in rivers.
The study limitations include the reliance of the study on observed variables only, since the area lacks past data on the same factors, especially on water quality and macro- invertebrates. Water quality assessment was limited to turbidity, pH, electrical conductivity, Dissolved Oxygen and Biological Oxygen Demand (BOD). The satellite data used in this study was medium-scale (30m resolution), and may not reveal land uses at finer scales. The determination of the temporal scale of assessment was also influenced by availability of cloud-free satellite imagery to minimize classification errors.
The hydrological model used in this study (SWAT) is designed to mimic hydrological conditions as much as possible in the real world. Although the model can be calibrated and
103 verified with real measurements on the ground, it requires sufficient, long-term monitoring data to do so. Available long-term water quality data is not always available, and this was experienced in this study. In addition, as with all models, models may not always be able to capture specific one-time events. In the case of hydrological processes, one-time events may include but not limited to; sudden point and non-point pollution events (e.g. release of nutrients in river channels through a burst sewer), sudden storms, sudden flood events amongst others.
Financial considerations limited the number of times that could be spent in the field to collect hydrological data. However, the study was designed in such a way that the data collected would be sufficient to overcome the challenges of financial constraints. In addition, the choice of methods was made in cognizance of financial limitations.
104
CHAPTER FIVE 5.0 RESULTS AND DISCUSSION The results of the study are presented and discussed within the context of the aims and objectives. Issues related to land use impacts on hydrological processes in the study area, while exploring the environmental and social implications of the same are evaluated. The outputs of the SWAT model are presented and compared with actual measured data.
Simulated water quality parameters are presented and their social and environmental implications discussed.
5.1 Land Use Types, Distribution and Changes
Landsat 7 and Landsat 8 data, both with a 30-metre resolution, were used to generate land use and land cover maps for the study area using supervised classification following basic principles of the USGS land use/land cover classification system (LULCCS) for use with remote sensor data level classification (Anderson et al., 1976). To avoid observer bias for a more accurate classification scheme, an Area of Interest (AOI) encompassing both watersheds was clipped and classified. An accuracy assessment using randomly generated points indicated in Figure 4-1 and ground truth data was performed to compute the quality of land use classification. The most recent image (2015) was used for this purpose because ground truth data was collected in December 2015 in combination with high resolution imagery available from Google Earth.
The supervised classification generated six (6) land use classes: (1) Forest composed of primary, secondary, plantations and on-farm woodlots, (2) Small-scale agriculture, consisting of areas under subsistence and cash-crop farming done at the family/ homestead
105 level, (3) Large-scale farming (commercial), (4) Urban and settlement areas, consisting of built up areas and human settlements, (5) water, and (6) Grasslands.
Figure 5-1: Random points generated for use in combination with ground truth data for accuracy assessment
Source: Modified from Landsat 8 image (2015)
Classification of Landsat imagery was performed in ArcMap, using the Maximum
Likelihood classifier. This classifier is based on two principles: (a) the cells in each class sample in the multidimensional space are normally distributed, and (b) Bayes' theorem of decision making process (ESRI Developer Network, 2016). For class-specific and overall accuracy estimates, a total of 60 samples were used. The resultant error matrix shown in
106
Table 5-1 gave a Kappa coefficient value of 84% while the overall classification accuracy for
the classification was 87%. The classes with the highest user accuracy (commission) were
urban and grassland, while those with the highest producer accuracy (omission) were Forests,
Large scale Agriculture and Water. Using ground truth data and high-resolution imagery from
google earth, the output of maximum likelihood was further refined to achieve a better
representation of land cover types in the study area.
Table 5-1: Land cover classification accuracies computed from ground truth
reference points over the 2015 maximum likelihood-classified image
Reference data Forest SSA LSA Water Urban/Settlement Grassland Row Total Commission Accuracy
Forest 8 1 0 0 0 0 9 0.89
SSA 0 13 0 0 1 0 14 0.93
LSA 0 0 10 0 1 0 11 0.91
Water 0 0 0 7 0 0 7 1.00
Urban/Settlement 0 0 0 0 6 0 6 1.00
Grassland 0 1 0 0 0 8 9 0.89
Column Total 8 15 10 7 8 8 52
Omission Accuracy 1.00 0.87 1.00 1.00 0.75 1.00
Overall Accuracy 0.87
Kappa Coefficient 0.84
The watersheds were then delineated using the SWAT interface and used to clip the classified
image. The resultant images encompassed the exact area covering each of the watersheds and
their corresponding land use and land cover types. In all the time steps, small-scale agriculture
dominated the landscapes of both Ruiru and Ndarugu Watersheds, with a range of 56.45 to
60.93% for Ruiru Watershed, and 54.02% to 55.96% for Ndarugu Watershed as shown in
Table 5-2.
107
Table 5-2: Land use characteristics in Ruiru and Ndarugu Watersheds, showing areas in square kilometers and percentage in brackets
2005 2010 2015
Ruiru Ndarugu Ruiru Ndarugu Ruiru Ndarugu
Forest 67.17 37.27 61.41 34.18 59.75 34.34
(18.26%) (16.20%) (16.70%) (14.86%) (16.25%) (14.93%)
Small-Scale Agriculture 207.62 128.77 222.39 130.28 224.11 124.31
(56.45%) (55.96%) (60.47%) (56.62%) (60.93%) (54.02%)
Large-scale Agriculture 74.92 28.13 63.31 27.02 57.16 27.45
(20.37%) (12.23%) (17.21%) (11.74%) (15.54%) (11.93%)
Urban and Settlement 13.54 25.13 18.03 30.03 21.33 33.81
(3.68%) (10.92%) (4.90%) (13.05%) (5.80%) (14.69%)
Water 4.32 1.47 0.67 0.94 0.73 0.57
(1.17%) (0.64%) (0.18%) (0.41%) (0.20%) (0.25%)
Grassland 0.24 9.32 1.99 7.64 4.73 9.61
(0.07%) (4.05%) (0.54%) (3.32%) (1.29%) (4.18%)
108
Ruiru 70 60 50 40 30 20 10 0 Forest Small-Scale Large-scale Urban and Water Grassland Agriculture Agriculture Settlement
2005 2010 2015
Ndarugu 60 50 40 30 20 10 0 Forest Small-Scale Large-scale Urban and Water Grassland Agriculture Agriculture Settlement
2005 2010 2015
Figure 5-2: Land use change in Ruiru and Ndarugu Watersheds (2005 to 2015)
A multivariate analysis of class distances between each pair of sequentially merged classes was used to produce a dendogram that revealed the relationship between land use classes and their means and variances in the three time epochs is shown in Figure 5-4. The diagram is graphically arranged to avoid crossing lines so that members of each pair of classes to be merged are neighbors in the diagram.
The Dendrogram tool uses a hierarchical clustering algorithm which computes the distances between each pair of classes in the input signature file developed during the signature development step in supervised classification. The tool then merges the closest
109 pair of classes and successively merges the next closest pair of classes and the succeeding closest until all classes are merged. After each merging, the distances between all pairs of classes are updated.
16
14
12
10
8 Ruiru Ndarugu
6
4
2
0 2005 2010 2015
Figure 5-3: Comparison of Urban and settlement in Ruiru and Ndarugu watersheds, showing a steady rise in this land use type between 2005 and 2015.
110
Figure 5-4: Dendrogram based on the hierarchical clustering algorithm, showing the different clusters from the three time epochs (2005, 2010 and 2015). A=Large-scale Agriculture; B= Small-scale Agriculture; C= Forest; D= Grassland; E= Urban and Settlement; F= Water
The cluster analysis in Figure 5-4 showed that the 12 sampling sites can be divided into three groups of relatively similar sites. By use of the most dominant land use types, the site groups were designated as forest-dominated sites, agriculture-dominated sites and urban-dominated sites as shown in Figure 5-4.
Accuracy assessment shown in Table 5-1 revealed that the produced land use classifications have agreeable accuracy levels, and therefore indicate that major land use categories and changes in the study area were accurately computed. An expansion of urban areas is attributed to a reduction in areas under large-scale agriculture, grasslands and forests along major infrastructural facilities (roads) to the east of the landscape as shown
111 in Figure 5-6. In these areas, grasslands decreased significantly. The pronounced change of these land use types is as a result of the urbanization process. For example, grasslands on these parts of the landscape have been observed to be ranging ground for wildlife, and have also been used by pastoralist communities as dry-season grazing grounds. The grasslands are located in the drier parts and have not provided the much-needed land for cultivation. Similarly, it was observed that large scale farms previously under coffee and sisal plantations to the east of the landscape have been sold off as commercial and residential plots, and are now being actively developed as residential areas. The fact that urban and settlement areas mainly increase in the eastern part of the catchment may be due to better accessibility (e.g. development of Thika super highway).
In addition, being a relatively drier area, the alternative to use the area for settlement than agriculture is reasonable. An increase in built-up areas can lead to an increase of the water yield and decrease in evapotranspiration as has been found in several studies (e.g. Im et al.,
2009; Wijesekara et al., 2012, Wagner et al., 2013). Numerous valleys in the area classified as large-scale farms are under water reservoirs. These reservoirs account for a large proportion of areas under water. From the analysis of classified imagery, a general shift from other land use types to urban and settlement is being observed in both watersheds as indicated in Figure 5-2. This pattern is strongly consistent with results of Thuo (2013) as well as Kiio and Achola (2015). Figure 5-5 shows how land use/ land cover shifted between
2005 and 2015 amongst four dominant land use types. Parts of the study area and particularly the lower reaches of Ruiru and Ndarugu basins form part of the Nairobi-Thika urban-rural fringe. The rural-urban fringe, or hinterland, is sometimes described as the landscape interface between town and country, and is often characterized by a mix of urban
112 and rural landscape characteristics. Rapid urbanization of the rural-urban fringe has led to new income opportunities for the people who originally worked in farms as farmers or labourers (Thuo, 2013).
A visual analysis of high resolution NASA and Digital Globe Imagery from Google Earth show that urban land use replaced areas previously occupied by sisal plantations and grasslands as shown in Plate 5-2. An increase in stone quarrying activities is also an indicator of high demand for building materials. Immigrating population from other areas presents the local residents with new opportunities, such as establishing of businesses and construction of rental houses to accommodate the immigrant population. In addition to new opportunities, residents have interacted with new comers who have brought in new technology and skills. Thuo (2013) observes that as a result of land use change, small- holder farming systems have been negatively affected as there is less availability of labour to work in the farms. There is also a general shift from growing of traditional food crops to crops such as kales, spinach and tomatoes which have a ready market among the residents. Traditionally, the area has undergone land sub division to smaller pieces of land.
Consequently, this new mode of farming is suitable for the smaller land parcels. Evidence of surface water scarcity in the area is already visible, with numerous boreholes that have been sunk. The Ruiru-Juja Water and Sanitation Company (RUJUWASCO) rations water as part of the management strategy to curb the problem of surface water availability. For example, information obtained from RUJUWASCO indicate that most estates in Juja Town receive water from the company once or twice a week, and residents have to acquire water storage facilities to cater for the days when water is not supplied.
113
The increasing population is leading to pressure for conversion of more farmland to residential areas (Figure 5-5). In future, opportunity for farmers to increase production by buying additional parcels of land in their locality will be unavailable. Demand for high value horticultural produce by urban consumers can stimulate production by small farmers
(Tacoli, 2002), but expansion of urban centres leads to competition over the use of essential natural resources, particularly land and water.
Changing land use systems have an influence of water flows in streams (Stonestrom et al.,
2009). For example, DeFries and Eshleman (2004) showed that land use changes could result in an increase of water shortage as a result of decreased flow in rivers especially in water-scarce countries. In addition, conversion of forest land to agriculture may result in changes in climatic patterns at the local and regional scales, which could affect the flow regime in rivers.
114
120
Ruiru 100
80 Grassland Water
60 Urban and Settlement Large-scale Agriculture 40 Small-Scale Agriculture Forest 20
0 2005 2010 2015
120 Ndarugu 100 Grassland
Water 80 Urban and 60 Settlement Large-scale Agriculture 40 Small-Scale Agriculture 20 Forest
0 2005 2010 2015
Figure 5-5: Land use change between 2005 and 2015 showing increasing urban areas and reducing agricultural areas in Ruiru and Ndarugu Watersheds
115
Plate 5-1: High Resolution Imagery of 2003, showing large parts previously under sisal in Juja (A) and stone quarrying B
Source: Google Earth (2016).
Plate 5-2: High resolution imagery showing the shift of urban land use replacing areas previously under sisal in Juja (A)- changing to settlement, and (B) Stone quarry expansion
Source: Google Earth (2016).
116
Plate 5-3: Coffee Plantations in Ruiru Watershed
Plate 5-4: Small-scale farming in Ndarugu watershed
117
Land use conversion is a global issue, and as revealed by this study, agricultural land is gradually being replaced by urbanization in the study area. In the short term, conflicts related to land acquisition and resistance of the local residents along with cultural shock are being experienced. In the long term, a serious food crisis is expected. In most African countries, agricultural production is achieved through extensification, or the conversion of more land for farming. Although agricultural intensification (use of more inputs such as fertilizers, technology and management) is taking shape, few farmers have the technical know-how, technology access and financial resources to adopt the new trend.
118
Figure 5-6: Land use and land cover characteristics of Ruiru and Ndarugu watersheds in
2005, 2010 and 2015
119
20
18
16
14
12
10 Ruiru Ndarugu 8
6
4
2
0 2005 2010 2015
Figure 5-7: Area under forest cover showing a decline between 2005 and 2015 in both
Ruiru and Ndarugu watersheds
In the Ruiru watershed, there was a general decline in forest cover and large-scale agriculture, while small-scale agriculture and urban and settlement areas increased. An increase in the coverage of urban and settlement category was observed, while there is a decline of small scale farms and grasslands. Ndarugu watershed on the other hand experienced a decline in forests and small-scale agriculture, an increase in urban and settlement areas, while large-scale agriculture remained almost the same. Although the area under water in both watersheds was small, a sharp decline in these areas was observed over the three time periods. This can be attributed to an increase in the number of built-up areas
120 which is faster in the Ruiru watershed due to a higher population and closer proximity to
Nairobi city, fuelling a demand for land and water resources.
Water 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 2005 2010 2015
Ruiru Ndarugu
Figure 5-8: Sharp decline in areas under water, indicative of pressure on water resources in the watersheds of Ruiru and Ndarugu
To project future scenarios on the observed land use changes, projections into the year
2020 were computed using the least squares method in the LINEST function of Microsoft
Excel. The projections were plotted on a bar graph as shown in Figure 5-9. These projections reveal that areas under forest, large scale agriculture and water will continue to decline as they are replaced by small scale agriculture and urban and settlement areas. In
Ndarugu watershed, all land use types will decline as they are replaced by areas under urban and settlement land use. Future policy review should focus on land management practices that address sustainability of land use systems for improved watershed management. These include the adoption of Green-Blue infrastructure.
121
70 60 50 40 30 2015 20 2020 projections 10 0 -10 Forest Small-Scale Large-scale Urban and Water Grassland Agriculture Agriculture Settlement
60 50 40 30
20 2015 10 2020 projections 0 Forest Small-Scale Large-scale Urban and Water Grassland -10 Agriculture Agriculture Settlement
Ndarugu
Figure 5-9: Projections of land use changes for year 2020 for Ruiru and Ndarugu watersheds
5.2 Land Use Impacts and seasonal Variations on Surface Water Quality
By use of the classified land use imagery, three site groups were identified at the outlet of the sub-basin levels and designated as forest-dominated (FD), agriculture-dominated (AD) and urban-dominated sites (UD) as shown in Figure 4-2. Small-scale and Large-scale agriculture land uses were combined because these classes were assumed to produce
122 similar outputs in rivers compared to other land use and land cover characteristics. In addition, multivariate analysis from the dendogram showed close similarity between the two land use classes in the 2005 and 2010 image as shown in Figure 5-4. The differences in the 2015 image are attributed to the fact that this image was derived from Landsat 8, while the previous two were derived from landsat 7, both of which are known to have some technical differences. To minimize the impacts of these technical differences, careful selection of spectral bands to be used during classification was done such that bands with similar spectral ranges were used in all time-steps, and classification was done repetitively until refined maps had been produced. Water and Grassland land uses were proportionally very small and were therefore not considered during the site group selection (Ranging from
0.18% to 1.17% for water; and 0.07% to 1.29% for grassland in both watersheds). The forested groups were in the upper reaches of both watersheds (4 sites), the agricultural group in the middle (4 sites) while urban (4 sites) were largely in the lower parts of the watersheds. This distribution also followed an altitude gradient that are determined by agro-ecological zones which influence the type of land use or land cover system that occurs in an area. Forested sites occurred at 2135m to 2231m, agricultural 1608m to 1971m and urban 1482m to 1551m as indicated in Figure 5-10.
123
2500
2000
1500
1000
Altitude(Metres) 500
0 FD AD UD
Ruiru Ndarugu
Figure 5-10: Distribution of sampling sites along the altitudinal gradient (FD-Forest
Dominated, AD-Agriculture Dominated, UD-Urban Dominated).
The concentrations of measured water quality parameters in the dry and wet seasons across sampling sites are shown in Table 5-3. The Kruskal-Wallis tests revealed significantly higher values of DO (H = 24.71, p < 0.01) and EC (H =7.98, p < 0.01) in the dry season than in the wet season. Temperature (H = 6.92, p < 0.01) was significantly lower in the wet season than in the dry season. The concentrations of pH and Turbidity showed no significant differences in the two seasons. Multiple comparisons of water quality parameters was revealed wide variations between Urban, Agriculture and Forest groups as seen in Figure 5-11. To test the significance of these variations, principle component analysis and multiple factor analysis were performed, with results presented in Figure 5-
12.
124
Figure 5-11: Mean ± standard error values for the water quality parameters among three land use-based site Groups in Ruiru and Ndarugu Watersheds
125
Table 5-3: Summary of surface water quality between the dry and wet seasons in the
Ruiru and Ndarugu Rivers, Kiambu County
Parameter N Dry Season Wet Season Kruskal Wallis p value
Mean± SD Mean± SD H
TEMP 48 17.9±2.35 16.27±1.87 6.92 0.009
DO 48 11.30±1.28 15.77±3.07 24.71 0.000
EC 48 75.40±42.82 102.26±42.36 7.98 0.005
TURB 48 23.94±18.40 36.99±28.52 2.65 0.103 pH 48 6.81±0.55 6.75±0.60 2.08 0.1488
Results of the Principal Component Analysis (PCA) revealed that three components explained 77.7% of the variance: Component 1 (32.5%), Component 2 (26%) and component 3 (18.8%). Component 1 distinguished temperature, component 2 distinguished pH, Turbidity and electrical conductivity while component 3 distinguished dissolved oxygen (Figure 5-12).
The variability in dissolved Oxygen, Electrical Conductivity and Temperature across land use types are explained by both point and non-point sources of pollution in urban and agricultural areas, which may have been caused by point sources which may include but not limited to sewerage effluent.
126
Figure 5-12: PCA analysis of water quality parameters across land use systems in Ruiru and Ndarugu basins, Central Kenya. The first two axis components 58.5% of the variation: Component 1 (32.5%) and Component 2 (26%).
According to Moldan and Cerny (1994), pH tends to vary with season and usually declines with increasing discharge. This agreed with the results of this study for agriculture and urban sub watersheds, which showed that pH, in both watersheds, were lower in the dry season. In addition, discriminant analysis (Figure 5-13) showed that pH was correlated largely with urban and agriculture land use, whereby the urban land use asserted the strongest influence. However, Sliva and Williams (2001) reported that the dynamics of the acid-neutralizing capacity of surface waters are most likely too complex to be determined by the few predictors (e.g. land use and seasonality alone), and further studies are required to explain this variability more deeply
127
Figure 5-13: Discriminant analysis on measured water quality parameters across land use types on two canonical axes showing water quality parameters and their associated land use types.
The study results show that land use and land cover has an impact on water quality parameters. The land use changes found in the study area, and particularly an increase in urbanization and agricultural areas, are anticipated for the highland areas of Kenya due to an increasing population as shown in Table 3-1.
The results of this study indicate that, using land use and seasonal variables, agricultural; and urban land uses are more important predictors of water quality (Figure 5-13). This relationship may be as a result of both point sources (e.g. storm water run-off, sewage leakages) and non-point sources (e.g. agricultural runoff) of pollution typically associated with urban land use and respectively (Sliva and Williams, 2002; Osborne and Wiley, 1988).
Results indicate that land use and land cover have significant impacts on water quality in
128 the two basins. Forested areas have better water quality than either agricultural dominated and urban dominated landscapes in both wet and dry seasons as shown in Figure 5-13.
Urban landscapes are associated with pollution from diverse sources, including domestic sources (e.g. human and animal waste) as well as industrial sources. Agricultural landscapes contribute chemical pollution from farm inputs, including pesticides and nutrient enrichment from fertilizers. Conversely, in forested landscapes, low anthropogenic activities means that there are fewer or none pollution sources.
Plate 5-5: The Ruiru River at Thika Road Bridge, showing urban land use and agricultural land use (small scale) adjacent to the river.
The macro invertebrate fauna of the Ruiru and Ndarugu basins is relatively poor as indicated in Table 5-4 in both species and numbers compared to other sites in Kenya (e.g.
129
Kibichii et al, 2007). In the two basins, Odonata (dragonflies and damselflies) were the most diversified group with six genera, followed by Hemiptera (true bugs) and Diptera
(flies) with three and two genera respectively. The generally poor status of macro- invertebrates is attributed to high modification of the landscape, leading to low abundance of aquatic insects. Leaching from agricultural land may lead to increased levels of organo- phosphate content in the rivers, which can influence macro invertebrate abundance.
Pollution of these rivers may also have contributed to low macro-invertebrate counts. This is attributed to agricultural runoff, sedimentation as well and domestic and industrial effluent.
130
Table 5-4: Macro invertebrate richness at sampled sites
Order Ndarugu Total Ruiru Total Family
NAD1 NFD1 NFD2 NFD3 NFD4 NUD1 NUD2 RAD1 RAD2 RFD1 RFD2 RUD1 RUD2 Annelida 1 1 1 1 Coleoptera 1 1 1 1 2 Dytiscidae 1 1 1 1 2 Decapoda 1 1 1 1 Diptera 1 1 1 1 2 Chironomidae 1 1 2 Simuliidae 1 1 Ephemeroptera 2 1 1 4 Baetidae 2 1 1 4 Hemiptera 1 1 1 2 5 1 1 1 3 Belostomatidae 1 1 1 3 1 1 1 3 Naucoridae 1 1 Notonectidae 1 1 Mollusca 1 1 Odonata 2 2 4 1 1 1 1 1 2 7 Aeshnidae 1 1 1 1 2 Coenagrionidae 2 2 1 1 1 3 Corduliidae 1 1 Lestidae 1 1 Libellulidae 1 1 Synlestidae Trichoptera 1 1 Hydropsychidae 1 1 Total 5 4 1 1 1 2 4 18 3 1 2 2 6 3 17
131
Table 5-5: MiniSASS Sensitivity Scores for Ruiru and Ndarugu Rivers
Order MiniSAS NA NF NF NF NF NU NU RA RA RF RF RU RU
S Score D1 D1 D2 D3 D4 D1 D2 D1 D2 D1 D2 D1 D2
Annelida 2 2 2
Coleoptera 5 5 5
Decapoda 6 6 6
Diptera 2 2 2 2
Ephemeroptera 5 10 5 5
Hemiptera 5 5 5 5 10 5 5 5
Mollusca 4 4
Odonata 4 8 10 8 6 6 4 12 18
Trichoptera 9 9
Total 23 24 2 6 5 7 20 18 6 11 10 30 20
ASPT(Average 7.7 8.0 2.0 6.0 5.0 3.5 6.7 6.0 6.0 5.5 5.0 5.0 10.
Score Per Taxa) 0
Moldan and Cerny (1994) demonstrate that most seasonal variations in river water chemistry are driven by climatic and biotic factors and are therefore largely governed by the processes that are taking place in the terrestrial part of the watershed such as natural or human induced vegetation cover changes.
In this study, land use and water quality interactions are better explained by interactions in the dry season as compared to the wet season, an observation attributed to the impact of higher run off (hence discharge) in the wet season (Elwood et al., 1983). Long term water quality data from WRA shows higher concentrations of nutrients in the wet season compared to the dry season, which also coincides with the period of fertilization of agricultural fields in the area. Wiley and Osborne (1998) observed that nutrients are easily transported to the channels via surface runoff and subsurface flows during periods of soil
132 fertilization. In this study, catchment-scale land use has been shown to have an impact on water quality parameters. Several studies have attempted to describe whether near-stream land use is of a greater influence to water quality as opposed to catchment-wide land use.
Although this study used the later approach, stream bank cultivation as well as development was observed in both watersheds as seen in Plate 5-6. Results from other studies show mixed outputs depending on the location, land use types and seasonal differences. For example, Sliva and Williams (2002) used a redundancy analysis and GIS to demonstrate that catchment-level land use aspects have a greater influence on water quality compared to near-stream land use in the Greater Toronto Area, Ontario, Canada. In contrast, Hunsaker and Levine (1995) found that when landscape was classified at the level of the whole catchment, the relationship between land use and water quality was distinctly stronger than if only a 200 or 400m buffer strip was considered. Similarly, Johnson et al.,
(1997) found that the whole catchment explained slightly less of the water quality variability than their 100m buffer. These studies demonstrate that both catchment-wide management of watersheds and stream-bank management are important factors to be taken into consideration to improve water quality in streams and rivers. For example, it is important to determine an adequate riparian zone along rivers and streams, as they provide a buffer that would be effective in mediating pollutant loading since it is affected by the spatial variations in physical, ecological and land use conditions within the streamside areas of the watershed. Indeed, riparian zone conservation is a legal requirement in most countries. In Kenya, the management of riparian zones is entrenched in the Water Act
(2016), in which the National Water Resource Strategy is established to define riparian areas for existing water resources and also provides measures for their protection,
133 conservation, control and management of water resources and approved land use for the riparian areas.
Xiang (1995) used GIS and modeling techniques to determine that, for one small Southern
Carolina watershed, the width of an effective buffer varied from 8 to 175m
Plate 5-6: Streambank cultivation (foreground) and development (background) of the
Ndarugu River
The variation in physico-chemico characteristics in Ruiru and Ndarugu basins were found to vary across different land use and land cover types, depending on season. The relatively high variation between water quality parameters associated with different land use types are characteristics that can be used to inform or predict water quality aspects in rivers that lack data. In recent years, unusual weather events have been observed in the region, where occasional heavy rains subject the area to flooding. In 2015 alone, at least three flooding events were recorded in Kiambu County. In June 2015, residents of Gitambaya village of
134
Ruiru constituency in Kiambu County lost property of unknown value due to sudden flash floods (Mediamax, 2015). In December 2015, twenty families were displaced in Ruiru’s
Membley estate following a sudden rain event (Mediamax, 2013). In November 2015, ten people were killed in Nyakibai, Gatundu North, Kiambu County, after their vehicle was swept off following sudden rain (Standard media, 2015). In the dry season, farmers in the study area have been practicing irrigation using water abstracted from the rivers and streams. As areas under agriculture are gradually being taken up by urbanization, agricultural intensification is observed. This includes increased farming activities in the dry season, which abstracts water from the rivers. Urbanization on the other hand, is known to result in more runoff during rainy season due to the increase of paved surface area
(Cunderlik, 2002).
In the study area, and most other regions in the country, agricultural activities are strongly tied to seasons. The onset of the long rains triggers a flurry of farm-based activities. The application of farm-based inputs, including fertilizers and chemicals, are also expected to follow this seasonal pattern. Agricultural activities alter hygienic and aesthetic aspects of water quality (Kithiia, 1992). Thus, this may aggravate the imbalance between water quality parameters and lead to seasonal water quality variations. Agricultural activities increase suspended solids, turbidity and colour, nitrates and phosphates from fertilizer application and animal wastes (Kithiia, 1997).
In this study, urbanization has been identified as the main driver of change in the study area. There is a general increase of urban/settlement areas more to the east than to the west in both watersheds. The eastern parts are generally drier, and settlement areas are cropping up there. The development of Thika Superhighway appears to have triggered urbanization
135 on these parts than elsewhere in the watersheds. Population projections in table 3-1 suggest that the towns of Ruiru and Juja are expected to continue growing. Figure 5-5, Plate 5-1 and Plate 5-2 reveal conversion of other land use types to settlement and urban areas.
To the extreme west of the watersheds lies the Gatamaiyu Forest, a protected forest that forms the southern-most part of the Aberdare Ranges, one of the key water catchments systems in Kenya. A buffer zone of the Nyayo Tea Zones protects the forest against human encroachment. This area is expected to remain unchanged as shown in Figure 5-5.
An increased demand for food is expected due to population growth as shown in table 3-1 and a decreased supply of food due to decreased cropland is expected. Figure 5-6 and Table
5-2 show a decrease in areas under small scale and large-scale agriculture and an increase in settlement and urban areas. This is seen as a consequence of rapid development in the area that may lead to conflicting land uses as described by Thuo (2013).
In water bodies, dissolved oxygen is associated with a response to biological activity, where higher levels are associated with presence of aquatic plants (Moyo, 1997; Schneider et al., 2000). In this study, measured levels of dissolved oxygen are higher in the wet season than in the dry season in all the land use types as shown in Figure 5-11. Although this trend is consistent with observations from other studies (e.g. Schneider et al., 2000), some studies have recorded lower DO levels in the dry season than in the wet season. For example, in a study conducted in lowveld sand river system in South East Zimbabwe, dissolved oxygen concentration levels were higher in the dry season compared to the wet season
(Tafangenyasha and Dzinomwa, 2005), and this was attributed to the loss of photosynthetic aquatic plants by faster current. However, the study was conducted in a semi-arid region, where ecological patterns could be different from areas that are wet, such as in the study
136 area for this research. Other factors that could result in variation in DO levels may include oxygen depletion resulting from the eutrophication of natural water that receives excessive amounts of nutrients normally limiting plant growth. Such nutrients are sometimes associated with urban land use (Schneider et al., 2000). In this study, urban/ settlement land use recorded lower levels of DO as indicated in Figure 5-11compared to both agriculture and forested land in both the dry and wet seasons. This is associated with sewage effluent from urban and settled areas. As the human population increases due to immigrating population from other areas, pressure is exerted on the current sewer infrastructure, which is already overstretched.
In all sampling stations, pH did not deviate from the 6.1 to 8.2 range expected for most surface freshwater systems, but seasonal variations in different land use types were observed. In the dry season, pH was higher in forests compared to Urban/settlement land uses and Agriculture. The wet season recorded higher pH levels in Urban/settlement areas and Agriculture, and lower values in Forest as shown in Figure 5-11. This suggests that although land use and seasonal changes may be a factor that influences pH, other external factors may also be major determinants of pH measurements in rivers. A number of factors may affect the proportion of major ions (hence pH) in a landscape, including geological, atmospheric, biological and anthropogenic activities (Fawzi et al., 2002). For example, consumption of carbon dioxide by aquatic plants results in an increase in pH during photosynthesis, while during respiration and decomposition, released carbon dioxide results in decreased pH (Schneider et al., 2000).
Theoretically, biological activity and human activity would be the primary determinants of pH, whereby the human activities trigger nutrient loading while biological activity plays a
137 secondary role (Tafangenyasha and Dzinomwa, 2005). The Ruiru and Ndarugu basins are enriched with nitrate while passing through agricultural dominated areas, and then they receive sewage effluent rich in phosphate from urban /settlement areas could support algal blooms. Data accessed from the Water Resources Management Authority (WRMA)
Kiambu regional offices shown in Appendix II and III show varying levels of Phosphates and Nitrates.
The implications of sudden heavy rains (which have been observed severally during the course of the study) in the Ruiru and Ndarugu watersheds are currently unknown. It is however expected that under such conditions, both turbidity and conductivity may be affected. Sudden rains may lead to flooding, which may add more insoluble material than soluble electrolytes. This could cause a dilution effect, which may either increase or decrease water quality depending on the land use, time, season and external factors.
Significant declines in salinity have been shown during high runoff periods associated with sudden rain events in other studies. For example, Russell (1999) showed short-term declines in salinity during high runoff periods, which suggests that dry season pollution events of waterbodies should be of concern.
Turbidity was highest in agricultural dominated sites and lowest in forest sites as Figure 5-
11 indicated. In forested sites, turbidity levels were minimal between the wet and dry seasons, but strongly variant in agricultural and urban/settled sites. Surface runoff, and its erosive action, is attributed to an increase in turbidity in the wet season in both agricultural areas and urban/settled areas. This periodic input of sediment into the rivers leads to decreased water clarity, thereby inhibiting light penetration and leading to reduced
138 biological activity. Highly turbid waters have more suspended solids and have been shown to be prone to oxygen depletion (Vesilind et al., 1994).
As shown by this study, the wet season increases erosive conditions in agricultural and urban/settled areas as rains and flood waters increase. This may lead to an overshoot on the recommended threshold of turbidity levels in surface water. The influence of seasons and land use may be compounded by other factors, such as the composition of and solubilities of materials in the rock, soil, primary production and inflows that the water flows through
(Tafangenyasha and Dzinomwa, 2005).
The Ndarugu river watershed straddles the towns of Kamwangi, Gatundu and Juja, while the towns of Ruiru, Kiratina and Githunguri are found within the Ruiru river watershed.
These towns produce large quantities of sewage, as some of them have high human population density. Ruiru town has the highest population in the entire Kiambu County, with 238,858 people (KNBS, 2009). Juja, Githunguri and Gatundu have 40,446, 10,007 and 5,550 people respectively. Organic pollution is known to lead to a decrease in species but an increase in individuals. In extreme conditions, survival of macro-invertebrates becomes impossible (Jackson and Jackson, 1998; Marshall, 1972). Septic conditions were recorded in at least six locations along the rivers (four in Ruiru River and two in Ndarugu), although all of them were not amongst the study sites.
Both point and non-point pollution sources from agricultural activities, sewage discharge and industrial discharge may increase nutrient levels in the water, including nitrogen, nitrates and phosphorous. Long term water quality monitoring data from Water Resources
Management Authority from monitoring stations of the two watersheds reveal that the
Ruiru and Ndarugu rivers are at acceptable levels for water quality as shown in Table 5-6.
139
However, higher than recommended levels in some instances as shown in appendix 4 may affect the quality and distribution of benthic macroinvertebrates.
Table 5-6: Mean and maximum values of water quality parameters recorded in Ruiru and
Ndarugu monitoring stations
Parameter NEMA Ruiru Ndarugu maximum
allowable value Mean Max Mean Max
pH 6.5 – 8.5 6.9 8.5 6.6 7.8
Suspended solids 30 (mg/L) 56.8 150 46.6 175
Nitrate-NO3 10 (mg/L) 1.95 2.94 2.62 9.67
Nitrite –NO2 3 (mg/L) 0.02 0.06 0.01 0.04
Total Dissolved 1200 (mg/L) 43.84 132 39.5 91 Solids
Fluoride 1.5 (mg/L) 0.2 0.29 0.19 0.29
Source: Water Resources Authority, Kiambu, (2016).
The Water Resources Management Authority (WRA) is mandate by the government of
Kenya under Section 7 of the Water Act (2016) as the lead agency in management of water resources in the country. Its role is to regulate, manage and equitably allocate the national water resources to all users. As revealed in this study, the growing human population and associated urban expansion, coupled with agricultural intensification, continues to exert
140 pressure on the scarce water resources. As the Country continues to develop, so does the water demand. WRMA’s mandate to the right to adequate and clean water to all citizens will be met by a number of challenges, some of which have been identified by this study.
WRMA embraces the principles of Integrated Water Resources management (IWRM) as provided in the Water Act of 2002 (e.g. Part IV, which provides guidelines on water supply). The Global Water Partnership (2000) defines IWRM as a “process which promotes the coordinated development and management of water, land and related resources in order to maximize the resultant economic and social welfare in an equitable manner without compromising the sustainability”. In this arrangement, the government role is of an overseer, and its major responsibilities should be to: a) formulate national water policies; enact water resources legislation; ensure separation of regulating and service provision functions; encourage and regulate the private sector; and encourage dialogue with neighbouring countries. IWRM call for practical management instruments, water resources assessment, communication and information, allocation and conflict resolution as well as formulation of regulatory systems.
Despite the existence of these mechanisms as provided for by IWRM and the Water Act of
2002, there are still problems that have been identified in the study area. For example, over- exploitation of rivers has been shown to lead to inequitable water allocation for domestic uses. These problems are not however unique to Kenya. Ngana et al., (2004) concluded that unsustainable utilization and management of natural resources (land, pasture and water resources) were the core problems in Tanzania that led to environmental degradation such as shortage of land and decreased river flows in Lake Manyara sub-basin; while Singh
(2000) showed that in the Haryana province of India, environmental degradation occurred
141 as a result of over exploitation of groundwater which caused an attendant increase of the shallow water table height of over 1m/a since 1985, culminating in water logging and attendant floods.
4.3 Watershed Modelling 4.3.1 Precipitation and Stream flow Long term discharge data for Ruiru and Ndarugu rivers are given in Appendix II. The monthly measured observations from WRA data and simulated observations from SWAT model shown in Figure 5-14 showed that the SWAT model mirrored the measured monthly precipitation closely, and chi square goodness of fit test was significant (ꭕ2 (1, N = 24) =
1.15-E65, p < 0.05).
250
200
150
100
50
0
Simulated PPT Measured PPT
Figure 5-14: Performance of SWAT Model under recorded (2002 to 2012) and simulated (2005 to 2014)
142
Figure 5-15 shows two main categories of flow regime between 2005 and 2014 as simulated by the SWAT model. The figure depicts a relatively lower but stable flow regime of the Ruiru River between 2005 and 2011, and a marked increase in flow between 2012 and 2014. The increase in flow does not follow a bimodal pattern (the normal climate scenario in the study area as shown in Figure 2-7) except in 2012, suggesting changes in climatic patterns and land use processes in the catchment.
Figure 5-15: Variation in OutFlow regime on Ruiru Basin as simulated by SWAT model
143
Figure 5-16: Variation in OutFlow regime on Ndarugu Basin as simulated by SWAT model
144
5.3.2 Water Temperature
Comparison between observed and simulated water temperature revealed that the model results mirrors observed results. Chi square goodness of fit was significant for Ndarugu watershed (ꭕ2 (1, N = 22) = 0.000282, p < 0.05) and not significant for Ruiru watershed
(ꭕ2 (1, N = 26) = 0.562570393, p < 0.05). However, observed records were significantly higher than simulated values in Ruiru watershed (by 28%) and Ndarugu watershed (by
15%) as shown in Figure 5-16. This is attributed to insufficient long-term data to calibrate the model. For calibration, the model requires at least four years of daily time-step data
(available data from WRA spans four years, and it is collected monthly).
Ruiru (oC)
30 20 10
0
1/12/… 2/12/… 3/12/… 4/12/… 5/12/… 6/12/… 7/12/… 8/12/… 9/12/… 1/12/… 2/12/… 3/12/… 4/12/… 5/12/…
11/12… 12/12… 10/12… 11/12… 12/12…
Observed Simulated
Ndarugu (oC)
30 20 10
0
1/14/… 2/14/… 3/14/… 4/14/… 5/14/… 6/14/… 7/14/… 8/14/… 9/14/… 1/14/… 2/14/… 3/14/… 4/14/… 5/14/…
11/14… 12/14… 10/14… 11/14… 12/14…
Observed Simulated
Figure 5-16: Observed and Simulated Temperature records for Ruiru and Ndarugu Rivers
145
5.3.3 Dissolved Oxygen
Simulated mean monthly dissolved oxygen is shown in Figure 5-17, revealing marked reduction but stable conditions between 2011 and 2014 and higher DO levels in Ndarugu watershed. Data from WRA for DO ranged from 1.74mg/l to 7.1mg/l in Ruiru watershed, and 2.3 mg/l to 6.0 mg/l in Ndarugu watershed
Figure 5-17: Simulated mean monthly dissolved oxygen for Ruiru (top) and Ndarugu (bottom)
Rivers
5.3.4 Sediment
The mean monthly sediment inflow as simulated by SWAT in Ruiru and Ndarugu rivers is shown in Figure 5-18 and 5-19. As with other simulated results, sediment flow peaked in
April 2013 for both watersheds. The high levels of sedimentation in the rivers are also shown in Plate 5-7 revealing their impacts on existing water infrastructure in Ndarugu
River.
146
Figure 5-18: Mean monthly sediment inflow in Ruiru River
147
Figure 5-19: Mean monthly sediment inflow in Ndarugu River
148
Plate 5-7: Sedimentation in Ndarugu Water Treatment Plant, showing the impacts of watershed sediment transport on water treatment infrastructure
149
4.3.5 Nitrite and Nitrate
Simulated Nitrite and Nitrate levels are shown in Figure 5-20. Measured values for Nitrite ranged between <0.01 to 0.06mg/l for the Ruiru river and <0.01 to 0.04 for Ndarugu river; while Nitrate levels ranged between 0.01 to 5.33 and 0.01 to 9.67 respectively.
Figure 5-20: Simulated Nitrite and Nitrate levels in Ruiru and Ndarugu River
Despite the erratic nature of long-term river monitoring data in both Ruiru and Ndarugu rivers, the SWAT model is known as a reliable estimate for watershed modelling, and is acceptable in the scientific community to estimate hydrological variables (e.g. Grey et al.,
2014). Available long-term monitoring data in the two monitoring stations is collected monthly and spans a total of four years. In some instances, it is often incomplete due to faulty or non-functional equipment. The use of models to provide planners with critical information is therefore a reasonable approach, but they require reliable inputs for them to
150 be verified and ensure that the simulations are correct. Like with most simulations, even when they are unrealistic, they can be used for planning for resources and performing of experiments. The need for Water Resources Authority (WRMA) to upscale their efforts to have more sustained water quality data cannot be over-emphasized.
In this study, the SWAT simulation revealed that the most profound effect of watershed processes is as a result of increased precipitation between 2011 and 2014. The higher than normal rainfall in this period is an impact factor to all other hydrological processes. The upper reaches of both watersheds are under montane forest, and are considered important as far as rainfall distribution is concerned compared to the mid- and lower-reaches. Rainfall in this area is observed to determine the agricultural gradient as described by the agro- ecological zonation.
Changing land use patterns have been shown to lead to impacts on stream flow processes whereby an increased urban landscape would lead to an increase in stream flow (Grey et al., 2014). In this study, a shift towards urban land use is recorded but it is not considered significant to have an influence in stream flow currently except during storm events.
Continuous spread of the urban landscape is expected to lead to a higher flow in subsequent years, although climate-change induced flooding is seen as a more immediate cause for concern to planners in the region.
An increase in forest cover is desirable as this may reduce the negative impacts of flooding that may be occasioned by increased precipitation. Between 2011 and 2014, several flooding events have been reported in the region, including urban areas adjacent to the region such as Nairobi (Standard Digital, 2015). Poor state of planning of infrastructural facilities has often led to disastrous environmental impacts. When this is coupled with an
151 increasing population and urbanization of urban fringe, the potential impacts of flooding to property and life is significant.
Globally, nutrient contamination is a very important aspect of water quality monitoring in rivers. In the Ruiru and Ndarugu rivers, the impact of nutrient contamination was evident especially where the rivers run near urban areas (Ruiru and Juja). Eutrophication is the most evident sign of nutrient contamination. In this study, high levels of both measured and simulated nutrients in the rivers suggest nutrient contamination. Data from Ruiru-Juja water and Sanitation Company collected at the inlet of its water treatment plants in Ruiru and Juja (sources from Ruiru and Ndarugu Rivers respectively) indicates exceeding levels of coliform bacterial in all instances (over 1100MPN/100ml).
Observed climatic changes are expected to have significant implications on the hydrologic processes, and climate variability is an important factor to be considered for controlling basin hydrologic processes (Qi et al., 2009).
4.3.6 Land use impacts on Simulated Sediment, Phosphate and Nitrate
The SWAT-simulated pattern of land use and land cover on selected water quality parameters is consistent with expectations in the Ruiru watershed but highly dynamic in
Ndarugu watershed (Figure 5-21). Land use patterns have been shown to lead to impacts on stream flow processes (Grey et al., 2014). Typically, an increased urban landscape would lead to an increase in stream flow.
However, in Ndarugu watershed, the SWAT simulation revealed that forest dominated sub- basins have higher sedimentation levels compared to Ruiru watershed. Further investigations in the field revealed that portions of the forest of this watershed have been
152 under cultivation through a KFS programme named Plantation Establishment and
Livelihood Improvement Scheme (PELIS). Cultivation is associated with higher sedimentation as described in the previous section, which may account for the higher sedimentation levels recorded in the forest part of Ndarugu watershed. The PELIS programme has received support for its improvement of local community’s livelihoods, where local communities are allowed to cultivate forest land for food crops as they establish forest plantation seedlings. While they enable faster establishment of forest plantations and at the same time improve their socio-economic situation, a robust erosion mechanism should be adopted by KFS and WRA to ensure that sedimentation in rivers is kept at minimal levels.
Because agriculture in the region is largely rain-fed, increasing amounts of rainfall means that there is higher availability of surface water and soil moisture that is available for agricultural practices. In this regard, the model is considered useful on interactions between precipitation and socio-economic activities, especially agriculture. Specifically, the model generates a range of hydrological parameters that can be incorporated by resource managers to arrive at rational water-related decisions. For example, incorporating nutrient loads generated from land use activities, which are then transported by surface and ground water predicts the high-potential agricultural areas.
153
Ndarugu 30 25 20 15 10 5 0 Agriculture Forest Urban and settlement
SEDPkg_ha ORGNkg_ha ORGPhg_ha
Figure 5-21: Levels of organic phosphates, nitrates and sediment loading in agriculture- dominated, forest-dominated and urban/settlement dominated sub basins.
154
Table 5-7: Simulated vs recommended water quality parameters on Ruiru and Ndarugu Watersheds, Kiambu County Parameter NEMA standard N Ruiru Ndarugu p value
Mean± SD Mean± SD F
SEDPkg_ha 30 (mg/L) 42 17.9±2.35 16.27±1.87 6.92 0.05
ORGNkg_ha 10 (mg/L) 42 11.30±1.28 15.77±3.07 24.71 0.05
ORGPhg_ha 35 (mg/L) 42 75.40±42.82 102.26±42.36 7.98 0.05
Water quality and balance can significantly be influenced by land use and land cover
(Abbaspour et al., 2007). When there in increased human activities, sources of organic nitrogen increase, including sewage from pit latrines and agricultural activities. In these watersheds, an increase in residential areas is observed in the lower reaches, where urban areas are expanding. This is expected to further have an impact on water quality parameters.
Steep slopes (up to 50%) are observed as a major characteristic of both Ruiru and Ndarugu watersheds, particularly in the upper reaches. In many tropical areas, agricultural activities on steep slopes have been recognized as critical in degradation of watersheds. For example, in Jamaica, agriculture on steep slopes is the single most important factor for watershed degradation (Hayman, 2001). The high sedimentation levels in areas under forest could be attributed to steep slopes, forest degradation, and unsustainable agriculture in these areas.
Although modelling results are only indicative, these processes should be further investigated.
155
CHAPTER SIX 6.0 SUMMARY OF FINDINGS, CONCLUSION AND RECOMMENDATIONS 6.1 Research Findings
The following are the key findings that resulted from this study:
Land use and land cover systems are changing as a result of an increasing population. There is a general shift from farmland and grasslands to urban areas and settlement in both Ruiru and Ndarugu watersheds. The Ruiru and Ndarugu watersheds form part of the Nairobi-
Thika rural urban fringe, which is rapidly urbanizing and causing significant impacts on water quality. Immigrating population from other areas presents the local residents with new opportunities, such as establishing of businesses and construction of rental houses to accommodate the immigrant population, thereby increasing the rate of urbanization. b) The extent of surface water resources declined sharply in Ruiru and Ndarugu watersheds from 4.32 km2 and 1.47 km2 in 2005 to 0.73 km2 and 0.57 km2 in 2015. This decline is attributable to an increase in urban and settlement areas which has fueled demand for land for agriculture and settlement as well as increased water demand. As areas under agriculture are gradually being taken up by urbanization, agricultural intensification is observed. An increase in irrigation activities as agriculture intensifies due to decreased land abstracts water from rivers, leading to fewer surface water resources. Increased farming activities and associated irrigation in the dry season also decreases surface water resources.
156
Land use and land cover has an impact on water quality parameters. The land use changes found in the study area, and particularly an increase in urbanization and areas under agricultural, are anticipated for the highland areas of Kenya due to an increasing population. In the study area, this increase is heightened by being part of the rural-urban fringe to the Nairobi city, fuelling demand for land for agriculture and settlement.
Water quality parameters in Ruiru and Ndarugu basins varied across different land use and land cover types, depending on season. Land use and land cover have significant impacts on water quality in the two basins. Forested areas have better water quality than either agricultural dominated and urban dominated landscapes in both wet and dry seasons due to fewer anthropogenic activities in forested lands. The relatively high variation between water quality parameters associated with different land use types reveal that sub-basin basin-level land use characteristics have an impact on surface water quality.
The wet season increases erosive conditions in agricultural and urban/settled areas as rains and flood waters increase leading to an overshoot on the recommended threshold of turbidity levels in surface water. However, the influence of seasons and land use may be compounded by other factors, such as the composition of solubilities of materials in the rock, soil, primary production and inflows that the water flows through.
The generally poor status of macro-invertebrates is attributed to high modification of the landscape, leading to low abundance of aquatic insects. Pollution of the rivers may have contributed to low macro-invertebrate counts. This is attributed to agricultural runoff, sedimentation as well and domestic and industrial effluent.
157
Hydrological modelling revealed increasing flow regime associated with increased precipitation between 2011 and 2014. The higher than normal rainfall occasioned with sudden storm and flooding events and changing land uses may lead to impacts on hydrological processes and should be further investigated.
6.2 Conclusion
Rapid development and population intensity are major drivers of land use change in Ruiru and Ndarugu watersheds, and are projected to lead to major impacts on the rivers’ health, particularly with regard to nitrogen contamination. Although there is a legal and institutional framework for watershed management at the watershed level, no specific river management practices have been recorded. This has been the norm over the past decades.
The findings of this study indicate that both urban and agricultural land use affect water quality in both Ruiru and Ndarugu watersheds by varying degrees.
This study has shown that the Ruiru and Ndarugu watersheds are increasingly becoming urbanized especially at the lower reaches where Thika highway is a major driver in this trend. Increased rainfall patterns are expected to lead to a higher frequency of flooding events and an impact on water quality parameters in both watersheds. Changing climate is expected to lead to a shift in agricultural practices. An increase in urban and residential landscape in both watersheds will have a multiplication effect on hydrology of the two watersheds particularly with regard to suspended sediments and nutrient composition in the stream flow. This study demonstrates that seasonal variations in the region have a strong influence on water quality. Therefore, the Ruiru and Ndarugu rivers rely on the episodes of dry and wet variations to maintain watershed health. As the study demonstrated, suspended sediment transport is greatest during the wet season. The
158 temporal pattern of short duration of heavy flow and long duration of light flow is important. This is because during the wet season, there is removal of material that may have accumulated during the dry season. These dynamics ensure an efficient system of material transport in the channel, but also check against excessive removal of material from the watershed. They also aid in maintaining and regulated rate of nitrification processes
An increasing population is going to lead to an intensification of agricultural activities, and this is strongly favoured by increasing rainfall due to climate change. In addition, urbanization of the urban fringe, which constitutes the lower reaches of both watersheds, will lead to increased runoff and therefore increased flow regime in the rivers. Higher rates of erosion are expected with an increase in rainfall. Without proper management, the watersheds of Ruiru and Ndarugu could potentially be degraded.
To effectively handle the pressures to our natural resources from both land use impacts (a direct result of burgeoning populations) and climate change, a number of innovative methods have been conceptualized, tested and adopted globally. These include integrated watershed management, integrated water management, integrated environmental management and adaptive management. Results of this study reveals that different land use practices in watersheds affect water resources differently, and supports integrated approaches to watershed management because components of watersheds cannot be managed in isolation. In addition, because the land use systems are changing (as revealed by the study), and this change is anticipated to lead to greater pressure on water resources, future water management systems should be sufficiently robust yet flexible to accommodate such changes. The study therefore supports the adoption of adaptive management of water resources, which accommodates changes as they take place. Such
159 approaches should be entrenched in the legal framework and adopted for watershed management in the study area.
6.3 Recommendations
Institutions mandated to manage water resources such as the Water Resources Authority should improve the data collection protocol for long term water quality monitoring in both
Ruiru and Ndarugu rivers. Existing data was seen to have gaps, and weaknesses related to data collection protocol, faulty equipment, timing and recording, which could compromise the overall quality of data. There are only two monitoring stations for the two rivers, which limits our knowledge on specific parts of the watersheds. There is need for establishment of additional monitoring stations and improve the monitoring of water quality at monitoring stations. The importance of having accurate, complete and representative ranges of datasets for landscape is of utmost importance for use in management. This allows for the use of such data when resources are limited to produce reliable information by using models such as SWAT, and analysis of secondary data when sufficiently large databases are available.
Areas under urban and agricultural land use have been shown by the study to be critical in affecting water quality, and therefore efforts for river restoration and management should be focused in these areas. To improve water quality in the landscape, several management techniques can be employed. Several forested areas exist in both catchments. Although most of these areas are under government protection and are not likely to be degraded in the near future, there is need for expansion of forested lands in other parts of the catchments. For example, agro-forestry can be promoted in both large-scale and small- scale farms, while urban parks can be forested to improve the infiltration capacity of the
160 watersheds during the rainy season. In addition, dams and recreational water pools can be established in the landscapes to increase the availability of surface water for different purposes to the growing population in the watershed. These techniques constitute the development of
Blue-Green infrastructure, an innovative approach that addresses catchment-scale water quality management.
In both the urban and agricultural areas, riparian vegetation should be encouraged. Riparian vegetation may play a key role on river water quality by filtering sediment and other pollutants carried in surface runoff. Domestic and industrial wastes are a common source of pollutants in urban areas. The study reveals that urban-dominated sub-watersheds have poor water quality, which is attributed to leaking sewer systems and below-standard septic systems which may all contribute to pollutants in urban rivers. To improve water quality in urban-dominated sub- basins, efficient water treatment plants and sewage conveyance systems should be constructed in urban areas to ensure adequately treated waste water is released back into rivers.
In order to achieve environmental sustainability, an Integrated Watershed Management (IWM) approach should be incorporated, especially because it relates to Ruiru and Ndarugu
Watersheds due to multiple land uses in the landscape. IWM enables an ecosystem-based approach to management, and ensures a holistic management that integrates multiple stakeholders in the watershed (Nobre et al., 2010).
Land use conversion is a major issue and should be done in a sustainable manner. This can be achieved by putting in place effective land use management framework. To avert unsustainable land use conversion, particularly agricultural land, there is every need to manage land use conversions sustainably.
161
REFERENCES
Abbaspour, K. C., Yang, J., Maximov, I., Siber, R., Bogner, K., Mieleitner, J. Zobrist, J.,
and Srinivasan, R. (2007). Modelling Hydrology and Water Quality in the
PreAlpine/Alpine Thur Watershed using SWAT. Journal of Hydrology, 333 (2-4),
413-430.
Abler, D., Shortle, J., Carmichael, J., and Horan, R. (2002). Climate change, agriculture,
and water quality in the Chesapeake Bay region. Clim. Chan, 55, 339–359.
Alexander, R.B., Smith, R.A., Schwarz, G.E., (2000). Effect of stream channel size on the
delivery of nitrogen to the Gulf of Mexico. Nature, 403 (6771), 758-761.
Allaby, A., and Allaby, M. (1999). A Dictionary of Earth Sciences. Oxford University
Press.
Allan J.D., and Johnson L.B. (1997) Catchment-scale analysis of aquatic ecosystems.
Freshwater Biology, 37, 107-111.
Alphan, H. (2003). Land use change and urbanization in Adana, Turkey. Land Degradation
and Development, 14(6), 575–586.
Amoros, C., and Roux, A.L. (1988). Interaction between water bodies within the floodplain
of large rivers: function and development of connectivity. Munstersche
Geographische Arbeiten 29, 125–130.
Arnold, J. G., Williams, J. R., Srinivasan, R., King, K. W. and Griggs, R. H. (1998). SWAT:
Soil Water Assessment Tool. USDA, Agricultural Research Service, Grassland, Soil
and Water Research Laboratory, Temple, TX.
Aron, J.L., Patz, J.A., (2001). Ecosystem Change and Public Health. Johns Hopkins Press,
Baltimore.
162
Baker, T.J., and Miller S.N. (2013). “Using the Soil and Water Assessment Tool (SWAT)
to assess land use impact on water resources in an East African watershed.” J.
Hydrol. Vol. 486, 100–111.
Bassett, R.L. (1997). Chemical modelling on the bare rock or forested watershed scale.
Hydrological Processes 11: 695-717.
Berrang, F.L. Malaria. In: Geist H.J. (Ed). (2006). Our Earth's changing land: An
encyclopedia of land-use and land-cover change. Westport London: Greenwood
Press.
Bicknell, B. R., J. C. Imhoff, J. L. Kittle, Jr., A. S. Donigian, Jr., and R. C. Johanson.
(1993). Hydrologic Simulation Program - FORTRAN (HSPF): User’s Manual for
Release 10. Report No. EPA/600/R-93/174. Athens, Ga.: U.S. EPA Environmental
Research Lab.
Biswas, S., Sudhakar, S., and Desai, V. R. (2002). Remote Sensing and Geographic
Information System Based Approach for Watershed Conservation. Journal of
Surveying Engineering, 128(3): 108-124.
Black, E., Slingo, J. and Sperber, K.R. (2003). An observational study of the relationship
between excessively strong short rains in coastal East Africa and Indian Ocean SST.
Mon. Weather Rev. 131, 74e94.
Bolstad, P. V. and Swank, W. T. (1997). Cumulative impacts of land use on water quality
in a southern Appalachian watershed. Journal of the American Water Resources
Association 33(3): 519-534.
163
Borah, D. K., and Bera, M. (2004). Watershed-scale hydrologic and nonpoint-source
pollution models: review of applications. Transactions of the American Society of
Agricultural Engineers (ASAE). 47(3): 789-803.
Brodie, J., and Mitchell, A. (2005). Nutrients in Australian Tropical Rivers: Changes With
Agricultural Development and Implications for Receiving Environments. Marine
and Freshwater Research, 56 (3):279-302.
Bunn, S.E. (1995) Riparian-stream linkages: research needs for the protection of in-stream
values. Australian Biologist. 6: 46-51.
Camberlin, P., Janicot, S., and Poccard, I. (2001). Seasonality and atmospheric dynamics
of the teleconnection between African rainfall and tropical sea-surface temperature:
Atlantic vs. ENSO. Int. J. Climatol. 21, 973-1005.
Cao, C., Matsumura, K., Ozaki, M., and Watanabe, Y., (1999) Lipopolysaccharide injected
into the cerebral ventricle evokes fever through induction of cyclooxygenase-2 in
brain endothelial cells. J. Neurosci. 19:716-725
Changnon, S.A., and Demissie, M., (1996) Detection of changes in stream flow and floods
resulting from climate fluuctuations and land use-drainage changes. Climatic
Change. 32: 411-421.
Chen, H., Jacobs, E., Schwarzschild, M.A., McCullough, M.L., Calle, E.E., Thun, M.J.,
Ascherio, A., Nonsteroidal anti-inflammatory drug use and the risk for Parkinson's
disease. Ann Neurol 2005a; 58: 963-967
Chen, J., Lu, J., (2014) Effects of Land Use, Topography and socio-economic factors on
river water quality in a mountainous watershed with intensive agricultural
production in east China. PLoS ONE, 9, doi:10.1371/journal.pone.0102714.
164
Clarke, K. C. (2003). Getting started with geographic information systems. New Jersey:
Prentice Hall.
Codjoe, S. N. A., (2007). Integrating Remote Sensing, GIS, Census, and Socioeconomic
Data in Studying the Population–Land Use/Cover Nexus in Ghana: A Literature
Update." Africa Development. 32 (2): 197–212.
Cogalton, R.G., (1991). A Review of Assessing the Accuracy of classification of remotely
sensed data. Remote Sensing and the Environment. 37 (1): 35-46.
Collins, A. E., (2001). Health ecology, land degradation and development. Land
DegradDev. 12: 237–50.
Collins, C.L., Mullen, M.W., Stewart, P.M., and Webber, E.C. (2008). Validation of an
invertebrate community index for urban streams for an Alabama coastal plains
watershed. J. Am. Wat. Resour. Assoc. 44(3):663-669.
Convention on Biological Diversity. (2006). Decision adopted by the conference of the 451
parties to the convention on biological diversity at its eighth meeting; VIII/9. 452
Implications of the findings of the millennium ecosystem assessment. CBD
Cornelissen, T., Diekkruger, B. and Giertz, S. (2013). A comparison of hydrological
models for assessing the impact of land use and climate change on discharge in a
tropical catchment. J. Hydrol. 498: 221–236
Cosgrove, D, Petts, G, eds (1990). Water, Engineering and Landscape. London: Belhaven
Press; 1990. 2.
Cunderlik, J. M. and Burn, D. H. (2002). The use of flood regime information in regional
flood frequency analysis. Hydrol. Sci. J. 47(1), 77–92.
165
Cunderlik, J. M. (2002). Analysis of the linkage between rain and flood regime and its
application to regional flood frequency estimation. Journal of Hydrology, 261 (1-
4): 115-131.
Daily, D.C. (1997). Nature’s Services. Island Press, Washington, DC.
Darnerud, P. O. (2003). Toxic effects of brominated flame retardants in man and in
wildlife. Environ Int29:841–53.
Davis, M. (2002). Late Victorian Holocausts: El Niño Famines and the Making of the Third
World. Verso. pp. 263–266.
DeFries, R. and Eshleman, K. N.(2004): Land-use change and hydrologic processes: a
major focus for the future, Hydrol. Process.,18: 2183–2186, 2004.
Deng, H., Shu, S.G.,Song, Y.Q. and Xing, F.L. (2007). Distribution of sand dunes and sand
shifts along the southern fringe of the Mu Us Desert since the Ming Dynasty. Chin.
Sci. Bull. 52: 3128–3138.
Dewar, R.E. and Wallis, J. R. (1999). Geographical patterning in interannual rainfall
variability in the tropics and near tropics: An L-moments approach. Journal of
Climate. 12: 3457–3466.
Dile, Y.T. and Raghavan, S. (2014). Evaluation of CFSR climate data for hydrologic
prediction in data-scarce watersheds: an application in the Blue Nile River Basin.
Journal of the American Water Resources Association (JAWRA). 50(5): 1226-1241.
DOI: 10.1111/jawr.12182
Di Luzio, M., Srinivasan, R., and Arnold, J.G., (2002). Integration of watershed tools and
SWAT model into BASINS. Journal of American Water Resource Association 4,
1127–1141.
166
Donnay, J.-P., M.J. Barnsley, and P.A. Longley (2001). Remote sensing and urban
analysis. In: Remote Sensing and Urban Analysis, Taylor and Francis, London,
UK, pp. 3-18.
Dudgeon, D. 1992. Endangered ecosystems: a review of the conservation status of tropical
Asian rivers. Hydrobiologia 248:167–191.
Edsel, B. D., Janey, V. C., Eugene, J. L., and Jessica, R. P. (2010). Watershed Modeling
Using GIS Technology: A Critical Review. Journal of Spatial Hydrology. 10(2):
13-28
Elwood, J. W., Newbold, J. D., O’Neil, R. V. and Van Winkle, W. (1983). Resource
spiraling: an operational paradigm for analyzing lotic ecosystems. In Dynamics of
Lotic Ecosystems, eds T. D. Fontaine and S. M. Bartell, pp. 3–27. Ann Arbor
Science, Ann Arbor, MI.
Faure, G., (1998). Principles and Applications of Geochemistry. Prentice Hall: Upper
Saddle River, NJ; 625.
Fawzi, B., Loudike, M., Oubraim, S., Sabour, B., and Chlaida, M. (2002). Impact of
wastewater effluent on the diatom assemblages’ structure of a brackish small
stream: OuedHassar (Morocco). Limnologica 32, 54- 65.
Ferreyra, C., and Beard, P. (2007). Participatory Evaluation of Collaborative and Integrated
Water Management: Insights from the Field. Journal of Environmental Planning
and Management. 50(2): 271-296.
Fisher, D. S., Steiner, J. L., Wilkinson, S. R. (2000). The relationship of land use practices
to surface water quality in the Upper Oconee Watershed of Georgia. Forest Ecology
and Management. 128: 39-48.
167
Fohrer, N., Haverkamp, S., Eckhardt, K., and Frede, H., (2001). Hydrologic response to
land use changes on the catchment scale. Physics and Chemistry of the Earth, Part
B 26 (7–8), 577–582
Freeman, G.K., Woloshynowych, M., Baker, R., Boulton, M., Guthrie, B., Haggerty, J.,
Car, J., and Tarrant, C. (2007). Continuity of Care 2006: What have we learned
since 2000 and what are policy imperatives now? London: National Institute for
Health Research.
Frissell, C.A., Liss, W.J., Warren, C.E., and Hurley, M.D. (1986). A hierarchical
framework for stream habitat classification: viewing streams in a watershed
context. Environmental Management.10: 199-214.
Frohlich, H.L., Breuer, L., Frede, H., Huisman, J.A., and Vache, K.B. (2008). Water source
characterization through spatiotemporal patterns of major, minor and trace element
stream concentrations in a complex, Mesoscale German catchment. Hydrol.
Process. 22: 2028–2043
Garnier, B.J. (1958). Some Comments on defining the humid tropics. Research Note 11.
Department of Geography, University of Ibadan.
Gburek, W. J., and Folmar, G. J. (1999). Flow and chemical contributions to stream¯ow in
an upland watershed: a base flow survey. Journal of Hydrology 217: 1-18.
Giertz, S., and Diekkruger, B. (2003). Analysis of the hydrological process in a small
headwater catchment in Benin (West Africa). Physics and Chemistry of the
Earth.28, 1333-1341.
Gikunju, J.K., Simiyu, K.W., Gathura P.B., Kyule, M., and Kanja, L.W. (2002). River
water fluoride in Kenya. Flouride. 35 (3):193–196.
168
Githui, F., Mutua, F., and Bauwens, W. (2001). Estimating the impacts of land-cover
change on runoff using the soil and water assessment tool (SWAT): case study of
Nzoia catchment, Kenya. Hydrol. Sci. J., vol 54, no 5, pp. 899-908.
Gleick, P. H. (1998). The world’s water 1998–1999: the biennial report on freshwater
resources. Island Press, Washington, D.C., USA.
Global Water Partnership. (2005). Intergrated Water Resources Management in Eastern
Africa. Technical Focus Report. Global Water Partnership, Sweden.
Gray, C.L., Lewis, O.T., (2014). Do riparian forest fragments provide ecosystem services
or disservices in surrounding oil palm plantations? Basic Appl. Ecol. 15, 693–700
Gregory, K. J. & Walling, D. E. (1973). Drainage Basin Form and Processes: A
Geomorphologic Approach. Norwich: Fletcher and Son Ltd, pp 22-58.
Gregory, K.J. (2006).The human role in changing river channels. Geomorphology, 79:172–
191.
Gregory, S.V., Swanson, F.J., Mckee, W.A. and Cummins, K.W. (1991). An ecosystem
perspective of riparian zones. Bioscience, 41: 540±551.
Grimm, N. B., Grove, J. M., Pickett, S. T. A., and Redman, C. L. (2000). Integrated
approach to long-term studies of urban ecological systems. Bioscience. 50(7): 571–
584
Gurtz, M. E., and J. B. Wallace. (1984). Substrate-mediated response of stream
invertebrates to disturbance. Ecology 65: 1556-1569
Hall, P., and Pfeiffer, U. (2000) Urban Future 21: A Global Agenda for Twenty-First
Century Cities. London: Spon.
169
Hayman, A. (2001). Rapid Rural Appraisal of the Great River Watershed: Ridge to Reef
Watershed Project. Burlington, VT: National Environment and Planning Agency
and the United States Agency for International Development.
Herold, M., Couclelis, H., and Clarke, K. C. (2005). The role of spatial metrics in the
analysis and modeling of urban land use change. Computers, Environ. & Urban
Systems, 29(2005), 369-399.
Hoekstra, A. Y. and Wiedmann, T. O. (2014) Humanity’s unsustainable environmental
footprint. Science 344: 1114–1117.
Hoekstra, A. Y., and Wiedmann, T. O. (2014) Humanity’s unsustainable environmental
footprint. Science 344: 1114–1117.
Holdgate, M. W. (1993). The sustainable use of tropical coastal resources – a key
conservation issue. Ambio, (22): 481-482.
Hooper, B. P. (2003). Integrated Water Resources Management and River Basin
Governance. Water Resources Update. 126: 8.
Horton, R.E. (1945). Erosional development of streams and their drainage basins.
Hydrophysical approach to quantitative morphology. Bull. Geo. Soc. pp 275-350
Hunsaker C.T., and Levine D.A. (1995). Hierarchical approaches to the study of water
quality in rivers. Bioscience. 45: 193-203.
Im, S., Kim, H., Kim, C., and Jang, C., (2009). Assessing the impacts of land use changes
on watershed hydrology using MIKE SHE. Environmental Geology 57, 231–239.
IUCN (International Union for the Conservation of Nature). (2000). Vision for water and
nature: a world strategy for conservation and sustainable management of water
170
resources in the 21st century. International Union for the Conservation of Nature,
Gland, Switzerland and Cambridge, UK.
Jackson, A.R., and Jackson, J.M. (1998). Environmental Science: the natural environment
and human impact. Longman, London. 370pp.
Jackson, I.J. (1989). Climate, Water and Agriculture in the Tropics. Longman, Harlow.
Jackson, T. (1988) The Law of Kenya. Kenya Literature Bureau. Nairobi, Kenya.
Jacobs, J.H., Angererb, J., Vitalec, J., Srinivasana, R., and Kaitho, R.,(2007). “Mitigating
Economic Damage in Kenya’s Upper Tana River Basin: An Application of Arc-
View SWAT.” Journal of Spatial Hydrology. Vol.7. No.1. Spring
Jaetzold, R., and Schmidt, H. (1983). Farm management handbook of Kenya, Vol. IIC
Eastern Province. Natural conditions and farm management information. Ministry
of Agriculture, Nairobi, Kenya.
Jain, M. K., Kothyari, U. C., and Ranga-Raju, K. (2004). A GIS based distributed rainfall-
runoff model. Journal of Hydrology, 299(1-2), 107-135.
Jakeman, A. J., and Letcher, R. A. (2003). Integrated Assessment and Modelling: Features,
Principles and Examples for Catchment Management. Environmental Modelling
and Software. 18 (6): 491-501.
Jayakrishnan, R., Srinivasan, R., Santhi, C., and Arnold, J., (2005). “Advances in the
application of the SWAT model for water resources management.” Hydrol.
Process., vol19, pp. 749–762.
Jensen, J. R., Cowen, D. J., Althausen, J. D., Narumalani, S., and Weatherbee, O. (1993).
An evaluation of the Coastwatch change detection protocol in South Carolina.
Photogrammetric Engineering and Remote Sensing, 59, 1039–1046.
171
Johnson, B. L., Richardson W. B., and Naimo, T. J. (1995). Past, present, and future
concepts in large river ecology. Bioscience, 45(3):134–141.
Johnson, L.B., Richards, C., Host, G., and Arthur, J.W. (1997) Landscape influences on
water chemistry in Midwest stream ecosystems. Freshwater Biology. 37: 193-208.
Kalnay et al. (1996). The NCEP/NCAR 40-year reanalysis project, Bull. Amer. Meteor.
Soc., 77, 437-470.
Kang, J., Lee, S.W., Cho, K.H., Ki, S.J., Cha, S.M., and Kim, J.H. (2010). Linking land-
use type and streamwater quality using spatial data of fecal indicator bacteria and
heavy metals in the Yeongsan riverbasin. Water Res., 44, 4143–4157.
Karr, J.R. (1981). Assessment of Biotic Integrity Using Fish Communities. Fisheries 6:
21-27
Kenya National Bureau of Statistics-KNBS. (2016). Urban Center Populations in Kiambu
County. KNBS, 2016.
Kibichii, S., Shivoga, W.A., Muchiri, M., and Miller, S.N. (2007). Macro invertebrate
assemblages along a land-use gradient in the upper River Njoro watershed of Lake
Nakuru drainage basin, Kenya. Lakes and Reservoirs: Research and Management.
Kiio, M.M., and Achola, O.P. (2015). Land Use Land Cover Changes and their Effects on
Agricultural Land: A Case Study of Kiambu County, Kenya. Kabarak Journal of
Research and Innovation. 3, 74-86.
King, R. T. (1966). Wildlife and man. NY Conservationist 20(6), 8–11
Kirkby, M. J. and Weyman, D. R. (1974). Measurements of contributing area in very small
drainage basins. Seminar Series B, No. 3. Department of Geography, University of
Bristol, Bristol.
172
Kithiia, S.M. (1992). Effects of industries and other land-use systems on the water quality
within the Nairobi River sub-catchments, Kenya. MSc Thesis. University of
Nairobi.
Kithiia, S.M. (1997). Land use changes and their effects on sediment transport and soil
erosion within the Athi drainage basin, Kenya. In: Human Impact on Erosion and
Sedimentation (ed., D.E. Walling and J.L, Probst). Proceedings of Rabat
Symposium S6, April 1997.
Köppen, Wladimir. (1936). The geographic system of climates. Handbuch der
Klimatologie. 1. Berlin: Borntraeger.
Körner, C., and Ohsawa, M. (2005). Mountain system. Ecosystems and Human Well Being,
Current states and trends. Millennium Ecosystem Assessment WHO report.
Kottek, J., Grieser, J. and Beck, C. (2006). World Map of the Köppen-Geigger Climate
Classification updated. Meteorologische Zeitchrift.15: 259-263
Lambin E.F., and Giest H.J., (Eds). (2006). Land Use and Land Cover Change: Local
Processes, Global Impacts. Springer.
Le, A.M., and Pricope, N.G. (2017) “Increasing the Accuracy of Runoff and Streamflow
Simulation in the Nzoia Basin, Western Kenya, through the Incorporation of
Satellite-Derived CHIRPS Data.” Water. Vol 9. No 114.
Lebel, J. Health. (2003). An Ecosystem Approach. International Development Research
Centre. Ottawa: Canada.
LeBlanc, R. T., Brown, R. D. and FitzGibbon, J. E. (1997). Modeling the effects of land
use change on the water temperature in unregulated urban streams. Journal of
Environmental Management.49: 445-469.
173
Lenat, D.R., and Crawford, J.K., (1994). Effects of landuse on water-quality and aquatic
biota of three North Carolina Piedmont streams. Hydrobiologia 294:185–99
Leopold, L.B., Wolman, M.G., and Miller, P. (1964). Fluvial Processes in
Geomorphology. W.H. Freeman and Company, 522 p.
Li, Y., Sperry, J.S., and Shao, M. (2009) Hydraulic conductance and vulnerability to
cavitation in corn (Zea mays L) hybrids of differing drought resistance. Environ
Exp Bot 66:341–346
Liu, J. G., Zehnder A. J. B., Williams J. R., and Yang H. (2007b), Development and testing
of the GEPIC model for global virtual water content estimation, Agric.
Syst., 94, 478–493.Pringle, 2001
Liu, X., and Lathrop, R. G. (2002). Urban change detection based on an artificial neural
network. International Journal of Remote Sensing, 23, 2513–2518.
Liverman, D., Emilio F. M., Ronald R. R., and Paul C. S. (1998). People and pixels:
linking remote sensing and social science. Washington, D.C., National Academy
Press.
Lopez, E., Bocco, G., Mendoza, M., and Duhau, E. (2001). Predicting land cover and land
use change in the urban fringe a case in Morelia City, Mexico. Landscape and
Urban Planning, 55(4): 271–285.
Lu, X., Zhao, M., Kong, H., Ca, J., Wu, J., Wu, M., Hua, R., Liu, J., and Xu G. (2004),
Characterization of complex hydrocarbons in cigarette smoke condensate by gas
chromatography–mass spectrometry and comprehensive two-dimensional gas
chromatography–time-of-flight mass spectrometry, Journal of Chromatography A,
vol. 1043(2), pp. 265-273.
174
Lubchenco, J. (1998). Entering the century of the environment: a new social contract for
science. Science 279: 491–495.
Macan, T.T. (1977). A guide to freshwater invertebrate animals. Longman publishers,
London.
Magilligan, F. J. and Stamp, M. L. (1997). Historical Land-Cover Changes and
Hydrogeomorphic adjustment in a Small Georgia Watershed. Ann. Assoc. Am.
Geogr. 87: 614–635.
Magurran, A.E. (1988). Ecological Diversity and its Measurement. Princeton University
Press, Princeton, NJ. pp 179
Mander, U., Kull, A., Tamm, V., Kuusemets, V. and Karjus, R. (1998). Impact of climatic
fluctuations and land use change on runoff and nutrient losses in rural landscape.
Landscape and Urban Planning41: 229-238.
Mango, L.M., Melesse, A.M., McClain, M.E., Gann, D., and Setegn, S.G. (2013) “ydro-
Meteorology and Water Budget of the Mara River basin Under Land Use Change
Scenarios. In Nile River Basin: Hydrology, Climate and Water Use, Melesse AM
(ed). Springer: Dordrecht; New York; 39–68.
Mango, L.M., Melesse, McClain, M.E., Gann, D. and Setegn, S.G., (2002). Land use and
climate change impacts on the hydrology of the upper Mara River Basin, Kenya:
results of a modeling study to support better resource management. Hydrol. Earth
Syst. Sci. vol. 15, pp. 2245–2258.
Margerum, R. D. (1999). Integrated Environmental Management: The Foundations for
Successful Practice. Environmental Management, 24(2): 151-166.
175
Mark, D. Abkowitz, and James, P. Dobbins. (2016). How Maximum Likelihood Works.
ESRI Developer Network.
Marshall, B.E. (1972). Some effects of organic pollution on benthic fauna. Rhodesia
Science News 6, 142-145.
Masese, F.O., Muchiri, M., and Raburu, P.O. (2009a). Macroinvertebrate assemblages as
biological indicators of water quality in the Moiben River, Kenya. Afr. Journ.
Aquat. Sc.
Masese, F.O., Raburu, P.O., and Muchiri, M. (2009b). A preliminary benthic
macroinvertebrate index of biotic integrity (B-IBI) for monitoring the Moiben
River, Lake Victoria Basin, Kenya. Afr. J. Aquat. Sc.
McGregor, G. R., and Nieuwolt, S. (1998). Tropical climatology. An introduction to the
climates of the low latitudes. Wiley, Paris.
Meadows, D. H., and Club of Rome. (1972). The Limits to growth: A report for the Club
of Rome's project on the predicament of mankind. New York: Universe Books.
Meadows, D. H., Meadows, D. L., Randers, J., et al. (1972). Limits to Growth. Universe
Books, New York
MediaMax. (2015). Flash Floods Wreak Havoc In Ruiru, Kiambu County.
http://www.mediamaxnetwork.co.ke/k24-tv/146997/flash-floods-wreak-havoc-in-
ruiru-kiambu-county/
MediaMax. (2015). Property worth millions destroyed as floods wreak havoc in Ruiru,
Kiambu County. http://www.mediamaxnetwork.co.ke/k24-tv/186227/property-
worth-millions-destroyed-as-floods-wreak-havoc-in-ruiru-kiambu-county/
176
Meigs, P. (1958). World’s distribution of Arid and Semi- arid homoclimates. In: Arid Zone
hydrology. UNESCO Arid Zone Research Series. 1: 203-209
Merritt, R.W., and Cummins K.W. (1996). An introduction to the Aquatic Insects of North
America.Kendall/Hunt Publishing Company, Dubuque, IA.
Meybeck M. (2002). Riverine quality at the Anthropocene: Propositions for global space
and time analysis, illustrated by the Seine River. Aquatic Sciences 64: 376–393.
Middleton, N.J., Thomas, D.S.J., and UNEP (Eds). (1997). World Atlas of Desertification.
UNEP
Millennium Ecosystem Assessment (2003). Ecosystems and Human Wellbeing: A
Framework for Assessment. Island Press, Washington, DC
Millennium Ecosystem Assessment (2005). Ecosystems and Human Wellbeing: General
Synthesis. Island Press, Washington, DC.
Minshall, G.W., Cummins, K.W., Petersen, R.C., Cushing, C.E., Bruns, D.A., Sedell, J.R.,
and Vannote, R.L. (1985). Developments in stream ecosystem theory. Canadian
Journal of Fisheries and Aquatic Sciences. 42: 1045-1055.
Moldan, B., and Cerny, K., (1994) Biogeochemistry of Small Catchments: A Tool For
Environmental Research. Wiley, Chichester, England.
Moyo, N.A.G. (ed) (1997). Lake Chivero: a polluted lake. University of Zimbabwe
Publications.
Musau, J., Sang, J., Gathenya, J., Luedeling, E., and Home, P.,(2014). SWAT model
parameter calibration and uncertainty analysis using the HydroPSO R package in
Nzoia Basin, Kenya. Journal of Sustainable Research in Engineering. vol 1. No 3.
Pp. 17-29.
177
Musau, J., Sang, J., Gathenya, J., and Luedeling, E., (2015) . Hydrological responses to
climate change in Mt. Elgon watersheds. J. Hydrol. vol. 3, pp. 233–246.
Nash, J.E., and Sutcliffe, J.V., (1970). River flow forecasting through conceptual models.
Part I – A discussion of principles. J. Hydrol. Vol 10. No 3. Pp. 282–290.
Ndaruga, A.M., George, G.N., Nathan, N.G., Wamicha, W.N., (2004). Impact of water
qualityu on macroinvertebrate assemblages along a tropical stream in Kenya. Afr.
J. of Ecol. 42: 208-216.
Neitsch, S.L., Arnold, J.G., Kiniry, J.R., and Williams, J.R. (2005). Soil and Water
Assessment Tool, Theoretical Documentation: Version 2005. Temple, TX. USDA
Agricultural Research Service and Texas A&M Blackland Research Center.
Ngana, J. O., Mwalyosi, R. B. B., Yanda, P., and Madulu, N. F. (2004). Strategic
development plan for integrated water resources management in Lake Manyara
Sub-basin, North Eastern Tanzania. Physics and Chemistry of the Earth 29: 1219-
1224.
Nilson, A. (ed). (1996). Aquatic Insects of North Europe. A taxonomic Handbook Vol 1.
Ephemeroptera-Plecoptera-HeteroteraNeuroptera-Megaloptera-Coleoptera-
Trichoptera-Lepidoptera. Apollo Books Aps.
Nobre, A. M., Ferreira, J. G., Nunes, J. P, Yan, X., Bricker, S., Corner, R., Groom, S., Gu,
H., and Zhu, M. 2010. Assessment of Coastal Management Options by Means of
Multilayered Ecosystem Models. Estuldarine, Coastal and Shelf Science.87(1):43-
62.
178
Odira, P.M.A., Nyadawa, M.A., Ndwallah, B.A., Juma, N.A., and Obiero, J.P.,(2010).
Impact of land use/cover dynamics on streamflow: A case of Nzoia river catchment,
Kenya. Nile Basin Water Sci. Eng. J. vol 3, pp. 64–78.
Ojea, E., Martin-Ortega, J., and Chiabai, A. (2012). Defining and classifying ecosystem
services for economic valuation: the case of forest water services. Environmental
Science & Policy, vol 19, pp 1–15
Omwoyo, M.A., Muthama, N.J., Opere, A., and Onwonga, R.,(2017) “Simulating
Streamflow in Response to Climate Change in the Upper EwasoNgiro Catchment,
Kenya.” Journal of Climate Change and Sustainability, vol 1, no. 2, pp. 12-29,
Oogathoo, S. (2006). Runoff Simulation in the Canagagigue Creek Watershed using the
MIKE SHE Model. Department of Bioresource Engineering, M.Sc. Thesis. McGill
University, Montreal, Quebec.
Osborne, L. L., and Wiley M. J. (1988). Empirical relationships between land use/cover
and stream water quality in an agricultural watershed. Journal of Environmental
Management. 26: 9–27.
Poff, N. L., J. D. Allen, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter, R. E.
Sparks, and J. C. Stromberg. 1997. The natural flow regime: a paradigm for river
conservation and restoration. BioScience. 47:769–784.
Pringle C.M. 2001. Hydrologic connectivity and the management of biological reserves: a
global perspective. Ecological Applications 11: 981–998.
Pringle, C.M., and Triska, F.J. (2000). Emergent biological patterns and surface–
subsurface interactions at landscape scales. In Stream and Groundwaters, Jones
JB, Mulholland PJ (eds). Academic Press: New York, 167–193.
179
Qi, S., Sun, G., Wang, Y., McNulty, S. G., and Myers Moore, J. A. (2009). Streamflow
Response to Climate and Landuse Changes in a Coastal Watershed in North
Carolina. American Society of Agricultural Engineers,52(3): 11.
Quigley, M. (1977). Invertebrates of streams and Rivers: a key to identification. Edward
Arnold Ltd. London.
Quinn, M.J., Cooper, B., Robert, J., Colley, D.J., Christopher, R., and Bruce, W. (1997).
Land use effects on habitat, water quality, periphyton, and benthic invertebrates in
Waikato, New Zealand, hill‐country streams. New Zealand Journal of Marine and
Freshwater Research Vol. 31 , Iss. 5,199
Raburu, P.O., Masese, F.O., and Mulanda, C.A. (2009). Macroinvertebrate Index of Biotic
Integrity (M-IBI) for monitoring rivers in the upper catchment of Lake Victoria
Basin, Kenya. Aquat Ecos. Health Manage. 12 (2): 1- 9
Reading, A.J., Thompson, R.D. and Millington, A.C. (1995). Humid Tropical
Environments. Blackwell, Oxford.
Rees, W.E., and Wackernagel, M. (1994). Ecological Footprints and appropriated
carrying capacity: measuring the natural capital requirements of the human
economy. In: Investing in Natural Capital: The Ecological Economics Approach to
Sustainability. Island Press, Washington.
Reid, M.A., Tibby, J.C., Penny, D., and Gell, P.A. (1995). The use of diatoms to assess
past and present water quality. Australian Journal of Ecology. 20: 57-64.
Republic of South Africa, Department of Water and Sanitation. (2016). Mini Stream
Assessment Scoring System (miniSASS). http://www.minisass.org/en/links/
180
Richards, C., Haro, R.J., Johnson, L.B. and Host, G.E. (1997). Catchment and reach-scale
properties as indicators of macroinvertebrate species traits. Freshwater Biology.37:
219-230.
Richards, C., Johnson, L.B., and Host, G.E. (1996). Landscape-scale influences on stream
habitats and biota. Canadian Journal of Fisheries and Aquatic Sciences. 53: 295-
311.
Richter, B. D., Baumgartner J. V., Wigington R., and Braun D. P. (1997b). How much
water does a river need? Freshwater Biology.37:231–249.
Richter, B. Mathews, R., Harrison, D., and Wigington, R. (2003). Ecologically Sustainable
Water Management: Managing River Flows for Ecological Integrity. Ecological
Applications, 13(3) 206-224.
Russell, I.A. (1999). Changes in the water quality of the wilderness and Swatvlei lake
systems, S. Africa. Koedoe 42: 57-72.
Saggerson, E.P. (1967). Geological map of Nairobi Area. Geological Survey of Kenya
Savage, J.M, Goldman, D.P. and Janos, D.P. (1982). Ecological Aspects of Development
of Tropical Climates. National Academy Press, Washington DC.
Schneider, S., Schranz, C., and Melzer, A. (2000). Indicating the trophic state of running
waters by submersed macrophytes and Epilitic Diatoms: Exemplary
implementation of a new classification of taxa into trophic classes. Limnologica,
30: 1-8.
Schofield, N.J., and Davies, P.E. (1996). Measuring the health of our rivers. Water
(May/June) 39-43.
181
Scott, A., Carter, C., Holzinger, O., et al. (2014). Tools, applications, benefits and linkages
for ecosystem science (TABLES). Report for the National Ecosystem Assessment
follow-on.
Sepulchre, P., Ramstein, G., Fluteau, F., Schuster, M., Tiercelin, J. and Brunet, M., (2006).
Tectonic uplift and Eastern Africa aridification. Science. 313, 1419.
Shand, P., Haria, A.H., Nealm C., Griffiths, K.J., Gooddy, D.C., Dixon, A.J., Hill, T.,
Buckley, D.K., Cunningham, J.E. (2005). Hydrochemical heterogeneity in an
upland catchment: further characterisation of the spatial, temporal and depth
variations in soils, streams and groundwaters of the Plynlimon forested catchment,
Wales. Hydrology and Earth System Sciences. 9: 621–644.
Shitakha, F.M. (1983). Detailed Soil Survey of the Coffee Research Station, Ruiru, Kiambu
Singh, K.P., Malik, A., and Sinha, S. (2005). Water quality assessment and apportionment
of pollution sources of Gomti river (India) using multivariate statistical technique
sea case study. Analytica Chimica Acta. 538: 355-374.
Sliva, L., and Williams, D.D. (2002). Buffer Zone versus Whole Catchment Approaches
To Studying Land Use Impact On River Water Quality. Wat. Res. 35 (14): 3462–
3472
Sombroek, W.G., Brown, H.M.H., and Van Der Pouw, B.J.A. (1980) Exploratory Soil Map
and Agro-climatic Zone Map of Kenya. Exploratory Soil Survey Report No. E,
Including Map. Nairobi: Kenya Soil Survey (KSS).
Sparks, R. E. (1995). Need for ecosystem management of large rivers and their floodplains.
BioScience. 45:169–182.
182
Sponseller, R.A., and Benreld E.F. (2001) In¯uences on land use on leaf breakdown in
southern Appalachian headwater streams: a multiple scale analysis. Journal of the
North American Benthological Society. 20: 44- 59.
Standard Digital. (2015). Ten killed as a 14-seater matatu is swept by floods in Gatundu,
Kiambu County. https://www.standardmedia.co.ke/article/2000183947/ten-killed-
as-a-14-seater-matatu-is-swept-by-floods-in-gatundu-kiambu-county
Stanford, J.A., Ward, J.V., Liss, W.J., Frisell, C.A., Williams, R.N., Lichatowitch, J.A.,
and Coutant, C.C. (1996). A general protocolfor restoration of regulated rivers.
REgul Rivers Res MNgt. 12: 391-413
Statzner, B., and Higler., B. (1985). Questions and comments on the River Continuum
Concept. Can. J. Fish. Aquat. Sci. 42: 1038-1044
Sthiannopkao, S., Takizawa, S., Homewong, J., and Wirojanagud, W. (2007). Soil erosion
and its impacts on water treatment in the northeastern provinces of Thailand.
Environ Int. 33 (5):706–11.
Stonestrom, D. A., Scanlon, B. R., and Zhang, L. (2009): Introduction to special section
on Impacts of Land Use Change on Water Resources, Water Resour. Res.
Stout I., and Ben, M. (2003). River Continuum Concept as an Analytical Template for
Assessing Watershed Health. Wheeling Jesuit University.
Stow, D. A., and Chen, D. M., (2002), Sensitivity of multitemporal NOAA AVHRR data
of an urbanizing region to land-use/land-cover change and misregistration. Remote
Sensing of Environment. 80, 297–307.
183
Strahler, A.N. (1964). Quantitative Geomorphology of Drainage Basins and Channel
Networks, section 4-11. In Handbook of Applied Hydrology, ed, by V.T. Chow.
McGraw-Hill, New York.
Stumm, W., and Morgan, J.J. (1996). Aquatic Chemistry: Chemical Equilibria and Rates
in Natural Waters. John Wiley and Sons: New York
Survey of Kenya (2016). Basemaps for Kenya in “Shapefile” Format. Survey of Kenya,
Nairobi.
Swank, W.T., Swift Jr., L.W., Douglass, J.E., (1988). Streamflow changes associated with
forest cutting, species conversions, and natural disturbances. In: Swank, W.T.,
Crossley Jr., D.A. (Eds.), Forest Hydrology and Ecology at Coweeta. Ecological
Studies, Vol. 66. Springer, New York, pp. 297-312
Tacoli, C. (2002). Changing Rural-Urban Interactions in Sub-Saharan Africa and their
Impacts on Livelihoods: A Summary. London: International Institute for
Environment and Development.
Tafangenyasha, C., and Dzinomwa, T. (2005). Land-use impacts on river water quality in
lowveld sand river systems in south-east Zimbabwe. Land Use. Water. Resou. Res.,
5: 3.1-3.10.
Taylor, L.H., Latham, S.M., and Woolhouse, M.E.J. (2001). Risk factors for human
disease emergence. Philos Trans R SocLond. 356:983–9.
Thornthwaite, C.W. (1948). An Approach Towards A Rational Classification of Climate.
Geographical Review. 38: 55-94
Thornton, K.W., Kennedy, R.H., Magoun, A.D. and Saul, G.E. (1982). Reservoir water
quality sampling design. Water Resources Bull. 18, 471-480.
184
Thuo, A. D. M. (2013). Impacts of Urbanization on Land Use Planning, Livelihood and
Environment in the Nairobi Rural-Urban Fringe, Kenya. International Journal of
Scientific and Technology Research, 2(7), 70–79. Retrieved from www.ijstr.org
Tong, H. (1990). Non-linear time series: a dynamical system approach. Oxford: Oxford
University Press.
Tong, T.Y., and Wenli, C. (2002). Modeling the relationship between land use and surface
water quality. Journal of Environmental Management.66, 377-393.
Torrey, B. B. (1998). We need More Research on the Impact of Rapid Urban Growth.
Chron. Higher Educ. 45: B6.
Turner, B. L., II (1994). Local faces, global flows: the role of land use and land cover in
global environmental change. Land Degradation and Development. 5: 71–78.
Turner, B.L., Matson, P.A., McCarthy, J.J., Corell, R.W., Christensen, L., and Eckley, N.
(2003). Illustrating the coupled human-environment system for vulnerability
analysis: three case studies. ProcNatlAcad Sci. 100(14):8080–5.
United Nations (2005). Rio Declaratrion on Environment and Development. Report of the
United Nations Conference on Environment and Development. Rio de Janeiro.
United Nations Centre for Human Settlements (UNCHS)- UN- Habitat. (2001). The state
of the world's cities. Nairobi, Kenya: UN- Habitat
United Nations University - Institute for Water Environment and Health (UNU-INWEH).
(2016). Waterbase: Landuse Maps (Africa).
http://www.waterbase.org/data/Global_Landuse_Data/af_landuse.zip
185
United Nations. (1994). Agenda 21: Program of Action for Sustainable Development; Rio
Declaration on Environment and Development; Statement of Forest Principles. Pp
166-185. UN Publication ISBN; 22-1-10059-4.
United Nations. (1994b). Agenda 21: Program of Action for Sustainable Development; Rio
Declaration on Environment and Development; Chapter 10 Integrated approach
to the planning and management of land resources. UN Publication ISBN; 22-1-
10059-4.
USACE. (1981). HEC-1 Flood Hydrograph Package: Program User's Manual. US Army
Corps of Engineering. Hydrologic Engineering Center, Davis, CA.
Vannote, R. L., Minshall G. W., Cummins K. W., Sedell J. R., and Cushing C. E. (1980).
The river continuum concept. Can. J. Fish. Aquat. Sci.37:130–137.
Vesilind, P.A., |Peirce, J.J. and Weiner, R. F., (1994) Environmental Engineering,
Butterworth Heinemann.
Vitousek, P.M., Mooney, H.A., Lubchenco, J., and Melillo, J.M. (1997). Human
domination of Earth's ecosystems. Science. 277(5335):494–9.
Vörösmarty, C.J., McIntyre, P.B., Gessner, M.O., Dudgeon, D., Prusevich, A., Green, P.,
Glidden, S., Bunn, S.E., Sullivan, C.A., and Liermann, C.R. (2010). Global threats
to human water security and riverbiodiversity. Nature, 467, 555–561.
Wagner, P. D., Kumar, S., and Schneider, K. (2013). An assessment of land use change
impacts on the water resources of the Mula and Mutha Rivers catchment upstream
of Pune, India. Hydrology and Earth System Sciences, 17(6), 2233–2246.
http://doi.org/10.5194/hess-17-2233-2013
186
Wang L. and Lyons J. (2003). Fish and Benthic Macroinvertebrate Assemblage as
Indicators of Stream Degradation in Urbanizing Watersheds. In: Simon TP (ed)
Biological Response Signatures: Indicator Patterns Using Aquatic Communities.
CRC Press, Boca Raton, FL. pp. 113-120
Ward J.V. (1997). An expansive perspective of riverine landscapes: pattern and process
across scales. River Ecosystems 6: 52–60
Ward J.V. and Stanford J.A. (1982). Thermal response in the evolutionary ecology of
aquatic insects. Annual Review of Entomology. 27: 97-117.
Watermeyer, Legge, Piesold, Uhlman, (1976). Upper Reservoir Scheme: Appraisal
Engineering Report. Tana and Athi River Development Authority. Nairobi.
Waters T.F. (1995). Sediment in Streams: Sources, Biological Effects and Control.
American Fisheries Society Monograph 7. American Fisheries Society, Bethesda,
Maryland.
Webster, J. R., M. E. Gurtz, J. J. Hains, J. L. Meyer, W. T. Swank, J. B. Waide, and J. B.
Wallace. (1983). Stability of stream ecosystems. In Stream Ecology. J. R. Barnes
and G. W. Minshall, editors. Plenum Press, New York, New York Pages 355-395
Weng Y. (2007). Spatio-temporal changes of landscape pattern in response to urbanization
Landsc. Urban Plann. 81: 341-353
Westra, L., Peter M., James R., Karr, et al. (2000). Ecological Integrity and the Aims of
the Global Ecological Integrity Project. In Ecological Integrity: Integrating
Environment, Conservation and Health, ed. David Pimentel, Laura Westra, and
Reed F. Noss. Washington, DC: Island Press.
187
Whitton, B.A. and Kelly, M.G. (1995). Use of algae and other plants for monitoring rivers.
Australian Journal of Ecology. 20: 45-56.
Wicklum, D. and Davies, R.W. (1995). Ecosystem health and integrity. Canadian Journal
of Botany. 73: 997-1000.
Wiens, J.A. (2002). Riverine landscapes: taking landscape ecology into the water. Freshw.
Biol. 47:501–15
Wijesekara, G.N., Gupta, A., Valeo, C., Hasbani, J.G., Qiao, Y., Delaney, P., and Marceau,
D.J. (2012). Assessing the impact of future land-use changes on hydrological
processes in the Elbow River watershed in southern Alberta. Canada J Hydrol. 412–
413: 220–232.
Wilson, E.O. (1998). Integrated science and the coming century of the environment.
Science. 279: 2048–2049.
Winter TC, Harvey JW, Lehn Franke O, Alley W.M. 1998. Ground water and surface
water: a single resource. Geological Survey Circular. 1139.
Withers, P.J. and Jarvie, H.P. (2008). Delivery and cycling of phosphorus in rivers: A
review. Sci. TotalEnviron., 400, 379–395.
World Bank (2007). Poverty and the Environment: Understanding Linkages at the
Household Level. World Bank.
World Resources institute (WRI); Department of Resource Surveys and Remote sensing;
ministry of Environment and Natural resources, Kenya; Central Bureau of
Statistics, Ministry of Planning and National development, Kenya; and
International Livestock Research Institute. 2007. Nature Benefits in Kenya, An
188
Atlas of Ecosystems and Human Well Being. Washington DC and Nairobi, World
Resources Institute.
Wright, J.F. (1995). Development and use of a system for predicting the macroinvertebrate
fauna in flowing waters. Australian Journal of Ecology.20: 181-197.
Wu, R. S. and Haith, D. A. (1993). Land use, climate, and water supply. Journal of Water
Resources Planning Management 119(6): 685-704.
Wunder, S., Engel, S., and Pagiola, S. (2008). Taking stock: A comparative analysis of
payments for environmental services programs in developed and developing
countries. Ecological Economics 65 (4): 834-852.
Xiang W. N. (1995). GIS-based analysis: injection of geographic information into
landscape planning. Landscape Urban Planning. 34: 1–10.
Xu, J., Sharma, R., Fang, J., and Xu, Y. (2008). Critical linkages between land-use
transition and human health in the Himalayan region. Environment International.
34: 239–247
Yang, H., Reichert, P., Abbaspour, K. and Zehnde, A. J. (2003). A water resources
threshold and its implications for food security, Environ. Sci.
Technol., 37(14), 3048–3054
Yu, D. and W. Changshan (2004). Understanding Population Segregation from Landsat
ETM+ Imagery: A Geographically Weighted Regression Approach. GIScience and
Remote Sensing 41(3): 187-206.
Yu., D. and W. Changshan. (2006). Incorporating remote sensing information in modelling
house values: a regression tree approach. Photogrammetric Engineering and
Remote Sensing 72: 129-138.
189
190
LIST OF APPENDICES
Appendix I: MiniSASS Table for Interpretation of Ecological Condition
Based On Composition Of Macro Invertebrates In Rivers
Ecological category River category
(Condition) Sandy Type Rocky Type
Unmodified
(NATURAL CONDITIONS) > 6.9 > 7.9
Largely natural/few modifications 5.8 to 6.9 6.8 to 7.9 (GOOD condition)
Moderately modified
(FAIR condition) 4.9 to 5.8 6.1 to 6.8
Largely modified
(POOR condition) 4.3 to 4.9 5.1 to 6.1
Seriously/critically modified
(VERY POOR condition) < 4.3 < 5.1
191
Appendix II: Long Term On-Site (In Situ) Water Quality Monitoring Data
From Water Resources Management Authority (Kiambu Regional Office)
Staff Sediment
Station Name Date Gauge Discharge Temp Cond TDS PH Load, TURBIDITY DO
NDARUGU 11/14/12 0.37 5.545 20.6 48.6 24 - - - -
NDARUGU 02/25/13 0.45 - 21 46.3 28 7.5 - - -
NDARUGU 07/02/13 0.35 - 18.3 66 33 7 - - -
NDARUGU 07/26/13 0.24 - 17.5 73 37 7.3 - - -
NDARUGU 08/20/13 0.21 - 19.7 75 38 7.1 - - -
NDARUGU 09/19/13 0.2 1.24 19.4 86 43 7 - - -
NDARUGU 10/25/13 0.1 0.004 27 99.5 50 7.8 - - -
NDARUGU 11/23/13 0.2 - 22.5 92 46 6.9 - - -
NDARUGU 02/20/14 0.2 - 21.6 75 38 7.1 - - -
NDARUGU 03/21/14 0.18 - 21.7 80 40 7.1 - - -
NDARUGU 05/20/14 0.18 0.947 20.1 68 34 7 - - -
NDARUGU 06/25/14 0.33 1.019 0.33 49 25 6.8 - - -
NDARUGU 07/17/14 0.28 1.487 - 54 27 - - - -
NDARUGU 08/26/14 0.19 0.945 18.6 65.5 33 - - - -
NDARUGU ------
NDARUGU 10/22/14 0.34 - 19.9 70 35 - - - -
NDARUGU 25/11/2014 0.5 - 22.4 38 29 6.4 - 40 6
NDARUGU 01/26/15 0.17 0.723 21.4 63.7 32 6.4 10.6 - -
NDARUGU ------
NDARUGU ------
NDARUGU ------
NDARUGU 01/29/15 - - 22 182 91 6.5 23.6 - -
NDARUGU 02/17/15 0.26 1.831 21.7 82 41 - - - -
NDARUGU 14/4/2015 0.15 - 22.5 87.5 45 - - - -
NDARUGU 18/5/2015 0.15 - 20.8 - - 7.1 - - -
NDARUGU ------
NDARUGU 29/7/2015 0.26 1.363 18.8 59.8 41 5.9 - 16.5 -
NDARUGU ------
NDARUGU 19/9/2015 0.2 - 21 74 55 5.9 - 7.14 2.31
192
NDARUGU 03/11/15 - - 21.8 87 43 - - 25 -
NDARUGU 19/01/2016 0.1 - 20.3 60.4 30 - - 175 -
NDARUGU 23/3/2016 - - 22.2 71 36 - - 16 -
NDARUGU ------
RUIRU 11/12/12 - - 20.1 52.6 26 - - - -
RUIRU 02/25/13 - - 21 55 35 8.5 - - -
RUIRU 03/28/13 - - 21.2 64.3 32 - - - -
RUIRU 07/02/13 - - 17.2 59 30 8 - - -
RUIRU 07/24/13 - - 17.5 63 31 7.2 - - -
RUIRU 08/19/13 - - 20.2 142 71 7.7 - - -
RUIRU 09/19/13 - - 19.3 69 35 6 - - -
RUIRU 10/22/13 - - 21 72 36 7.9 - - -
RUIRU 11/23/13 - - 22.8 84 42 7.3 - - -
RUIRU 02/17/14 - - 20.5 82 46 7 - - -
RUIRU 03/21/14 0.58 - 20.2 82 41 7.2 - - -
RUIRU 05/19/14 0.63 1.642 20.2 70 35 7.2 - - -
RUIRU 07/16/14 0.61 1.176 - 75 37 - - - -
RUIRU 08/27/14 0.63 0.883 19.2 72 36 - - - -
RUIRU ------
RUIRU 10/21/14 - - 20.1 69.3 35 - - - -
RUIRU 25/11/2014 0.74 - 20.2 68 51 6.3 - 21 6.1
RUIRU 01/27/15 0.62 1.676 20.1 60 30 6.4 7.5 - -
RUIRU 18/5/2015 - - 20.3 264 132 7.6 - - -
RUIRU 26/6/2015 - - 18.5 53 36 5.6 - 47.5 7.1
RUIRU 29/7/2015 - - 18.1 51 34 6 - 26.7 0
RUIRU ------
RUIRU 15/9/2015 - - 21 58 43 5.9 - 8.3 1.74
RUIRU 03/11/15 - - 22 192 97 - - 60 -
RUIRU 19/01/2016 1 - 20.1 75.8 38 - - 150 -
RUIRU 17/2/2016 0.72 - 19.8 71 35 - - 58 -
RUIRU 23/3/2016 - - 22.3 64 32 - - 83 -
193
Appendix III: Long Term Water Quality Monitoring Lab Data From Water
Resources Management Authority (Kiambu Regional Office)
Date
and
Time Total
Station Sample Turbid Iro Co M T. Alkali Chlor Sulph Nitr Nitr TD
Name d pH ity n Mn nd Na K g Hard Ca F nity ide ate ite ate S
21/05/2 8.0 3.4 86. 5. 0.0 11. 0.2 < < 53.
RUIRU 012 2 108 8 0.6 2 4.8 7 1 28 2 2 30 3 < 0.3 0.01 0.01 4
07/03/1 7.7 0.3 0.0 95. 4. 0.9 0.2 58.
RUIRU 2 8 11 3 6 2 6.6 1 8 28 9.6 2 36 6 < 0.3 0.8 16
06/15/1 7.3 2.3 59. 0.1 <0.0 37.
RUIRU 2 9 48.5 9 0.3 9 3.7 3 3.4 18 1.6 4 24 4 <0.3 1 7.4 2
07/18/1 7.4 61. 2. 1.9 0.1 <0.0 38.
RUIRU 2 5 45.8 2.9 0.3 5 4.3 3 5 20 4.8 8 20 4 <0.3 1 2.08 1
08/09/1 7.1 1.5 72. 3. 1.9 0.1 44.
RUIRU 2 5 40.2 7 0.2 2 7.1 6 5 16 3.2 2 18 6 <0.3 0.06 2.91 76
08/09/1 7.2 1.8 71. 3. 1.9 0.1 44.
RUIRU 2 3 41.5 7 0.2 9 8.1 3 5 14 2.4 2 22 2 <0.03 0.06 2.94 58
02/25/1 6.9 0.5 0.0 60. 0.9 0.1 <0.0 372
RUIRU 3 6 19.32 96 6 1 7.1 2 7 12 3.2 3 20 4 2.02 1 0.68 .6
03/28/1 6.9 2.8 68. 3.8 0.1 <0.0 42.
RUIRU 3 6 97.3 4 0.1 5 1 3 9 28 4.8 9 24 4 <0.3 1 0.93 5
07/02/1 6.8 0.7 0.1 61. 7.3 1.4 0.2 <0.0 37.
RUIRU 3 6 15.5 2 8 1 2 2 6 12 2.4 6 20 3 0.63 1 0.95 88
07/24/1 7.0 59. 0.9 0.2 <0.0 37.
RUIRU 3 7 26.3 0.5 0.1 9 4.1 4 7 16 4.8 9 24 3 0.2 1 0.85 14
08/19/1 6.8 0.4 0.0 63. 2.4 0.1 <0.0 39.
RUIRU 3 8 23.2 8 8 5 6 4 3 20 4 8 24 1 3.37 1 0.7 4
11/23/1
RUIRU 3 6.9 0.2 1.23
01/08/1 6.9 <0. 86. 10. 2.4 0.2 <0.0 53.
RUIRU 4 2 6.9 0.7 01 4 5 3 20 4 3 26 10 2.1 1 0.77 57
02/17/1 6.7 3.7 0.0 74. 7.8 0.2 46.
RUIRU 4 3 77.1 3 2 4 4 3.4 20 2.4 6 24 4 0.5 0.03 1.41 13
03/21/1 7.3 0.8 <0. 96. 10. 0.9 0.2 59.
RUIRU 4 5 2 3 01 1 78 8 22 7.2 1 24 8 <0.3 0.03 5.33 6
04/28/1 7.1 0.5 0.0 79. 10. 4. 1.9 0.2 <0.0 49.
RUIRU 4 2 9.2 3 8 5 4 3 5 18 4 6 44 2 0.51 1 2.35 3
NDARU 21/05/2 7.0 3.4 39. 2. 0.4 0.1 < 24.
GU 012 8 109.9 9 0.1 5 2.2 3 9 12 4 3 12 1 < 0.3 0.01 1.74 5
NDARU 06/03/1 7.5 0.3 0.0 85. 11. 3. 0.9 0.2 51.
GU 2 8 8 3 4 6 2 4 7 14 5 36 2 0.71 0.78 4
194
NDARU 06/12/1 7.3 3.0 44. 2. 1.9 0.1 <0.0 27.
GU 2 2 53.3 4 0.3 1 2 7 6 14 2.4 1 10 1 0.4 1 4.9 3
NDARU 07/18/1 7.3 51. 1. 2.9 0.1 <0.0 32.
GU 2 2 52.4 3 0.2 8 3.8 8 2 20 3.2 7 18 3 <0.3 1 1.62 1
NDARU 09/28/1 6.7 0.8 <0. 75. 3. 1.4 <0.0 46.
GU 2 5 13.3 2 01 4 8 3 6 16 4 0.2 26 7 <0.3 1 1.7 75
NDARU 10/24/1 0.0 <0. 77. 2. 1.9 0.1 <0.0 47.
GU 2 6.5 33.7 26 01 2 7 4 5 20 4.8 9 22 7 0.3 1 3.3 9
NDARU 02/25/1 7.0 1.2 0.0 62. 8.6 0.9 0.1 <0.0 38.
GU 3 1 28.7 7 6 7 2 2 7 10 2.4 1 26 1 <0.03 1 0.68 9
NDARU 03/26/1 7.6 0.9 0.0 119 1.9 14. 0.1 74.
GU 3 6 32.9 9 8 .6 5.4 3 5 44 4 7 48 4 1.3 0.02 1.11 1
NDARU 07/26/1 7.2 <0. 68. 1.9 0.2 <0.0 42.
GU 3 3 25.8 0.6 01 2 5 4 5 18 4 9 24 5 4.34 1 0.66 3
NDARU 08/20/1 7.0 0.0 72. 10. 0.9 0.1 <0.0 44.
GU 3 2 14.2 0.2 4 2 2 0 7 14 4 9 32 1 1.2 1 0.65 8
NDARU 11/23/1
GU 3 14.3 0.9 1.16
NDARU 02/20/1 8.2 1.3 0.0 92. 6.4 3.4 0.2 57.
GU 4 5 49.9 3 8 6 8 1 32 7.2 6 10 1 <0.3 0.04 9.67 41
NDARU 03/21/1 8.0 0.8 109 3.8 11. 0.2
GU 4 3 14.9 9 0.3 .7 5 9 44 2 1 20 6 <0.3 0.04 7.07 68
NDARU 04/30/1 6.7 0.0 3. 1.4 0.2 <0.0 48.
GU 4 4 7 0.4 8 79 9.9 4 6 20 5.6 3 28 2 0.74 1 2.29 98
195
Appendix IV: NEMA Water Quality Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
196
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Appendix IV: NEMA Water Quality Appendix IV: NEMA Water Quality
Standards For Domestic Use Standards For Domestic Use
Source: Legal notice no. 120, 2006, NEMA Water Quality Regulations
197
Appendix V: SWAT Tables
Detailed Landuse, Soil and Slope Distribution per sub-basin
Ruiru Watershed
Number of subbasins: 23
------
Area [ha]
Watershed 43866.25
------
Area [ha] %Watershed
Landuse
SAVA 3344.91 7.63
CRDY 2688.76 6.13
FODB 13345.43 30.42
FOEB 23155.17 52.79
CRWO 1322.57 3.01
Soil
Nh2-2c-848 6568.35 14.97
Bc14-2bc-440 5183.33 11.82
Fo48-2ab-42 767.36 1.75
Tm9-2c-948 21874.52 49.87
198
Tm10-2bc-941 9463.28 21.57
Slope
0-10.0 22497.31 51.29
10.0-9999 21359.53 48.69
------
------
Area [ha] %Watershed %Subbasin
Subbasin 1 5262.03 12.00
Landuse
CRDY 295.14 0.67 5.61
FODB 41.92 0.10 0.80
FOEB 4921.55 11.22 93.53
CRWO 3.42 0.01 0.07
Soil
Tm9-2c-948 1129.23 2.57 21.46
Tm10-2bc-941 4132.81 9.42 78.54
Slope
0-10.0 2577.55 5.88 48.98
10.0-9999 2684.48 6.12 51.02
199
------
Area [ha] %Watershed %Subbasin
Subbasin 2 5665.82 12.92
Landuse
SAVA 96.67 0.22 1.71
FODB 2967.65 6.77 52.38
CRDY 323.37 0.74 5.71
CRWO 0.86 0.00 0.02
FOEB 2277.28 5.19 40.19
Soil
Nh2-2c-848 1868.36 4.26 32.98
Tm9-2c-948 3195.20 7.28 56.39
Tm10-2bc-941 602.26 1.37 10.63
Slope
0-10.0 2283.27 5.21 40.30
10.0-9999 3382.55 7.71 59.70
------
Area [ha] %Watershed %Subbasin
Subbasin 3 1487.67 3.39
Landuse
200
FODB 480.78 1.10 32.32
FOEB 969.25 2.21 65.15
CRWO 37.64 0.09 2.53
Soil
Tm9-2c-948 760.52 1.73 51.12
Tm10-2bc-941 727.15 1.66 48.88
Slope
0-10.0 505.59 1.15 33.99
10.0-9999 982.09 2.24 66.01
------
Area [ha] %Watershed %Subbasin
Subbasin 4 1153.18 2.63
Landuse
SAVA 82.13 0.19 7.12
FODB 219.86 0.50 19.07
CRDY 103.51 0.24 8.98
FOEB 747.69 1.70 64.84
Soil
Tm9-2c-948 1153.18 2.63 100.00
Slope
201
0-10.0 384.96 0.88 33.38
10.0-9999 768.22 1.75 66.62
------
Area [ha] %Watershed %Subbasin
Subbasin 5 988.93 2.25
Landuse
CRDY 12.83 0.03 1.30
FODB 39.35 0.09 3.98
FOEB 936.75 2.14 94.72
Soil
Tm9-2c-948 705.77 1.61 71.37
Tm10-2bc-941 283.16 0.65 28.63
Slope
0-10.0 347.32 0.79 35.12
10.0-9999 641.61 1.46 64.88
------
Area [ha] %Watershed %Subbasin
Subbasin 6 2327.75 5.31
Landuse
FODB 404.64 0.92 17.38
202
FOEB 1923.11 4.38 82.62
Soil
Tm9-2c-948 224.99 0.51 9.67
Tm10-2bc-941 2102.76 4.79 90.33
Slope
0-10.0 1105.28 2.52 47.48
10.0-9999 1222.48 2.79 52.52
------
Area [ha] %Watershed %Subbasin
Subbasin 7 1140.35 2.60
Landuse
CRDY 290.01 0.66 25.43
FODB 18.82 0.04 1.65
FOEB 831.52 1.90 72.92
Soil
Tm9-2c-948 1140.35 2.60 100.00
Slope
0-10.0 452.55 1.03 39.68
10.0-9999 687.80 1.57 60.32
------
203
Area [ha] %Watershed %Subbasin
Subbasin 8 1181.41 2.69
Landuse
SAVA 84.69 0.19 7.17
FODB 313.10 0.71 26.50
CRDY 37.64 0.09 3.19
FOEB 745.98 1.70 63.14
Soil
Tm9-2c-948 944.45 2.15 79.94
Tm10-2bc-941 236.97 0.54 20.06
Slope
0-10.0 491.90 1.12 41.64
10.0-9999 689.51 1.57 58.36
------
Area [ha] %Watershed %Subbasin
Subbasin 9 3901.83 8.89
Landuse
CRDY 95.81 0.22 2.46
FODB 279.74 0.64 7.17
FOEB 3526.27 8.04 90.37
204
Soil
Tm9-2c-948 2606.64 5.94 66.81
Tm10-2bc-941 1295.19 2.95 33.19
Slope
0-10.0 1515.05 3.45 38.83
10.0-9999 2386.78 5.44 61.17
------
Area [ha] %Watershed %Subbasin
Subbasin 10 1886.32 4.30
Landuse
CRDY 66.73 0.15 3.54
FODB 238.68 0.54 12.65
FOEB 1580.92 3.60 83.81
Soil
Tm9-2c-948 1803.34 4.11 95.60
Tm10-2bc-941 82.98 0.19 4.40
Slope
0-10.0 770.78 1.76 40.86
10.0-9999 1115.54 2.54 59.14
------
205
Area [ha] %Watershed %Subbasin
Subbasin 11 1096.72 2.50
Landuse
CRDY 10.27 0.02 0.94
FODB 492.75 1.12 44.93
FOEB 593.70 1.35 54.13
Soil
Tm9-2c-948 1096.72 2.50 100.00
Slope
0-10.0 593.70 1.35 54.13
10.0-9999 503.02 1.15 45.87
------
Area [ha] %Watershed %Subbasin
Subbasin 12 1375.61 3.14
Landuse
CRDY 78.70 0.18 5.72
FODB 550.07 1.25 39.99
FOEB 746.83 1.70 54.29
Soil
Nh2-2c-848 1206.22 2.75 87.69
206
Bc14-2bc-440 53.89 0.12 3.92
Tm9-2c-948 115.49 0.26 8.40
Slope
0-10.0 787.89 1.80 57.28
10.0-9999 587.71 1.34 42.72
------
Area [ha] %Watershed %Subbasin
Subbasin 13 2473.18 5.64
Landuse
FODB 933.32 2.13 37.74
FOEB 1539.86 3.51 62.26
Soil
Nh2-2c-848 597.98 1.36 24.18
Bc14-2bc-440 299.42 0.68 12.11
Tm9-2c-948 1575.79 3.59 63.71
Slope
0-10.0 1286.64 2.93 52.02
10.0-9999 1186.55 2.70 47.98
------
Area [ha] %Watershed %Subbasin
207
Subbasin 14 532.96 1.21
Landuse
SAVA 86.40 0.20 16.21
FODB 249.80 0.57 46.87
CRDY 45.34 0.10 8.51
FOEB 151.42 0.35 28.41
Soil
Tm9-2c-948 532.96 1.21 100.00
Slope
0-10.0 183.07 0.42 34.35
10.0-9999 349.89 0.80 65.65
------
Area [ha] %Watershed %Subbasin
Subbasin 15 2852.16 6.50
Landuse
CRDY 319.09 0.73 11.19
FODB 1999.25 4.56 70.10
FOEB 388.39 0.89 13.62
CRWO 145.43 0.33 5.10
Soil
208
Nh2-2c-848 175.37 0.40 6.15
Tm9-2c-948 2676.78 6.10 93.85
Slope
0-10.0 1218.20 2.78 42.71
10.0-9999 1633.96 3.72 57.29
------
Area [ha] %Watershed %Subbasin
Subbasin 16 59.03 0.13
Landuse
FOEB 59.03 0.13 100.00
Soil
Tm9-2c-948 59.03 0.13 100.00
Slope
0-10.0 26.52 0.06 44.93
10.0-9999 32.51 0.07 55.07
------
Area [ha] %Watershed %Subbasin
Subbasin 17 82.13 0.19
Landuse
CRDY 9.41 0.02 11.46
209
FODB 40.21 0.09 48.96
FOEB 32.51 0.07 39.58
Soil
Tm9-2c-948 82.13 0.19 100.00
Slope
0-10.0 24.81 0.06 30.21
10.0-9999 57.32 0.13 69.79
------
Area [ha] %Watershed %Subbasin
Subbasin 18 104.37 0.24
Landuse
CRDY 9.41 0.02 9.02
FODB 65.02 0.15 62.30
FOEB 29.94 0.07 28.69
Soil
Tm9-2c-948 104.37 0.24 100.00
Slope
0-10.0 27.38 0.06 26.23
10.0-9999 76.99 0.18 73.77
------
210
Area [ha] %Watershed %Subbasin
Subbasin 19 1337.11 3.05
Landuse
SAVA 544.94 1.24 40.75
FODB 790.46 1.80 59.12
FOEB 1.71 0.00 0.13
Soil
Bc14-2bc-440 1058.22 2.41 79.14
Nh2-2c-848 278.89 0.64 20.86
Slope
0-10.0 1269.53 2.89 94.95
10.0-9999 67.58 0.15 5.05
------
Area [ha] %Watershed %Subbasin
Subbasin 20 2440.67 5.56
Landuse
CRDY 195.05 0.44 7.99
FODB 1402.13 3.20 57.45
SAVA 159.12 0.36 6.52
FOEB 684.38 1.56 28.04
211
Soil
Nh2-2c-848 473.08 1.08 19.38
Tm9-2c-948 1967.60 4.49 80.62
Slope
0-10.0 897.39 2.05 36.77
10.0-9999 1543.28 3.52 63.23
------
Area [ha] %Watershed %Subbasin
Subbasin 21 1175.42 2.68
Landuse
CRDY 1.71 0.00 0.15
FODB 878.57 2.00 74.75
FOEB 295.14 0.67 25.11
Soil
Nh2-2c-848 1175.42 2.68 100.00
Slope
0-10.0 705.77 1.61 60.04
10.0-9999 469.66 1.07 39.96
------
Area [ha] %Watershed %Subbasin
212
Subbasin 22 1436.34 3.27
Landuse
SAVA 489.33 1.12 34.07
FODB 772.49 1.76 53.78
FOEB 171.95 0.39 11.97
CRWO 2.57 0.01 0.18
Soil
Bc14-2bc-440 643.32 1.47 44.79
Nh2-2c-848 793.03 1.81 55.21
Slope
0-10.0 1272.95 2.90 88.62
10.0-9999 163.40 0.37 11.38
------
Area [ha] %Watershed %Subbasin
Subbasin 23 3905.25 8.90
Landuse
SAVA 1801.63 4.11 46.13
FODB 166.82 0.38 4.27
CRDY 794.74 1.81 20.35
213
CRWO 1132.65 2.58 29.00
Soil
Bc14-2bc-440 3128.48 7.13 80.11
Fo48-2ab-42 767.36 1.75 19.65
Slope
0-10.0 3769.23 8.59 96.52
10.0-9999 126.61 0.29 3.24
------
Ndarugu Watershed
Number of subbasins: 19
------
Area [ha]
Watershed 30707.32
------
Area [ha] %Watershed
Landuse
CRDY 3067.74 9.99
CRWO 4806.06 15.65
SHRB 69.29 0.23
SAVA 4540.01 14.78
214
FODB 9742.16 31.73
FOEB 8482.05 27.62
Soil
Nh2-2c-848 13949.39 45.43
Bc14-2bc-440 1610.86 5.25
Fo48-2ab-42 5520.39 17.98
Tm9-2c-948 9195.51 29.95
Tm10-2bc-941 431.16 1.40
Slope
0-10.0 17138.61 55.81
10.0-9999 13568.71 44.19
------
------
Area [ha] %Watershed %Subbasin
Subbasin 1 5841.19 19.02
Landuse
SAVA 54.75 0.18 0.94
FODB 1442.33 4.70 24.69
CRDY 717.74 2.34 12.29
CRWO 911.94 2.97 15.61
215
FOEB 2714.43 8.84 46.47
Soil
Nh2-2c-848 797.30 2.60 13.65
Tm9-2c-948 4612.73 15.02 78.97
Tm10-2bc-941 431.16 1.40 7.38
Slope
0-10.0 2359.40 7.68 40.39
10.0-9999 3481.79 11.34 59.61
------
Area [ha] %Watershed %Subbasin
Subbasin 2 3642.62 11.86
Landuse
CRDY 604.82 1.97 16.60
FODB 567.18 1.85 15.57
SAVA 155.70 0.51 4.27
CRWO 576.59 1.88 15.83
FOEB 1738.33 5.66 47.72
Soil
Nh2-2c-848 775.92 2.53 21.30
Tm9-2c-948 2866.70 9.34 78.70
216
Slope
0-10.0 1210.50 3.94 33.23
10.0-9999 2432.12 7.92 66.77
------
Area [ha] %Watershed %Subbasin
Subbasin 3 2544.19 8.29
Landuse
FODB 356.73 1.16 14.02
FOEB 2042.02 6.65 80.26
CRWO 145.43 0.47 5.72
Soil
Nh2-2c-848 1823.88 5.94 71.69
Tm9-2c-948 720.31 2.35 28.31
Slope
0-10.0 766.51 2.50 30.13
10.0-9999 1777.68 5.79 69.87
------
Area [ha] %Watershed %Subbasin
Subbasin 4 1525.31 4.97
217
Landuse
CRDY 70.15 0.23 4.60
FODB 459.39 1.50 30.12
FOEB 995.77 3.24 65.28
Soil
Nh2-2c-848 529.54 1.72 34.72
Tm9-2c-948 995.77 3.24 65.28
Slope
0-10.0 526.97 1.72 34.55
10.0-9999 998.34 3.25 65.45
------
Area [ha] %Watershed %Subbasin
Subbasin 5 1847.83 6.02
Landuse
CRDY 152.27 0.50 8.24
FODB 1505.64 4.90 81.48
218
FOEB 189.92 0.62 10.28
Soil
Nh2-2c-848 1847.83 6.02 100.00
Slope
0-10.0 1053.09 3.43 56.99
10.0-9999 794.74 2.59 43.01
------
Area [ha] %Watershed %Subbasin
Subbasin 6 1841.84 6.00
Landuse
FODB 1841.84 6.00 100.00
Soil
Nh2-2c-848 1841.84 6.00 100.00
Slope
0-10.0 596.27 1.94 32.37
10.0-9999 1245.57 4.06 67.63
------
Area [ha] %Watershed %Subbasin
Subbasin 7 1377.32 4.49
Landuse
219
SAVA 274.61 0.89 19.94
FODB 905.95 2.95 65.78
CRWO 196.76 0.64 14.29
Soil
Nh2-2c-848 1305.46 4.25 94.78
Bc14-2bc-440 71.86 0.23 5.22
Slope
0-10.0 1319.14 4.30 95.78
10.0-9999 58.17 0.19 4.22
------
Area [ha] %Watershed %Subbasin
Subbasin 8 1089.02 3.55
Landuse
SAVA 142.86 0.47 13.12
FODB 221.57 0.72 20.35
CRDY 396.09 1.29 36.37
CRWO 328.50 1.07 30.16
Soil
220
Nh2-2c-848 1033.42 3.37 94.89
Bc14-2bc-440 55.61 0.18 5.11
Slope
0-10.0 1071.91 3.49 98.43
10.0-9999 17.11 0.06 1.57
------
Area [ha] %Watershed %Subbasin
Subbasin 9 1106.99 3.60
Landuse
SHRB 65.87 0.21 5.95
SAVA 445.70 1.45 40.26
CRWO 595.41 1.94 53.79
Soil
Fo48-2ab-42 1106.99 3.60 100.00
Slope
0-10.0 1063.36 3.46 96.06
10.0-9999 43.63 0.14 3.94
------
Area [ha] %Watershed %Subbasin
Subbasin 10 1112.97 3.62
221
Landuse
SAVA 810.14 2.64 72.79
CRWO 302.84 0.99 27.21
Soil
Fo48-2ab-42 1112.97 3.62 100.00
Slope
0-10.0 1067.63 3.48 95.93
10.0-9999 45.34 0.15 4.07
------
Area [ha] %Watershed %Subbasin
Subbasin 11 1784.52 5.81
Landuse
FODB 1366.20 4.45 76.56
FOEB 418.33 1.36 23.44
Soil
Nh2-2c-848 1784.52 5.81 100.00
Slope
0-10.0 657.86 2.14 36.86
10.0-9999 1126.66 3.67 63.14
------
222
Area [ha] %Watershed %Subbasin
Subbasin 12 305.40 0.99
Landuse
FODB 305.40 0.99 100.00
Soil
Nh2-2c-848 305.40 0.99 100.00
Slope
0-10.0 82.98 0.27 27.17
10.0-9999 222.42 0.72 72.83
------
Area [ha] %Watershed %Subbasin
Subbasin 13 299.42 0.98
Landuse
FODB 143.72 0.47 48.00
FOEB 155.70 0.51 52.00
Soil
223
Nh2-2c-848 299.42 0.98 100.00
Slope
0-10.0 129.18 0.42 43.14
10.0-9999 170.24 0.55 56.86
------
Area [ha] %Watershed %Subbasin
Subbasin 14 443.99 1.45
Landuse
CRDY 28.23 0.09 6.36
FODB 189.06 0.62 42.58
FOEB 226.70 0.74 51.06
Soil
Nh2-2c-848 443.99 1.45 100.00
Slope
0-10.0 236.97 0.77 53.37
10.0-9999 207.03 0.67 46.63
------
Area [ha] %Watershed %Subbasin
Subbasin 15 961.56 3.13
224
Landuse
SAVA 350.75 1.14 36.48
FODB 437.15 1.42 45.46
FOEB 0.86 0.00 0.09
CRWO 172.81 0.56 17.97
Soil
Nh2-2c-848 961.56 3.13 100.00
Slope
0-10.0 620.22 2.02 64.50
10.0-9999 341.33 1.11 35.50
------
Area [ha] %Watershed %Subbasin
Subbasin 16 248.09 0.81
Landuse
SAVA 4.28 0.01 1.72
CRDY 2.57 0.01 1.03
CRWO 241.24 0.79 97.24
Soil
Bc14-2bc-440 62.45 0.20 25.17
Nh2-2c-848 185.64 0.60 74.83
225
Slope
0-10.0 183.07 0.60 73.79
10.0-9999 65.02 0.21 26.21
------
Area [ha] %Watershed %Subbasin
Subbasin 17 3816.28 12.43
Landuse
SHRB 3.42 0.01 0.09
SAVA 1382.45 4.50 36.23
CRDY 1095.86 3.57 28.72
CRWO 1334.54 4.35 34.97
Soil
Bc14-2bc-440 1420.95 4.63 37.23
Nh2-2c-848 13.69 0.04 0.36
Fo48-2ab-42 2381.65 7.76 62.41
Slope
0-10.0 3363.73 10.95 88.14
10.0-9999 452.55 1.47 11.86
------
226
Area [ha] %Watershed %Subbasin
Subbasin 18 307.97 1.00
Landuse
SAVA 307.97 1.00 100.00
Soil
Fo48-2ab-42 307.97 1.00 100.00
Slope
0-10.0 298.56 0.97 96.94
10.0-9999 9.41 0.03 3.06
------
Area [ha] %Watershed %Subbasin
Subbasin 19 610.81 1.99
Landuse
SAVA 610.81 1.99 100.00
Soil
Fo48-2ab-42 610.81 1.99 100.00
Slope
0-10.0 531.25 1.73 86.97
10.0-9999 79.56 0.26 13.03
227
HRU Distribution Summary
Ruiru Watershed
Detailed Landuse/Soil/Slope Distribution 25 August 2016 12.37
Using percentage of subbasin as a threshold
Multiple HRUs Landuse/Soil/Slope option Thresholds: 10/10/10 [%]
Number of HRUs: 129
Number of subbasins: 23
Numbers in parentheses are corresponding values before HRU creation
------
Area [ha]
Watershed 43866.25
------
Area [ha] %Watershed
Landuse
SAVA 3017.06 (3344.91) 6.88 (7.63)
CRDY 1472.82 (2688.76) 3.36 (6.13)
FODB 13553.31 (13345.43) 30.90 (30.42)
FOEB 24636.89 (23155.17) 56.16 (52.79)
CRWO 1186.18 (1322.57) 2.70 (3.01)
228
Soil
Nh2-2c-848 6569.36 (6568.35) 14.98 (14.97)
Bc14-2bc-440 5101.81 (5183.33) 11.63 (11.82)
Fo48-2ab-42 802.73 (767.36) 1.83 (1.75)
Tm9-2c-948 21812.00 (21874.52) 49.72 (49.87)
Tm10-2bc-941 9580.36 (9463.28) 21.84 (21.57)
Slope
0-10.0 22777.72 (22497.31) 51.93 (51.29)
10.0-9999 21088.53 (21359.53) 48.07 (48.69)
Ndarugu Watershed
Using percentage of subbasin as a threshold
Multiple HRUs Landuse/Soil/Slope option Thresholds: 10/10/10 [%]
Number of HRUs: 96
Number of subbasins: 19
Numbers in parentheses are corresponding values before HRU creation
------
Area [ha]
Watershed 30707.32
------
229
Area [ha] %Watershed
Landuse
SAVA 4355.04 (4540.01) 14.18 (14.78)
CRDY 2849.30 (3067.74) 9.28 (9.99)
FODB 9973.35 (9742.16) 32.48 (31.73)
FOEB 8788.75 (8482.05) 28.62 (27.62)
CRWO 4740.88 (4806.06) 15.44 (15.65)
SHRB (69.29) (0.23)
Soil
Nh2-2c-848 14196.69 (13949.39) 46.23 (45.43)
Bc14-2bc-440 1599.35 (1610.86) 5.21 (5.25)
Fo48-2ab-42 5528.42 (5520.39) 18.00 (17.98)
Tm9-2c-948 9179.92 (9195.51) 29.89 (29.95)
Tm10-2bc-941 202.94 (431.16) 0.66 (1.40)
Slope
0-10.0 17302.17 (17138.61) 56.35 (55.81)
10.0-9999 13405.14 (13568.71) 43.65 (44.19)
230