Urban Wetland Ecology and Floods in Kumasi, Ghana

URBAN WETLAND ECOLOGY AND FLOODS IN KUMASI, GHANA

Benjamin Betey Campion

Institut für Geographie an der Universität Bremen May, 2012

URBAN WETLAND ECOLOGY AND FLOODS IN KUMASI, GHANA

Benjamin Betey Campion

Institut für Geographie an der Universität Bremen May, 2012

Submitted for the degree of Doctor of Natural Sciences (Dr. rer. nat.) Supervisors: 1. Prof. Dr. Jörg-Friedhelm Venzke 2. Prof. Dr. Michael Flitner

DECLARATION

I, Benjamin Betey Campion, hereby declare that this thesis has been independently written by me and is original. Where information has been derived from other sources, I confirm that this has been indicated in the thesis.

Bremen, …27. 09.2012.…………………….…….. …………. Date Benjamin Betey Campion

iii

ACKNOWLEDGEMENT My soul gives thanks to the Lord my God for taking me this far. May the Name of God be blessed forever! I am very grateful to Prof. Dr. Jörg-Friedhelm Venzke for the bold initiative he took and accepted me as his student. Finding such a father figure from the internet is a million dollar lottery ticket. Prof. Venzke, your friendship, encouragement, sense of humour, humility and simplicity have been my inspiration and support in writing this thesis. I love your open door approach to supervision. I am also very grateful to Prof. Dr. Michael Flitner for accepting to be the second supervisor. Thank you for the invaluable suggestions and criticisms in shaping me into an up- and- coming scientis t. I also received tremendous moral and scientific support from Bruni Hans, Birte Habel, PD. Dr. Karin Steinecke, Dr. Steffen Schwantz and Dr. Marco Langer. I love this small family that nurtured me. You will always be cherished for everything you did to make my stay in Uni- Bremen, UFT, Abeitsgruppe Klima- und Vegetationsgeographie, and above all, Room 0060 a wonderful one. I am also grateful and indebted to my colleagues in the Faculty of Renewable Natural Resources, KNUST, for their support. Prof. Steve Amisah, thank you for advising me to get a PhD. For the second time, the Katholischer Akademischer Ausländer-Dienst (KAAD) deemed me fit for their highly competitive scholarship. I am highly indebted to you and appreciate your contribution in shaping my life this far. Marko, Simone and the many others at the KAAD office, thank you and God will richly bless you for the trust and confidence you had in me. I look forward to a future of sharing the task of developing humanity with you. Thank you. To my inspirational friends: Dr. Callistus Mahama, Dr. Nana Ato Arthur, Seth Mensah Abobi, Dr. Julius Fobil, Cephas Kwesi Zuh, Isaac Sackey, Kwabena Asubonteng, and Dr. David Yegbey, thank you. I will never forget the continuous and diverse support of the Campions: Kofi, Jatuat, Teni, Jakper, Yaw, Yenukwah, Victor, Dumbian, Lardi, Tisob, Sanjaung, Kibirsiar and Cynthia. You are the pillar behind this work. Thank you and let us uphold the family spirit. I could not have made it without a partner. A friend for life, my hope in times of despair, my all: Elma.

“You placed gold on my finger You brought love like I've never known You gave life to our children And to me, a reason to go on. …… But most of all, you're my best friend” (Don Willia ms ). Palmaak and Iloam, you are my motivation and reason for studying. Thank you for always accepting me after the long periods away from home.

DEDICATION To my dad who said, ‘Ben, go to school’; and to my mum who made it all possible.

John Limuub Campion & Mary Konjit Lambon

LIST OF ACRONYMS

As Arsenic Cd Cadmium CF Contamina tio n Factor Cr Chromium Cu Copper DEMC District Environmental Management Committee DO Dissolved Oxygen EF Enrichment Factor EPA Environmental Protection Agency FRNR Faculty of Renewable Natural Resources Hg Mercury HSD Hydrological Services Department IFRC International Federation of Red Cross and Red Crescent Societies Igeo Geo-accumulation Index IPCC Intergover nme nta l Panel on Climate Change ITCZ Intertropical Convergence Zone KMA Kumasi Metropolitan Assembly KNUST Kwame Nkrumah University of Science and Technology MDGs Mille nnium Development Goals MDS non- Metric Multid ime ns io na l Scaling NADMO National Disaster Management Organisation NEPAD New Partnership for African Development Ni Nickel P Phosphorus Pb Lead PCA Principa l Components Analys is PLI Pollution Load Index PRS Poverty Reduction Strategy TCPD Town and Country Planning Department TDS Total Dissolved Solids Zn Zinc

TABLE OF CONTENTS DECLARATION ...... iii

ACKNOWLEDGEMENT ...... iv

DEDICATION ...... v

LIST OF ACRONYMS ...... vi

TABLE OF CONTENTS...... vii

LIST OF FIGURES ...... xi

LIST OF TABLES ...... xiv

LIST OF PHOTOGRAPHS ...... xv

APPENDICES ...... xvi

LIST OF PUBLICATIONS FROM THIS THESIS ...... xvii

ABSTRACT...... xviii

ZUSAMMENFASSUNG ...... xix

1 RESEARCH THEME AND THEORETICAL FRAMEWORK ...... 1 1.1 INTRODUCTION...... 1

1.2 RESEARCH BACKGROUND AND JUSTIFICATION ...... 2

1.2.1 Urban Environmental Issues ...... 2

1.2.2 Research Justification ...... 4

1.3 RESEARCH GOAL ...... 7

1.4 RESEARCH OBJECTIVES ...... 7

1.5 RESEARCH QUESTIONS...... 7

1.6 CONCEPTUAL AND THEORETICAL FRAMEWORKS ...... 8

1.7 HYPOTHESES ...... 12

2 WETLANDS, URBANISATION AND FLOOD NEXUS ...... 14 2.1 INTRODUCTION...... 14

2.2 WETLANDS ...... 14

2.2.1 Morphology and Hydrodynamics of Wetlands ...... 17 vii

2.2.2 Distribution, Values and Degradation of Wetlands ...... 18

2.3 URBANISATION ...... 21

2.3.1 Concept and History of Urbanisation ...... 21

2.3.2 Modern Trends in Urbanisation ...... 23

2.3.3 Urbanisation Trends in sub-Saharan Africa...... 25

2.4 URBANISATION AND THE ENVIRONMENT ...... 26

2.5 WETLANDS AND URBANISATION ...... 28

2.5.1 Urban Wetlands Vegetation ...... 31

2.5.2 Wetlands, Urbanisation and Agriculture...... 32

2.5.3 Wetlands, Urbanisation and Floods ...... 35

2.6 URBAN POVERTY AND WETLANDS...... 37

2.6.1 Urban Poverty in Africa ...... 37

2.6.2 Urban Poverty and the Environmental Kuznets Curve ...... 38

2.7 CLIMATE CHANGE, URBANISATION AND WETLANDS...... 40

2.8 REMOTE SENSING AND URBAN LANDSCAPE PLANNING AND MANAGEMENT...... 42

3 POLITICAL ECOLOGY OF WETLANDS IN KUMASI...... 45 3.1 INTRODUCTION...... 45

3.2 METHODOLOGY ...... 46

3.2.1 Physical and Socio-economic Survey of Wetland Communities of Kumasi..... 46

3.2.2 Data Analysis ...... 48

3.3 RESULTS...... 48

3.3.1 The Physical Characteristics of Kumasi ...... 48

3.3.2 Socio-economic Characteristics of Wetland Communities in Kumasi ...... 52

3.3.3 Community Efforts at Wetland Management in Kumasi ...... 60

3.4 DISCUSSION ...... 63

4 FLOODS AND ENVIRONMENTAL MANAGEMENT IN KUMASI, GHANA ...... 67

viii

4.1 INTRODUCTION...... 67

4.2 ENVIRONMENTAL MANAGEMENT AND THE NATIONAL ENVIRONMENTAL ACTION PLAN ...... 69

4.2.1 The Environmental Protection Agency...... 69

4.2.2 Water Resources Commission ...... 70

4.2.3 Lands Commission...... 72

4.2.4 Local Government and Environmental Management ...... 73

4.3 WETLANDS, FLOOD PREVENTION AND MANAGEMENT ...... 74

4.3.1 Stakeholders’ Capacity for Managing Wetlands and Floods...... 74

4.4 DISCUSSION ...... 79

5 WETLAND ECOLOGY AND VEGETATION GEOGRAPHY IN THE KUMASI METROPOLIS ...... 85 5.1 INTRODUCTION...... 85

5.2 METHODOLOGY ...... 86

5.2.1 Reconnaissance Survey of Wetlands in Kumasi and Selection of Study Sites.. 86

5.2.2 Ecological Survey of Wetlands in Kumasi ...... 87

5.2.3 Data Analysis ...... 91

5.3 RESULTS...... 91

5.3.1 Anthropogenic Activities in the Wetland Areas of Kumasi ...... 91

5.3.2 Wetland Edaphic Characteristics, Heavy Metals and Stream Physicochemical Characteristics ...... 94

5.3.3 Wetland Vegetation of Kumasi...... 107

5.4 DISCUSSION ...... 123

6 CLIMATE, LANDUSE CHANGES AND FLOODING IN THE KUMASI METROPOLIS ...... 128 6.1 INTRODUCTION...... 128

6.2 METHODOLOGY ...... 129

6.2.1 Rainfall and Flood Data Collection and Analysis...... 129

6.2.2 Assessment of Urbanisation of Kumasi ...... 129

6.3 RESULTS...... 130

6.3.1 Rainfall Patterns and Trends in Kumasi ...... 130

6.3.2 Floods in Kumasi ...... 139

6.3.3 Landcover Change in Kumasi...... 146

6.4 DISCUSSION ...... 150

7 GENERAL DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS ...... 154 7.1 GENERAL DISCUSSION...... 154

7.2 CONCLUSIONS ...... 157

7.3 RECOMMENDATIONS ...... 159

8 REFERENCES ...... 162 9 APPENDICES ...... 179

LIST OF FIGURES Figure 1.1. Conceptual framework for the investigation of urbanisation, socio-economic activities and incidence of floods in Kumasi ...... 10 Figure 1.2. A composite model of the effects of human interference on a wetland ecosystem in an urban setting ...... 11 Figure 2.1. An illustration of the continuum between the terrestrial, wetland and aquatic systems ...... 15 Figure 2.2. Distribution of wetlands in the world ...... 20 Figure 2.3. Major vegetable farming areas in the Kumasi Metropolis ...... 35 Figure 3.1. Map of Ghana with an insert showing limits of the Kumasi Metropolitan Assembly and the study sites ...... 49 Figure 3.2. Population growth trajectory of the city of Kumasi ...... 50 Figure 3.3. The occupation of respondents from flood prone suburbs of Kumasi ...... 54 Figure 3.4. Residential types and homeownership of respondents in the different suburbs of Kumasi ...... 55 Figure 3.5. Respondents’ sources of land for construction of houses in the different suburbs of Kumasi ...... 57 Figure 3.6. Reasons for choice of wetland area for construction of houses in the flood prone suburbs of Kumasi ...... 58 Figure 3.7. Factors or instruments that will motivate residents of the various flood prone communities to relocate ...... 59 Figure 3.8. The predominant uses of wetlands in the various suburbs of Kumasi ...... 61 Figure 3.9. Perception of wetland protection in Kumasi ...... 61 Figure 3.10. Reasons for the continual destruction of wetlands in the various suburbs of Kumasi ...... 62 Figure 3.11. A proposed model to describe the decision to move or stay in the flood prone suburbs of Kumasi ...... 65 Figure 4.1. Assessment of stakeholders’ responsibility for the management of floods and wetlands in Kumasi ...... 74 Figure 4.2. Spatial distribution of houses marked for demolition and attitude towards demolition of houses in waterways by the Kumasi Metropolitan Assembly...... 77 Figure 4.3. A conceptual model for management of floods and wetlands within the Kumasi Metropolis ...... 83 Figure 5.1. A schematic diagram of a sample study site along a stream X showing the transects and quadrat layout ...... 87

Figure 5.2. Landcover and Landuse activities within 100 m from the stream channel at the different study sites in Kumasi ...... 93 Figure 5.3. Soil map of the Kumasi showing study sites ...... 95 Figure 5.4. A cluster diagram of the correlation matrix of the various environmental variables measured in the different suburbs of Kumasi ...... 100 Figure 5.5. A principal components analysis plot of the measured environmental parameters ...... 101 Figure 5.6. Heavy metals enrichment in the stream sediments of the different suburbs of Kumasi in the rainy season ...... 102 Figure 5.7. Heavy metals enrichment in the stream sediments of the different suburbs of Kumasi in the dry season ...... 103 Figure 5.8. Heavy metals accumulation in the stream sediments of the different suburbs of Kumasi in the rainy season ...... 104 Figure 5.9. Heavy metals accumulation in the stream sediments of the different suburbs of Kumasi in the dry season ...... 105 Figure 5.10. Seasonal differences in pollution by metals in the stream sediments of the different suburbs of Kumasi...... 106 Figure 5.11. Number of species identified at the various wetland sites in Kumasi ...... 116 Figure 5.12. Species richness, evenness and Shannon diversity at the study sites in Kumasi ...... 116 Figure 5.13. Species diversity indices with respect to distance from the water channel at the different sampling sites in Kumasi ...... 117 Figure 5.14. Analysis of similarities of the community composition among wetland sites in Kumasi according to (a) presence or absence of surface water and (b) distance from the water channel ...... 119 Figure 5.15. A hierarchical cluster of the community composition among wetland sites in Kumasi according to (a) presence or absence of surface water and (b) distance from the water channel ...... 121 Figure 5.16. A non-metric multidimensional scaling plot of the community composition among wetland sites in Kumasi using the presence or absence of surface water as a factor. 122 Figure 5.17. A non-metric multidimensional scaling plot of the wetland sites in Kumasi using the measured environmental parameters...... 122 Figure 5.18. A principal components analysis plot of the sampled wetland sites in Kumasi superimposed with vector plots (circle and lines) of the measured environmental parameters ...... 123 Figure 6.1. Average monthly rainfall pattern of Kumasi from 1961 to 2010 ...... 131 Figure 6.2. Historical annual rainfall pattern of Kumasi from 1961 to 2010 ...... 131 xii

Figure 6.3. Trends in rainfall of Kumasi for the months of January from 1960 to 2010...... 133 Figure 6.4. Trends in rainfall of Kumasi for the months of February from 1960 to 2010..... 133 Figure 6.5. Trends in rainfall of Kumasi for the months of March from 1960 to 2010...... 134 Figure 6.6. Trends in rainfall of Kumasi for the months of April from 1960 to 2010...... 134 Figure 6.7. Trends in rainfall of Kumasi for the months of May from 1960 to 2010...... 135 Figure 6.8. Trends in rainfall of Kumasi for the months of June from 1960 to 2010...... 135 Figure 6.9. Trends in rainfall of Kumasi for the months of July from 1960 to 2010 ...... 136 Figure 6.10. Trends in rainfall of Kumasi for the months of August from 1960 to 2010 ..... 136 Figure 6.11. Trends in rainfall of Kumasi for the months of September from 1960 to 2010 137 Figure 6.12. Trends in rainfall of Kumasi for the months of October from 1960 to 2010 .... 137 Figure 6.13. Trends in rainfall of Kumasi for the months of November from 1960 to 2010 138 Figure 6.14. Trends in rainfall of Kumasi for the months of December from 1960 to 2010. 138 Figure 6.15. Trends in total annual rainfall of Kumasi from 1960 to 2010 ...... 139 Figure 6.16. Respondents’ flood experience in the various suburbs of Kumasi...... 140 Figure 6.17. Relationship between different rainfall events and total monthly rainfall for Kumasi from 1961 to 2010 ...... 141 Figure 6.18. Trends in heavy rainfall events in Kumasi from 1961 to 2010 ...... 142 Figure 6.19. Recession time of water at the various flood prone suburbs of Kumasi ...... 144 Figure 6.20. Adaptations of residents to floods in the various suburbs of Kumasi ...... 144 Figure 6.21. Flood response model of new and old residents of flood prone suburbs of Kumasi ...... 145 Figure 6.22. Landcover changes within the jurisdiction of the Kumasi Metropolitan Assembly between 1986 and 2007...... 147

xiii

LIST OF TABLES Table 2.1. Likely effects of urbanisation on wetland hydrology and geomorphology ...... 18 Table 2.2. Anthropogenic effects on urban wetlands...... 29 Table 2.3. Likely effects of urbanisation on wetland ecology ...... 29 Table 3.1. Stakeholder list and the respective data collection tools used in the survey of wetlands in Kumasi, Ghana ...... 47 Table 4.1. Ghana’s international commitments to the environment ...... 68 Table 4.2. Stakeholder capacity assessment on wetlands and flood prevention and management in Kumasi...... 76 Table 5.1. Classes of the geoaccumulation index used to define sediment quality ...... 89 Table 5.2. The Braun-Blanquet scale and the description of the values used for sampling of wetland species in Kumasi ...... 90 Table 5.3. Edaphic characteristics of wetland sites sampled in the Kumasi Metropolis ...... 98 Table 5.4. Physicochemical characteristics of water running through the wetlands in Kumasi ...... 98 Table 5.5. Species identified in the different wetlands of Kumasi and their importance values ...... 109 Table 6.1. Descriptive statistics of monthly rainfall pattern of Kumasi from 1961 to 2010 . 132 Table 6.2. Rainfall events and reported floods in some suburbs of Kumasi in 2009 ...... 140

xiv

LIST OF PHOTOGRAPHS Photograph 1. Some uses of wetlands in the Kumasi Metropolis...... 97 Photograph 2. State of some of the wetlands in the Kumasi Metropolis ...... 108 Photograph 3. Some high importance value wetland species in the Kumasi Metropolis ...... 114 Photograph 4. More high importance value wetland species in the Kumasi Metropolis ...... 115 Photograph 5. House being built on stilts in a flood prone suburb in the Kumasi Metropolis ...... 148 Photograph 6. Raised walkways in the compound of a flood prone house in the Kumasi Metropolis ...... 148 Photograph 7. Embankment built around a flood prone house in the Kumasi Metropolis.... 149 Photograph 8. Embankment built around the main door to an apartment in a flood prone suburb of the Kumasi Metropolis ...... 149

APPENDICES Appendix 1. Survey instrument used for the wetland associated communities in Kumasi ... 179 Appendix 2. Survey instrument used for wetland associated NGOs and government agencies in Kumasi ...... 185 Appendix 3. Field data sheet used for wetland vegetation and ecological studies in Kumasi ...... 186 Appendix 4. Study sites and anthropogenic activities within 100 m from the water channel ...... 187 Appendix 5. A Wilcoxon Signed Ranks Test comparison of the seasonal differences in physicochemical parameters of the wetlands in Kumasi ...... 187 Appendix 6. Metal concentration (mg/kg) in sediments of the different study sites in Kumasi ...... 189 Appendix 7. Pearson correlation matrix of heavy metals and other measured environmental variables (significant relationships shown in bold) ...... 190 Appendix 8. Principal Component Analysis of environmental variables and heavy metals assessed in the wetlands of Kumasi ...... 191 Appendix 9. Comparison of means of heavy metal enrichment factors for rainy and dry season in Kumasi ...... 192 Appendix 10. Comparison of means of heavy metal geoaccumulation indices for rainy and dry season in Kumasi ...... 193 Appendix 11. Comparison of means of Pollution Load Indices for the different sites in the rainy and dry season in Kumasi ...... 194

xvi

LIST OF PUBLICATIONS FROM THIS THESIS The following manuscripts have been prepared with material from this thesis and submitted for publication: 1. Campion, B. B. & Venzke, J. -F. (2011). Spatial patterns and determinants of wetland vegetation distribution in the Kumasi Metropolis, Ghana. Wetlands Ecology and Management, 19(5): 423-431 2. Campion, B. B. & Venzke, J. - F. Rainfall variability, floods and adaptations of the urban poor to flooding in Kumasi, Ghana. This manuscript has been accepted for publication in Natural Hazards published by Springer. 3. Campion, B. B. & Venzke, J. -F. Beyond defining wetlands to feeling wet: the political ecology of wetlands in Kumasi, Ghana. This manuscript is under review by Wetlands Ecology and Management published by Springer.

In all these papers, I developed the concept of the study, collected data and wrote the manuscript with scientific advice from Venzke, J. –F.

xvii

ABSTRACT Urbanisation and climate change are two of the greatest environmental challenges the world is experiencing. Wetlands bear the heaviest effects of urbanisation in Kumasi, Ghana. Also, current forecasts show that climate change will very much affect sub - Saharan Africa and may cause flooding in many areas. Already, Kumasi has been battling with a growing incidence and scale of floods, mostly, at the suburbs closer to wetland areas. This research therefore sought to investigate the interlocked issues of urbanisation, climate change, wetland ecology and degradation and flooding in Kumasi. The vegetation, edaphic characteristics, hydrology and physicochemical conditions of the wetlands were analysed. Various stakeholders were interviewed on wetlands management, floods and adaptations to living with floods. Rainfall data from 1961 to 2010 was analysed for trends. The landuse change from vegetated to built environment was also analysed. Whilst there were some plant species common to all the wetlands, the species composition for each wetland was unique. In each wetland, the level of soil water saturation was found to influence species composition. The percentage organic carbon of the wetland soils was the strongest factor that explained the observed spatial heterogeneity of wetland vegetation in Kumasi. All the wetlands were found to be polluted by all the heavy metals investigated. Furthermore, the built-up and bare areas within the Kumasi Metropolitan Assembly (KMA) have been growing at a rate of 1.9 % per annum between 1986 and 2007. Conversely, the results indicated no significant change in rainfall of Kumasi for the flood prone months of June and July and, therefore, the floods were not due to climate change. Two types of floods were, however, identified in the suburbs studied: flash floods due to heavy rainfall events and “effluent stream floods” caused by a rise in the water table during the peak rainy seasons. Adaptations to floods by residents included dredging of streams, erection of embankments around houses or apartments, moving property to higher grounds, building networks of raised walkways and building houses on stilts. People continued to live in the flood prone areas because, other forms of capital developed after living in these suburbs for a while, far outweighed the push factors to relocate. Based on these findings, it was recommended that the KMA should use the presence of effluent floods and the typical wetland vegetation identified by this research as defining factors to delineate the wetland boundaries. Considering the limited financial resources available to the KMA for an engineering- biased urban drainage system, the approach to managing floods in Kumasi should be a combination of the reconstruction of the natural wetlands and community specific social interventions.

xviii

ZUSAMMENFASSUNG Verstädterung und Klimaänderungen stellen zwei der großen Herausforderungen für die globale Umwelt dar. Urbanisierungsflächen stehen demzufolge unter hohem Nutzungsdruck; dies gilt besonders für Feuchtgebiete. Eine hohe Verstädterungsrate, eine langsam vonstatten gehende Raumplanung sowie die Unfähigkeit und Ungleichheit in der Bereitstellung grundlegender sanitärer Anlagen führen in Ghana zur Bildung städtischer Ballungsräume mit unterschiedlicher Umweltgüte. Armenviertel (slums) und informelle Siedlungen entwickeln sich dabei meist in und um Überschwemmungsgebiete. Zudem zeigen aktuelle Vorhersagen, dass der Klimawandel insbesondere den subsaharischen Raum Afrikas u. a. mit zunehmenden Überschwemmungsereignissen betreffen wird. Bereits über einen langen Zeitraum mussten sich die Behörden in Kumasi mit einer zunehmenden Anzahl an Überschwemmungsereignissen in Armenviertel und informellen Siedlungen auseinandersetzen.

Die vorliegende Arbeit stellt die notwendigen Informationen bereit, die einen ganzheitlichen Lösungsansatz zu der komplexen Problematik aus Verstädterung, Verlust und Degradation von natürlichen Feuchtgebieten sowie Überschwemmungsereignissen in Kumasi bieten. Folgende Zielvorgaben stehen dabei im Fokus:

Beschreibung und Charakteristik städtischer Feuchtgebiete in Kumasi, Ghana,

Bestimmung von Überflutungsparameter sowie Bewältigungsstrategien gegenüber Überflutunge n in Kumasi,

Überprüfung der Anwendbarkeit der Umwelt-Kuznets-Kurve (Environmental Kuznets Curve [EKC]) als Instrument der Planung für überflutungsgefährdete Stadtteile in Kumasi.

Um diese Ziele zu erreichen, wurden zunächst sämtliche Feuchtgebiete des Großraumes Kumasi registriert. Neben Angaben zu den Vegetations-, Boden- und hydrologischen Verhältnissen wurden die physisch-chemischen Zustände der Feuchtgebiete analysiert. Die Informationen zu Feuchtgebietsmanagement, Überflutungen und Anpassungsstrategien an die Lebensbedingungen wurden durch Befragungen verschiedener Interessensvertreter gewonnen. Zudem wurden die Niederschlagsereignisse der vergangenen 50 Jahre

xix

ausgewertet. Darüber hinaus wurde auch der Effekt des Landnutzungswandels von vegetationsbedeckten zu bebauten Flächen untersucht.

Alle untersuchten Feuchtgebiete gehören zu einem Flusssystem. Während einige der Pflanzenarten in jedem Überflutungsbereich anzutreffen sind, besitzt jedes Feuchtgebiet trotzdem eine spezifische Artenzusammensetzung. In jedem Feuchtgebiet folgt die Verteilung der Vegetationsgesellschaften einem hydrologischen Gradienten; die Bodenwassersättigung ist der dominante Faktor für die entsprechenden Artenzusammensetzungen. Der prozentuale Anteil an organischem Kohlenstoff bildet eine weitere wesentliche Komponente zur räumlichen Heterogenität von Vegetationsgesellschaften in Feuchtgebieten. Während saisonale Unterschiede die chemischen Zustände des Wassers prägen (pH, DO), gibt es hingegen keine saisonalen Veränderungen im Vegetationsbestand der Feuchtgebiete. Die Artenvielfalt in den Feuchtgebieten wird nicht durch die benachbarte Landnutzung beeinflusst; sondern vielmehr durch die hydrologischen und edaphischen Verhältnisse innerhalb der Feuchtgebiete selbst. Aufgrund der Verhältnisse nahe der KNUST [Kwame Nkrumah University of Science and Technology]-Polizeistation werden hier die höchsten Raten an Kohlenstoffbindung erreicht. In allen untersuchten Feuchtgebieten kann eine Kontamination mit Schwermetallen festgestellt werden. Wahrscheinlich stammt der Anteil an Hg, Ni, Zn und Pb aus derselben Quelle, während die Immissionen von Cu und Cr andere Quellen haben.

Überflutungen in Kumasi sind nicht als Folge veränderter Niederschlagmuster in den Regenmonaten Juni und Juli zu verstehen. Es gibt zwei Typen von Überschwemmungen, die einen Einfluss auf die untersuchten Feuchtgebietsgesellschaften ausüben: plötzlich auftretende Überflutungen aufgrund starker Niederschlagsereignisse sowie Folgen eines steigenden Grundwasserspiegels, der zur Bildung von Flussläufen beiträgt. Die bebaute und vegetationsfreie Fläche innerhalb der Großstadtregion von Kumasi (Kumasi Metropolitan Assembly [KMA]) ist im Zeitraum von 1986 bis 2007 um 1,9 % pro Jahr angestiegen. Dieser Nutzungswandel führt aufgrund eines reduzierten Infiltrationsvermögen, kürzeren Ablaufzeiten und folglich höheren Abflussmaxima zu höherem Oberflächenabfluss. Dies ist die Hauptursache für plötzlich auftretende Überflutungen in Kumasi. Hingegen dauert die Überflutungszeit an der Oberfläche als Folge ansteigenden Grundwasserpegels während der Hauptregenzeit länger an.

Die Einwohner haben sich an die Überflutungssituation weitgehend angepasst. Die Anpassungsstrategien sind der Bau von Drainagegräben und der Bau von Dämmen um Gebäude, Verlagerung von Besitz in höher gelegene Bereiche, Aufbau eines höher liegenden Fußwegenetzes sowie Hausbau auf Pfählen. Wegen dieser Anpassungsmöglichkeiten ist die ortsansässige Bevölkerung nur bereit fortzuziehen, wenn alternative Wohnmöglichkeiten angeboten werden. Entsprechende Entscheidungen in den überflutungsgefährdeten Vororten von Kumasi folgen dabei nicht immer den Vorgaben der EKC. Auch bei steigendem Einkommen wandert die ortsansässige Bevölkerung nicht ab, auch nicht unter staatlichem Druck. Die anderen Formen der Lebenssicherung, die sich nach einer gewissen Aufenthaltszeit in diesen Vororten entwickelt haben, überwiegen die Gründe, die Feuchtgebiete wegen der Überflutungsgefahr zu verlassen.

Die Mehrheit der für diese Studie befragten Interessensvertreter ist der Auffassung, dass die Feuchtgebiete sehr wichtig sind und daher geschützt werden sollten. Die wichtigsten Verantwortlichen für das Feuchtgebiets- und Überschwemmungsmanagement sind die lokalen Clanchefs sowie die KMA. Zu den wichtigsten Aufgabenbereichen gehören die Verbesserung von Landnutzungsplänen, Abbruch flutgefährdeter Gebäude und die Errichtung öffentlicher Parks.

Aus den Ergebnissen geht hervor, dass die KMA das Auftreten von Grundwasserbedingten Überflutungsereignisse sowie von charakteristischer Feuchtgebietsvegetation, die in dieser Studie erfasst worden ist, als bestimmende Faktoren zur Abgrenzung von Feuchtgebieten herangezogen werden sollten. Diese Methodik wurde besonders in den Feuchtgebieten nahe der KNUST- Polizeistation, in den FRNR [Faculty Renewable Natural Resources]- Bauernhöfen, in Atonsu, Family Chapel, Dakwadwom, Spetinpom und Golden Tulip Hotel exemplarisch angewendet. Nach Festlegung der Grenzen der Überschwemmungsgebiete sollten zunächst alle dort vorhandenen Bauten zerstört werden. In den mit neuen Namen gekennzeichneten, sich mit Vegetation überziehenden Feuchtgebieten werden sich die natürlichen Ökosystemfunktionen regenerieren. Dazu sollte die KMA die Zusammenarbeit von Asantehene (König von Ashanti) anstreben, um mögliche Widerstände gegenüber den Planungsmaßnahmen einzuschränken.

Zukünftig sollten stadtökologische Fragestellungen in den Blickpunkt wissenschaftlicher Studien mehr berücksichtigt werden. Dazu gehören die mögliche Nutzung von Vegetation zur

xxi

Behandlung von städtischem Abwasser sowie die Bedeutung von Feuchtgebietsvegetation zum Hochwassermanagement und zur Kohlenstoffbindung. Zudem sollten diese wissenschaftlichen Untersuchungen die Grundlage für die Bewirtschaftungsmaßnahmen in abgegrenzten Feuchtgebietsregionen darstellen, um die Biodiversität zu erhalten sowie den Erholungswert dieser Räume zu steigern.

Um den direkten Eintrag von Schadstoffen in die Gewässer zu verringern, sollte die KMA die Unterbringung von Abfallsammelbecken in Feuchtgebieten einstellen. Wo es für notwendig erscheint, sollte eine Mauer um eine derartige Anlage eingerichtet werden, um Überläufe der Abwässer zu verhindern.

Die Feuchtgebiete um die KNUST-Polizeistation sollte aufgrund ihres ökologischen Wertes für die Erhaltung von Torfvorkommen geschützt werden. Dieser Standort könnte aufgrund seines Potenzials als „K ohlenstoffsenke“ im Rahmen des CO 2-Handels zur Abschwächung des Klimawandels als mögliche Einnahmequelle für die KMA fungieren. Beispielsweise böte die Errichtung einer Gartenstadt eine praktische Umsetzung des von den Vereinten Nationen verabschiedeten Gesetzes zur „Reduzierung von Emission aus Abholzung und Wald- Degradierung“ (REDD+). Dies sollte der zukünftige Weg von Kumasi sein, um den Forderungen der Konvention für globale Städtepartnerschaften für Biodiversität nachzukommen.

Aufgrund der begrenzten finanziellen Möglichkeiten der KMA zur Errichtung eines technisch adäquaten städtischen Drainagesystems zur Bewältigung des Überschwemmungsproblems wären eine Kombination aus Umgestaltungen der Feuchtgebiete, wie schon oben beschrieben, sowie sozialpolitische Maßnahmen notwendig. Ein derartiges Überschwemmungsmanagement sollte in jedem Vorort dazu dienen, Schwerpunkte bei der Reduzierung von Überschwemmungsflächen und -häufigkeiten zu setzen. In diesem Zusammenhang müssen KMA und die „National Disaster Management Organisation“ erkennen, dass die Einbindung der lokalen Clanchefs, Abgeordneter und relevanter NGO’s wichtig sind.

xxii

1 RESEARCH THEME AND THEORETICAL FRAMEWORK 1.1 INTRODUCTION Ghana became a signatory to the Ramsar Convention in 1988. However, more than two decades later, wetlands are still considered as ‘waste lands’, ‘flood zones’ or ‘mosquito breeding grounds’ and as such, dredged to facilitate drainage of the water, reclaime d for other uses, or simply designated as dumping sites for all types of refuse (Ministry of Lands and Forestry, 1999a). Even with various efforts to protect or improve the wetlands according to the dictates of the Convention, recent developments in human geography and socioeconomics continue to be the major threats to wetlands in Ghana.

Urbanisation is one of the greatest environmental changes the developing world is uszndergoing. The urban land area is therefore under severe pressure for various uses and wetlands bear the heaviest of them all (Ramsar Convention Secretariat, 2007). Even with technological progress in city infrastructure, a growing population puts pressure on biodiversity (Eppinka et al., 2004). Therefore, to safeguard and possibly enhance the livelihood of many urban and peri-urban communities who subsist on wetlands, it is imperative that the benefits of the natural wetland ecosystems, including their values for subsistence economies, are recognised when planning and implementing development projects (Silvius et al., 2000). This requires identification and delineation these urban wetland ‘ecosystem’ boundaries for their management and sustainability to reduce their conflicting and competing uses for housing, infrastructure and agriculture.

The problems that climate change, urbanisation and floods present are becoming impossible to differentiate. The scale, frequency and cost of urban floods are on the rise and are sometimes attributed to climate change without the requisite support from research. Global climate change impacts natural vegetation, affects ecological and physiological processes, alters growing season length, biomass production, competition and leads to shifts in species ranges and possible extinctions (Levy et al., 2004; Intergovernmental Panel on Climate Change [IPCC], 2007). Climate change mitigation has therefore been the motivation for efforts to further understand carbon budgets and carbon accounting at local scales to contribute to regional and global carbon budget information (Risbey et al., 2007; Tschakert et al., 2008). The need for information on carbon pathways in major urban areas in Ghana can

therefore not be overemphasised. Terrestrial ecosystems represent a large store of carbon, approximately, three times that of the atmosphere (Levy et al., 2004). Also, metre for metre, wetlands lock up more carbon than any other terrestrial ecosystems in active exchange with the atmosphere. For this reason, as interest in climate change mitigation grows, interest in wetlands intensifies in parallel (Ramsar Convention Secretariat, 2006; Maltby & Barker, 2009). With such growing interest, wetland ecosystems in Kumasi have the potential to serve effective ly as the basis for mitigating and adapting to climate change.

1.2 RESEARCH BACKGROUND AND JUSTIFICATION

1.2.1 Urban Environmental Issues Urbanisation is a threat to many natural habitats and species because species richness, evenness, and density on farms, lawns or gardens in urban areas are directly controlled by man (Niemelä, 1999). These human-dictated urban vegetation communities form the template for other functional groups of species (Taylor, 1993; Barrat-Segretain & Amoros, 1996). Urbanisation may also lead to increased patch fragmentation (natural, semi-natural or artificial) and changes in structure (biomass, density, stratification, etc.) and floristic composition and spatial diversity (Ehrenfeld, 2000; Altobelli et al., 2007; Grimm et al., 2008). Like most sub-Saharan countries, urbanisation has been a dominant demographic trend in Ghana. By the year 2020, urban population in Ghana will more than double that at 2000 whilst rural population would have increased by only 3 % (Ghana Statistical Service, 2002a). Given that urban landuse and footprint will continue to expand, the prognosis for maintaining diversity and function of biological communities and their associated ecosystem services within and near Ghanaian cities seems dire. Unlike the natural habitats, the biodiversity of urban habitats is poorly documented in Ghanaian cities, and thus baseline information is scarce and the possibilities of applying ecological knowledge in urban planning are limited (Niemelä, 1999).

Factors such as rates of population growth, urbanisation, inappropriate macroeconomic, social, and technological policies and choices have been considered as some of the causes and implications of habitat change in Africa (Kidane-Mariam, 2003; Satterthwaite, 2003). For example, in Ghana, almost all the non-agricultural economic activities take place in the urban areas. Therefore, peri-urban and rural landowners are usually excited about the potential opportunities that will be available to them when the urban economy and space merge with the peri-urban. This leads to an increase in demand for land for infrastructure, industry, 2

service and housing, thereby, causing a drastic landuse change in these transition zones. The land values rise dramatically and much income is accrued from the sale of lands as they are put to more productive and profitable uses. Due to the high economic incentives and developers’ need for lands at already populated areas, wetlands are also sold, although, at cheaper prices (Payne, 1997; Brook & Dávila, 2000). Urbanisation and economic expansion then outpace environmental controls and planning (Grimm et al., 2008). With the high rates of urbanisation, slow pace of urban spatial planning, inability and inequalities in the provision of basic sanitation services in Ghana, the cities differentiate into spatial clusters of different environmental quality. The worst zones are the clusters of informal communities or slums associated with wetlands (World Bank, 2007). The situation is further compounded by the deliberate filling and dumping of solid and liquid wastes into these wetlands and along waterways (Water Resources Commission, 2008). In most towns and cities, skip bins are peculiarly placed at wetlands leading to the gradual conversion of the area into rubbish dumps (Fobil & Atuguba, 2004).

The two major Ghanaian cities, Kumasi and Accra have been battling with a growing frequency and scale of floods. These floods are mostly associated with informal and slum communities in wetland areas. The nature and growth of these cities however make it difficult to identify the real cause of the floods. Whilst the multitude of physical factors described above cannot be overlooked, changing climatic conditions may also be a likely cause of floods.

Slum population in many Sub-Saharan cities accounts for more than 70 % of the urban population and are dependent on natural resources associated with the already pressured wetlands (Rietbergen et al., 2002). Poverty alleviation measures for these slum communities (e.g. industrialisation, agriculture, flood control and protection works etc.) are also drivers of ecosystem and environmental degradation (World Meteorological Organisation, 2006; Satterthwaite, 2003). These communities are also the most vulnerable, both physically and socially, to environmental hazards like the annual floods and tend to be the least influential economically and politically. Wetland degradation is therefore likely to continue to increase in slum areas because of the increasing incidence of urban poverty (Baharoglu & Kessides, 2000; Baker & Schuler, 2004).

The above vulnerability situation may be widespread in Ghana because the country has no urban policy. National responses to urbanisation have so far been very piecemeal and not sufficiently integrated and holistic. According to the United Nations Human Settlements Programme (UN-HABITAT), the challenges and priorities facing Ghana among others, are urban planning and management, urban environment and development and vulnerability reduction (UN-HABITAT, 2008a). Following these rising urban environmental problems, the Ministry of Local Government and Rural Development announced that it was in the process of developing a national urban policy (GNA, 2010b). This is supposed to functionally integrate social, economic, institutional, cultural, environmental and spatial aspects of urban development. However, the ecological justification of such a policy, if formulated, is likely to be weak because of the substantial dearth of information on urban ecosystems in Ghana.

1.2.2 Research Justification Global interest in wetlands, climate change and urbanisation has been on the increase in recent years. Notable recent efforts are the United Nations Conference on Environment and Development (UNCED) and its aftermath Conventions on Biodiversity and Climate Change, the International Peat Society Commission on Land use Planning and Environment, IUCN Wetlands Programme, Ramsar Convention, UN World Charter for Nature, UNESCO Man and the Biosphere Reserve Programme, UNESCO Convention for the Protection of World Cultural and Natural Heritage, World Wide Fund for Nature (WWF), the United Nations Environment Programme’s Cities as Sustainable Ecosystems initiative and the International Human Dimensions Programme on Global Environmental Change.

The Ramsar Convention Secretariat (2008) has called on all Parties to review the state of their urban and peri-urban wetlands and to put in place schemes for restoration and rehabilitation; formulate landuse planning and management to minimize any future impacts on urban wetlands; and become involved in the Convention on Biological Diversity’s Global Partnership on Cities and Biodiversity. Another output of the Conference is the call for preparation of guidelines for managing urban wetlands, in accordance with an ecosystem approach, taking into account issues such as climate change, ecosystem services and livelihoods. The following year, the UN chose ‘Planning our Urban Future’ as its theme for the 2009 World Habitat Day Celebration. The World Disasters Report 2010 also focused on urban risk impacts of changing precipitation patterns (International Federation of Red Cross and Red Crescent Societies [IFRC], 2010). 4

These international concerns are relevant to Ghana. Being a member of the international community and signatory to several treaties and conventions, the Government of Ghana recognises the importance of wetlands (Ministry of Lands and Forestry, 1999a). Ghana’s wetland commitments are incorporated into an extensive poverty reduction strategy (PRS) process that integrates environmental issues with development and poverty reduction (Griebenow & Kishore, 2009). Principal commitments in the PRS are the Millennium Development Goals (MDGs), the Kyoto Protocol, the New Partnership for African Development (NEPAD) and the Convention on Biological Diversity (National Development Planning Commission, 2005). The peer to peer review by NEPAD of the PRS process recommended that Ghana should ensure effective environmental sustainability in progr amme and policy designs of ministries, departments and agencies of government and improve their capacity to practice environmental sustainability (African Peer Review Mechanism Secretariat, 2005). These recommendations have not been implemented by the Kumasi Metropolita n Assembly (KMA) and all the other assemblie s.

In assessing Ghana’s efforts at attaining environmental sustainability (i.e. MDG 7), it is critical to note that, some of the most cost-intensive interventions have so far not been addressed. These include developing systems for environmental monitoring, landuse management, management of watersheds and freshwater ecosystems and biodiversity conservation among others (Sachs et al., 2004). Projections by the National Development Planning Commission (2005) and National Development Planning Commission & UNDP (2010) show that Ghana may not meet the targets of the MDG 7 and therefore set out three critical areas to be considered: institutional resources for monitoring and evaluation, regulation and enforcement and environmental management; resources (human and financial) development for education and training; and specific investment in institutional capacity and policy reforms. It also proposed the development of a landuse master plan that demarcates all lands in dispute; initiation of measures to stem land degradation; minimization of the impact of climate change/variability; and promotion of human centred biodiversity conservation initiatives.

Environmental research and information to meet these targets have been lacking, not only in Ghana but in developing countries as a whole (Sánchez-Rodríguez et al., 2005). With the 5

issue of climate change rapidly gaining the right political attention and momentum, scientists need to take advantage of this trend of events to identify possible local climate impacts. It is now known that climate change will very much affect sub - Saharan Africa (Niasse et al., 2004; Millennium Ecosystem Assessment, 2005; Intergovernmental Panel on Climate Change [IPCC], 2007; Satterthwaite, 2008), especially, its water resources (Reinelt et al., 1998; Choi et al., 2003; United Nations Department of Economic and Social Affairs, 2005; Jenerettea et al., 2006; Obuobie et al., 2006) and may cause flooding in many areas (Fedeski & Gwilliam, 2007; IPCC, 2007; International Recovery Platform, 2009; McClean, 2010). This information on sub-Saharan African countries is generated from running climate models. However, no studies have been conducted on specific local climate change effects and adaptations by the vulnerable urban wetland communities in Kumasi, Ghana.

Relevant research conducted in Kumasi and its environs have been on pollution and peri- urban agriculture (Brook & Davila, 2000; Keraita et al., 2002; Cornish & Kielen, 2003; Keraita et al., 2003; Cofie et al., 2005), livelihood changes in peri-urban environments (Holland et al., 1996; Danso et al., 2002; Quansah et al., 2005; Danso et al., 2006; Drechsel, 2006; McGregor et al., 2006), urban planning (Mabogunje, 1990), land and land tenure (Blake & Kasanga, 1997; Botchie et al., 2007). There is certainly a shortfall in depth and direction of research that would provide a basis for climate mitigation in the urban environment in Kumasi, Ghana. Also, the GPRS II does not provide room for GIS-based ecological and poverty maps and climate change scenarios to offer practical information and innovation required to better inform development policy. There is no information to use to develop instruments for the vulnerable to cope with the impacts of climate change by themselves. There is no information on the effects of climate change on urban wetland vegetation dynamics, wetland hydrology and carbon sequestration in the light of the increasing pollution, urbanisation, agriculture and for carbon budgeting. In order to understand the link between urbanisation and the incidence of flooding, there is the need to know the spatial extent and characteristics of wetlands, land cover changes and fragmentation as an information input for better spatial planning in Kumasi. This research is therefore an attempt to provide the needed information for a holistic solution to the interlocked issues of climate change, urbanisatio n, wetlands degradation and flooding in Kumasi.

1.3 RESEARCH GOAL This research therefore sought to contribute to knowledge on the state of wetlands (ecology, carbon stocks, and drivers of change), incidence of flooding and stakeholder needs for management of wetlands in Kumasi. It will also be an initial attempt to develop the information base for indicators of wetland biotic integrity in the Kumasi Metropolitan area as influenced by urbanisation.

1.4 RESEARCH OBJECTIVES To describe and characterise urban wetlands in Kumasi To identify the determinants of floods and coping strategies to flooding in Kumasi To test the applicability of the Environmental Kuznets Curve in the decision making of members of flood prone communities in Kumasi

1.5 RESEARCH QUESTIONS In order to achieve these objectives, this research answered the following questions: 1. What are the characteristics of wetlands in Kumasi and in which ways are they changing spatially and temporally? a. What are the physical and chemical characteristics of wetlands in Kumasi how are they related to the adjacent socio-economic activities? b. What are the ecological drivers of change in these urban wetland ecosystems? c. What are the wetland vegetation species and the spatial and temporal differences in distribution of these species in Kumasi? d. What is the carbon sequestration potential of these urban wetland ecosystems? 2. What are the effects of climate and landcover- landuse change on the incidence of floods in Kumasi? a. Is there any relationship between the incidence of floods and the climatic (precipitation) characteristics of Kumasi? b. What are the landuse-landcover changes in Kumasi and how do they affect the incidence of floods in Kumasi? 3. How does the economy of wetland communities influence their environmental concerns? a. What are the socio-economic activities associated with wetland communities in Kumasi?

b. Does the trend of wetland degradation in Kumasi follow the Environmental Kuznets Curve? i. How far to left of the continuum of the Environmental Kuznets Curve are members of these wetland communities in Kumasi? ii. What do these communities need to reach the peak or flatten the Curve? 4. How are wetland associated communities and authorities managing the associated wetlands and floods? a. What are the existing wetland management practices and adaptations to life by wetland associated communities to life in such environments? b. What are the opportunities for implementing the Ghana Wetlands Management Strategy in Kumasi?

1.6 CONCEPTUAL AND THEORETICAL FRAMEWORKS This research is based on the following conceptual framework represented in Figure 1.1. The total amount of wetlands in Kumasi is considered as a non-renewable stock. This wetland stock is being reduced through the influence of urbanisation from the conversion of wetland areas into built environment. Also, these wetland areas, by virtue of their nature, do not attract high prices and in most places are not even sold. They are therefore invaded by poor people and converted into informal or slum settlements and the remaining wetland areas are associated with intense economic activities of these wetland inhabitants. The dense built up environment and clearing of vegetation in the wetlands therefore lead to increased frequency and scale of flooding of these suburbs. Storm drains are then built as a flood mitigation strategy but these have not been successful in solving the problem of floods.

This research assumes that, there are two pathways available to the KMA for the management of the wetlands and flooding: (i) continue with the business-as-usual engineering solution or (ii) put in place some wetland management interventions based on activities and programmes that will be investigated in this study (research themes [RT]). The latter will lead to an improved stock and state of wetlands in Kumasi offering positive environmental services including the reduction in floods and or its effects.

The business-as-usual and engineering solution will result in a vicious cycle of environmental ills. This is because, with increasing urbanisation, the city will frequently have to expand the storm drains to cope with the increasing runoff. Meanwhile, degradation and

loss of the wetland areas continue. The consequence of this is an increase in the frequency and intensity of floods and loss of environmental quality.

The above conceptual framework is based on the assumption that, urban processes and natural wetland ecosystems can co-exist. This is because, the perception of wetlands as a limit to urban expansion and economic growth could be changed to a rather active r ole in reducing urban poverty, achieving higher living standards and increasing human development levels in a sustainable manner (Costantini & Monni, 2008). However, to achieve this, the very complex links between urbanisation, wetlands, floods and human activities have to be understood. The theoretical basis of this research, therefore, piggybacks an adapted model of McDonnell & Pickett (1990) and Ehrenfeld (2000) on urbanisation and ecological phenomena (Figure 1.2). This research envisages that, there are interactions between the features of urban areas (column A) and ecological phenomena (column B). Manifestations of these interactions in Kumasi are shown in column C. Ehrenfeld (2000) proposed some components and indicator variables for assessment of such urban wetlands (column D).

In this model, the urban environment is divided into four entities: 1. human physical constructs, 2. the natural areas, 3. humans with their accompanying socio-economic activities and 4. climate. The human dwellings, roads, factories, offices etc. make up the physical abiotic developments; the urban greens, parks, wetlands etc. form the natural areas; the human influence is characterised by various socio-economic activities such as trading, carpentry, agriculture etc.; and all these interact and are influenced by the local climate. These form the drivers of change (A) of the environment. Urbanisation provides an opportunity for all these factors to closely interact (B) to produce the current wetlands of Kumasi, C (results). These wetlands have reduced ecological value because of encroachment and activities of the informal and slum communities. The research themes (D) are derived from phenomena taking place in rows 1, 2, 3 and 4. That is, this research seeks to investigate how the changes in the phenomena represented by the rows 1, 2, 3 and 4 of column D lead to the manifes tatio ns in column C in Kumasi.

outcome decision current state decision outcome

Figure 1.1. Conceptual framework for the investigation of urbanisation, socio-economic activities and incidence of floods in Kumas i (RT= research themes or areas being investigated) 10

A B C D

Divers of Change Biotic and environmental Result Research effects of urbanisation

Floods, Slums 1 Structural features of Physical and chemical Urbanisation (spatial and urban areas environment temporal changes in physical structures, pollution) ? Current state of 2 Wetland areas Population and urban wetland Ecology (soil, vegetation, community effects Ecosystems ? hydrology, physicochemical characteristics) ? Reduced 3 Socio-economic factors Management and capital value and Socio-economic conditions apportionment function of (floods, wetland management, wetlands anthropogenic activities)

4 Climatic effects

Figure 1.2. A composite model of the effects of human interference on a wetland ecosystem in an urban setting (adapted from McDonnell & Pickett, 1990; Ehrenfeld, 2000)

1.7 HYPOTHESES The following hypotheses have been formulated to guide this study.

I. According to the Federal Interagency Committee for Wetlands Delineation (1989), Keddy (2000) and Maltby & Barker (2009), wetland vegetation composition is dependent on the hydrological characteristics of the site. That is, water level fluctuations and depth of inundation during the year are important in the development of plant associations (Reinelt et al., 1998; Sassa et al., 2010). In an urban setting, however, the vegetation is often manipulated for aesthetics and design, biological conservation, and for reduction in some forms of pollutants (Botkin & Beverage, 1997; Owen, 1999). The urban vegetation is therefore often a mix of biological deserts and areas that are rich in biodiversity. Therefore, the structure, floristic composition and spatial distribution of urban wetland vegetation may rather be the result of deliberate human interventions to satisfy human interests or adjacent anthropogenic economic activities (Altobelli et al., 2007). It can therefore be argued that, in urban wetland environments, the anthropogenic influences or interests rather than the hydrology determine the vegetation.

Ho: Wetland vegetation composition in the Kumasi Metropolis is not determined by the hydrological characteristics of the wetland.

II. Nature is the original source of all heavy metals. The presence or absence of a heavy metal in any environment therefore depends on the site characteristics. Recent developments in less developed countries have led to a wave of rapid urbanisation (UNFPA, 2007). Even though these urban areas currently cover less than 3 % of the total land area of the earth, they are places of intense economic activities (Millennium Ecosystem Assessment, 2000). The direct effect of urbanisation and the associated intense activities is pollution of ecosystems (Sánchez-Rodríguez et al., 2005). For this reason, while heavy metals may be ubiquitous in urban environments due to natural processes, the increased human activities due to urbanisation and population growth could make a significant contribution beyond their natural concentrations (Sekabira et al., 2010). Spatial variations in heavy metals in urban environments could therefore be a reflection of the local anthropogenic activities.

Ho: Concentration of heavy metals in the wetland sediments of Kumasi is not greater than the natural sediment concentration.

III. Flooding has claimed more lives than any other single natural catastrophe (IPCC, 2007). In the decade 1986-1995, flooding accounted for 31 % of the global economic losses from natural catastrophes and 55 % of the casualties (Petrov et al., 2005). Also, recent analyses by the IPCC (2007) show that sub-Saharan Africa will experience the worst effects of climate change. On the other hand, when urbanisation occurs, some or all of the functions and values of wetlands are affected by direct activities such as dredging, filling, draining or outlet modification, while others are affected by secondary activities, such as increased impervious surfaces, leading to increased quantity of inflow water (Mumphrey & Whalen, 1977; Reinelt & Horner, 1991; Brody et al., 2007). Roy (2009) therefore concludes that the causality and vulnerability of urban areas to floods are clearly linked to the very nature of the urbanisation process. The floods at the various suburbs may therefore be outcomes of the urbanisation process rather than climate change. From the above arguments, it is therefore hypothesized that:

Ho: The flooding in the suburbs of Kumasi is not due to climate change.

IV. Studies show that an inverted U-shaped relationship exists between environmental degradation and per capita income (Magnani, 2001; Dinda, 2004; Costantini & Monni, 2008; Wagner, 2008). Panayotou (1993) therefore adduces from this relationship that, economic growth could be a powerful way for improving environmental quality in developing countries. That is, efforts at improving the economy of developing countries will lead to improved environmental quality. However, will there be an automatic improvement in environmental quality as income levels increase or it requires conscious institutional and policy reforms? If such a relationship exists, does the current environmental quality in the flood prone suburbs of Kumasi correspond to the economic status of its residents? Are the wetland slum communities an indicator of a stage of economic development of Kumasi? If such an inverted U-shaped relationship is true, why are rich countries or economies still degrading their wetlands? It is therefore hypothesized that:

Ho: The relationship between income and wetland degradation in the various suburbs of Kumasi will not follow an inverted- U- shaped relations hip.

2 WETLANDS, URBANISATION AND FLOOD NEXUS 2.1 INTRODUCTION Extensive literature on wetlands, urban wetlands, tropical wetlands, urbanisation, urbanisation in developing countries, urban ecology, urban poverty, flooding, climate change and carbon budgets was thoroughly reviewed. From this review, research gaps pertaining to wetlands, urbanisation and floods in Kumasi were identified. Also, the appropriate research methods were developed to ensure that methods are standard scientific procedure and comparable to similar research conducted elsewhere. Literature used in this review were generally peer reviewed documents from the following databases and resources: CAB Direct, African Journals Online (AJOL), ELDIS, ScienceDirect, Elsevier, Kwame Nkrumah University of Science and Technology (KNUST) and University of Bremen library resources and Government of Ghana reports and policy documents. The search was conducted using wetlands, urban, flooding, urban planning, urban poverty, ecosystem change, climate change, urban ecology, Kumasi, Ghana etc. as search words.

The search results showed that on the global level, significant research has been published in these thematic areas. Much of these publications are, however, focused on developed countries. There is relatively very little research conducted in these thematic areas involving Kumasi, Ghana or sub-Saharan Africa. This review is, therefore, very much based on the specific literature obtained and others which may not be that specific to the study area or region, but relevant to the assessment of the research and knowledge gaps on wetlands, urbanisatio n, poverty, climate change and flooding in Kumasi, Ghana.

2.2 WETLANDS Man has had a very long association with wetlands at different places. The concept of wetlands therefore has different meanings or sentiments because of the distinctive relationships with different interest groups in different parts of the world (Maltby & Barker, 2009). The different meanings may be related to the culture or history of a people or socioeconomic uses to which it has been put etc. Man, in changing from the wanderer to a settler, has been associated with water bodies. This man-water interaction has been beneficial and hazardous to both ends of the continuum and the meanings of wetlands sometimes reflect these experiences. However, the term wetland is a recent vocabulary compared to man’s association with it (Keddy, 2000). Wetlands are usually associated with floods, swamps, wastelands, mosquito-infested mud holes etc. It is therefore worthy to note that some of the

earlier delineations of wetlands in the USA were for their possible conversion to damp sites (Federal Register, 1980).

According to Haslam (2003), ‘wetland’ is originally a North American term and is generally referred to, interchangeably, as bogs, fens, marsh, swamps, sloughs, morasses, moor, mires, flood meadows etc. in non-ecological literature on the other side of the Atlantic. Wetlands have a combination of aquatic and terrestrial conditions that produce complex ecosystems that do not fit completely into aquatic or terrestrial ecology (Figure 2.1). As transitional zones or ecotones, they have characteristics and species typical of the communities they divide (Mitsch & Gosselink, 1993). Therefore, the characteristics of wetlands are determined largely by hydrologic processes which may exhibit daily, seasonal or longer-term fluctuations, in relation to regiona l climate and geographic location of the site (Halls, 1997).

Figure 2.1. An illustration of the continuum between the terrestrial, wetland and aquatic systems (source: Mitsch & Gosselink, 1993)

From the illustration in Figure 2.1, wetlands are ecological areas which may be marshes, peatlands, floodplains, rivers and lakes, and coastal areas such as salt marshes, mangroves, and seagrass beds, but also coral reefs as well as human-made wetlands such as waste-water treatment ponds and reservoirs. Because of this broad spectrum of ecosystems and the long 15

historical list of interchangeable words used to describe wetlands in different locations by different cultures, the definition of wetlands has been of much controversy. For example, in the 2005 Millennium Ecosystem Assessment report, ‘wetlands’ and ‘inland waters’ are used interchangeably. Also, “wetlands and mangroves” and “lakes, rivers, wetlands, and shallow groundwater aquifers” are mentioned as habitat types (Millennium Ecosystem Assessment, 2005). It is not surprising that such a renowned publication by a team of experts will have problems defining wetlands. Most authors therefore prefer to mention the various wetland ecosystem types rather than attempt a definition.

One of the earliest formal definitions is that of Ramsar as “areas of marsh, fen, peatland or water, whether natural or artificial, permanent or temporary, with water that is static or flowing, fresh, brackish or salt, including areas of marine water the depth of which at low tide does not exceed six metres” (Ramsar Convention Secretariat, 2006). This definition embraces almost all land area. Large tracts of land are sometimes temporary flooded due to anthropogenic activities or changes to the landscape. Also, changes in rainfall patterns sometimes create large flood zones that may only be temporal. In places where demand for land is very high, e.g. urban areas, such a broad definition will only exacerbate landuse conflicts. A less “land grabbing” definition is that by US Army Corps of Engineers and the US Environmental Protection Agency which considered wetlands as “areas that are inundated or saturated by surface or ground water at a frequency and duration sufficient to support, and that under normal circumstances do support, a prevalence of vegetation typically adapted for life in saturated soil conditions” (Federal Interagency Committee for Wetlands Delineation, 1989). This definition does not include artificial wetlands because under normal circumstances, they may hitherto have been a different land type and therefore, not supported wetland vegetation. It is also worthy to note that this definition was formulated as a guideline for selecting places to be designated as landfill sites in the USA and therefore d id not include coastal zones as wetlands.

Keddy (2000) emphasizes on inundation and defines wetlands as product of a cause-effect relationship. According to Keddy, ‘a wetland is an ecosystem that arises when inundation by water produces soils dominated by anaerobic processes and forces the biota, particularly rooted plants to exhibit adaptations to tolerate flooding’. The use of ‘forces’ in the definition undermines the fact that wetlands are a natural part of the environment with plants and animals adapted to such an environment. Therefore, there is no forcing of biota to adapt. The description of wetlands forcing biota makes wetlands look like a recent artificial creation such

that the biota that are ‘caught up’ in such a new environment are forced to adapt. With these diverse definitions and shortfalls, some wetland interest groups (International Mire Conservation Group and International Peat Society with support from the Dutch Ministry of Foreign Affairs, IUCN, Alterra, Wetlands International and Environment Canada) commissioned a study that revised the definition as follows: “a wetland is an area (of marsh, fen, peatland or water, whether natural or artificial, permanent or temporal) that is inundated (up to a depth of 6 m at low tide) or saturated by water (surface/ground, static/flowing, fresh/brackish) at a frequency and for a duration sufficient to support a prevalence of vegetation typically adapted for life in saturated soil conditions” (Joosten & Clarke, 2002). This all-encompassing definition could be considered the most sufficient for delineation, conservation and management of wetlands. According to Maltby & Barker (2009), the definitions of wetlands communicate the importance of substrate, vegetation and water in the identification, development and persistence of wetlands. It is therefore often argued that a robust definition that can provide effective guidance towards wetlands identification and delineation must include these three elements. For the purpose of this study therefore, the definition of Joosten & Clarke (2002) will be adopted. The Government of Ghana however, uses the Ramsar definition.

2.2.1 Morphology and Hydrodynamics of Wetlands Wetland morphology is as diverse as the landscape in which it is situated. The source and state of the water, topography and geology all determine the morphology of the wetland. The prevailing climatic conditions, wetland morphology and hydrology determine the wetland ecosystem attributes: species composition of the plant community, primary productivity, organic deposition and flux and nutrient cycling (Halls, 1997; Ot’ahel’ová et al., 2007). In an urban setting, wetlands are further impacted by anthropogenic activities leading to modifications in the vegetation, hydrological and morphological conditions as shown in Table 2.1. Upstream activities affect the watershed hydrology through changes in the average runoff supply and changes in the frequency and severity of extreme events, including flooding and drought. However, these have been difficult to assess due to complexities in the processes at work as well as a non-uniform and, in many parts of the world, deteriorating monitoring network (Millennium Ecosystem Assessment, 2005). In Ghana, gauges are rare on rivers and data on runoff is therefore difficult to come by.

Table 2.1. Likely effects of urbanisation on wetland hydrology and geomorphology (Ehrenfeld, 2000) Hydrology: Decreased surface storage of stormwater results in increased surface runoff (Decreased surface water input to wetland)

Increased stormwater discharge relative to baseflow discharge results in increased erosive force within stream channels, which results in increased sediment inputs to recipient coastal systems

Changes occur in water quality (increased turbidity, increased nutrients, metals, organic pollutants, decreased O2, etc.)

Culverts, outfalls, etc., replace low-order streams; this results in more variable baseflow and low-flow conditions

Decreased groundwater recharge results in decreased groundwater flow, which reduces baseflow and may eliminate dry-season streamflow

Increased flood frequency and magnitude result in more scour of wetland surface, physical disturbance of vegetation

Increase in range of flow rates (low flows are diminished; high flows are augmented) may deprive wetlands of water during dry weather

Greater regulation of flows decreases magnitude of spring flush

Geomorphology: Decreased sinuosity of wetland/upland edge reduces amount of ecotone habitat

Decreased sinuosity of stream and river channels results in increased velocity of stream water discharge to receiving wetlands

Alterations in shape of slopes (e.g., convexity) affects water gathering or water- disseminating properties

Increased cross-sectional area of stream channels (due to erosional effects of increased flood peak flow) increases erosion along banks

2.2.2 Distribution, Values and Degradation of Wetlands The variations in the definitions of wetlands make it a near impossibility to k now the total area and distribution of the world’s wetlands (Barbier, 1994; Finlayson et al., 2000). It is, however, generally estimated to be about 6 % of the earth surface (Haslam, 2003; Ramsar Convention Secretariat, 2006; World Meteorological Organisation, 2006; Maltby & Barker

2009). Wetlands are very fragmented and represented in all parts of the world (Figure 2.2.) varying in flora and fauna with diverse ecological functions and anthropogenic uses.

Whilst the definition of wetlands is in controversy, its values and functions are undisputed. When properly measured, the total economic value of a wetland's ecological functions, services and resources may exceed the economic gains of converting the area to alternative uses (Phillips, 1998; Turner et al., 2000). Challenges, however, arise in applying economic tools such as cost benefit analysis to assess wetlands because of the non-availability of data, especially, in developing countries (Phillips, 1998). In particular, economic analysis of the environmental functions of tropical wetlands is underdeveloped. However, the physical/hydrological and environmental services provided by wetlands around the world have been well documented and can easily be assessed using various scientific methods. The hydrological services of wetlands include its prevention of storm damage, flood and water control (Barbier, 1994; Shutes et al., 1999; Haslam, 2003; World Meteorological Organisation, 2006), coastal protection and aquifer recharge (Zedler & Leach, 1998), waste management and improvement of water quality (Biney, 1990; Barbier, 1994; Green & Upton, 1994; Zedler & Leach, 1998; Shutes, 2001; Haslam, 2003; Ramsar Convention Secretariat, 2006; Cooper, 2009) and climate change mitigation (Joosten & Clarke, 2002; IPCC, 2007; Ramsar Convention Secretariat, 2007). Also, lower soil horizons of wetlands support sulphate reduction to sulphides and provide the environment for metal-sulphide precipitation by trapping heavy metals in sediments (Maltby & Barker, 2009).

Wetlands are habitats of enormous biological productivity and high conservation value, especially, for migratory birds. They typically have high species diversity and are habitats for various stages in the life cycle of species (Barbier, 1994; Zedler & Leach, 1998; Haslam, 2003; Botkin & Beverage, 1997). Apart from benefit to animals, the socioeconomic development of many cities can be traced to their association with wetlands through the provision of potable water (Biney 1990; Barbier, 1994; Millennium Ecosystem Assessment, 2005; Ramsar Convention Secretariat, 2006), expression of cultural values, capture fisheries, aquaculture and extraction of agricultural and horticultural products (Biney, 1990; Mitsch & Gosselink, 1993; Barbier, 1994; Millennium Ecosystem Assessment, 2005). The use of wetlands for recreation and tourism is being heavily exploited in England, the Northern Territory of Australia, and Kashmir, northern Germany (Zedler & Leach, 1998; Millennium Ecosystem Assessment, 2005).

Figure 2.2. Distribution of wetlands in the world (Millennium Ecosystem Assessment, 2005)

There are increasing anthropogenic demands for water and other natural resources and services provided by wetlands and these impact adversely on wetlands (Ramsar Convention Secretariat, 2007). The capacity of wetlands to continue to provide these resources and services is therefore declining. A major cause of this decline is the increasing abstraction of water, especially, for agriculture (Barbier, 1994; Joosten & Clarke, 2002; Millennium Ecosystem Assessment, 2005; Ramsar Convention Secretariat, 2007), port expansion, industrialisation and urbanisation (World Meteorological Organisation, 2006). Many surface and groundwater-dependent wetland systems and their catchments are already water-stressed, and demand for water is projected to continue to increase (Millennium Ecosystem Assessment, 2005). For example, access to freshwater is declining for 1-2 billion people worldwide and this, in turn, negatively affects food production, human health, and economic development, and has a potential to increase societal conflict (Ramsar Convention Secretariat, 2008).

Following the varied uses of wetlands described above and its consequent degradation, the allocation of wetland resources requires careful analysis of their values, in particular, their indirect use values of environmental regulatory functions. The failure to take these values into account is often a major factor behind decisions that lead to overexploitation or excessive degradation of wetland systems (Barbier, 1994). The Government of Ghana in celebrating the World Wetlands Day, 2010, reiterated its commitment to review policies or amend the relevant legislations that impede the protection of the country's wetlands to ensure long-term ecological viability (GNA, 2010a). It is unfortunate however, that, this review, if it ever happens, will not have a strong backing of empirical data, information and literature from the Ghanaian context.

2.3 URBANISATION

2.3.1 Concept and History of Urbanisation The term urban areas are settlements or localities defined as "urban" by national statistical agencies. The cut-off points for identifying urban areas or populations therefore vary from country to country (Millennium Ecosystem Assessment, 2005). The most common ways of classifying urban settlements are by population, economic conditions, and location. The minimum population density criterion commonly ranges between 400 and 1,000 persons/km2 (Qadeer, 2004; Pacione, 2009). When population size is used, other economic or social indices are described. The minimum population size ranges between 1,000 and 5,000 residents; and maximum agricultural employment is usually in the vicinity of 50–75 % 21

(UNFPA, 2007). In each case, however, cut-off points outside these ranges can easily be found. For example, in Sweden, the urban population threshold is 200 inhabitants; 10,000 in Benin, Switzerland and Burkina Faso; 20,000 in Nigeria and 30,000 in Japan (Pacione, 2009). A settlement with a minimum of 5,000 persons is used to describe urban settings in Ghana (Ghana Statistical Service, 2002a). Although country-specific definitions will remain central to defining and assessing urban centres for many years to come, the basis for a more uniform definition is emerging from work using remote sensing and geographical information systems (Fedeski & Gwilliam, 2007). According to Pacione (2009), whatever the reasons are for urbanisation, it is likely to be a thing of the past in the near future. This is because, with globalisation and advances in telecommunication technologies, time becomes instantaneous and the urban frame will be useless and spatial congregation becomes unnecessary. Man will return to more ‘rural’ and isolated settlement types but still perform all urban functions. Following these arguments, the definition of an urban setting is expected to become more controversial in the future.

Also following recent spatial development trends, urban and suburban landuse form a continuum as much urban land area is occupied by dense-to-moderate residential housing. Thus, there are no clear physical boundaries between urban, suburban lands and rural lands, or between the urban cores of the larger cities and suburbia (Ehrenfeld, 2000). However, to differentiate economically, urban residents are less likely to work in agriculture and more likely to work in industry or services. For example, in India, more than 75 % of adult male population in a town has to be engaged in non-agricultural work to be considered urban (Pacione, 2009).

As seen above, there is no international agreement on the defining characteristics of urban, nor are there any scientifically accepted criteria by which to identify urban areas and populations. Moreover, some urban researchers believe that the distinction between rural and urban is becoming less relevant (Cohen, 2004) and that the boundary definitions are inevitably somewhat arbitrary. Even though there are differences in what is urban, the concept is well studied among sociologists and geographers. Statistically, urbanisation reflects an increasing proportion of the population living in settlements defined as urban, primarily through net rural to urban migration. The level of urbanisation is the percentage of the total population living in towns and cities while the rate of urbanisation is the rate at which it grows (Mille nnium Ecosystem Assessment, 2005; UNFPA, 2007).

2.3.2 Modern Trends in Urbanisation In the olden days, man was an agricultural being and most of the daily activities involved procurement of food. With time, technological advances changed the face of agriculture (Kaplan et al., 2009; Pacione, 2009). Higher productivity in agriculture resulted in food surplus sufficient to feed non-farmers. This led to division of labour such that, other than farming, people could become full time craftsmen, traders, soldiers, religious leaders, politicians etc. Increased geographical mobility and diffusions of goods, technologies and ideas resulted in concentration of populations. Rather than the earlier focus on agriculture, trade hubs, political centres and massive public projects to improve human wellbeing started to become important (Jacobs et al., 1997; Bhattacharya, 2002; Millennium Ecosystem Assessment, 2005; Smith, 2005; Kaplan et al., 2009). The larger human settlements grew as cultural centres of literacy, arts, religion and science. This then led to establishment of political hubs of imperial power and later, economic domination of cities over rural areas (Harvey, 1985a, b). Then came the period between 1750 and 1950 when most regions in North America and Europe experienced industrialisation resulting in a wave of urbanisation to produce the urban industrial societies. Since then, urbanisation and urban growth have continued to be major demographic trends in all parts of the world (Millennium Ecosystem Assessment, 2005). The increased urbanisation resulted in increase from only two cities containing 10 million people in 1950 to seven by 1975 (World Bank, 1991).

Recent developments in less developed countries have led to a second wave of rapid urbanisation where the number of urbanites will grow from 309 million in 1950 to 3.9 billion in 2030 (UNFPA, 2007) with Africa being the most urbanising (Bhattacharya, 2002; Kaplan et al., 2009). According to the Millennium Ecosystem Assessment (2005), decolonisation and the development of independent nation states had heavy influences on urbanisation in Africa and much of Asia as controls on the rights of the inhabitants to live in or move to urban centres were dismantled, and in part because the building of the institutions for independent governments increased urban employment. The UNFPA (2007) projects the urban population of Africa and Asia to double between 2000 and 2030 and Grimm et al. (2008) estimates that more than 95 % of the net increase in the global population will be in cities of the developing world. Meanwhile, the urban population of the developed world is expected to grow relatively little: from 870 million to 1.01 billion within the same time frame. At the global level therefore, all future population growth will thus be in towns and cities. Already, in 2008 the proportion of the world’s human population residing in urban areas reached 50 % as a result of rapid urban growth in developing countries (United Nations, 2008; UN-HABITAT, 23

2008b). Urban places will therefore be of fundamental importance for the distribution of population; in the organisation of economic production, distribution and exchange; in the structuring of social reproduction and cultural life; and in the allocation and exercise of power (Pacione, 2009).

It is also argued that the main driver of current urbanisation is economic growth. In general, the most urbanised nations are those with the highest per capita incomes, and the nations with the largest increases in urbanisation are those with the largest economic growth (World Bank, 1991; Millennium Ecosystem Assessment, 2005). The current rates of urbanisation in developing countries could therefore have global economic implications. It can shape prospects for global economic growth, poverty alleviation, population stabilisation and environmental sustainability. Cities are already the locus of nearly all major economic, social, demographic and environmental transformations. They cover only about 2 % of the earth’s surface, yet consume 75 % of all resources and produce 75 % of all waste (UNFPA, 2007). Therefore, improving urban planning to make cities more sustainable and increase the quality of urban life will require integrated urban planning and management and cooperation among responsible stakeholders (Berkowitz et al., 2003; United Nations Department of Economic and Social Affairs, 2005). Such stakeholders, however, have different and often conflicting priorities or insufficient resources to carry out their mandate. This is particularly so in developing countries which do not have the financial and economic resources to meet the ever increasing challenges of their uncontrollable urban expansion.

The growth of cities throughout the world presents new challenges to meet societal needs, which are vulnerable to both climate and population changes. For example, urban water withdrawal often extends far beyond city boundaries, as municipalities seek to secure new water sources from the hinterlands (Jenerettea et al., 2006). Cities consume less than 10 % of the total water used globally (Millennium Ecosystem Assessment, 2005). However, it is the concentrated nature of that demand which poses a heavy burden on the limited local water supplies. This is both in terms of volume of demand and pollution from wastewater discharge. The high costs of collecting, treating and distributing clean drinking water, as well as the costs of wastewater collection and treatment, are a major burden on public budgets and are beyond the capacity of municipal or national governments in most developing countries like Ghana (Millennium Ecosystem Assessment, 2005; Grimm et al., 2008). Any effort at improving wetlands will therefore be a step in the right direction to reduce the immediate and future water burden of these growing cities.

2.3.3 Urbanisation Trends in sub-Saharan Africa Urban population growth in sub-Saharan Africa is close to 5 % per annum and is the highest growth rate in the world (Kaplan et al., 2009). Contrary to current urbanisation trends of developed countries, the urbanisation of Sub-Saharan African countries can best be described as low density urban sprawl. Urban sprawl, loosely defined as dispersed and inefficient urban growth (Hasse & Lathrop, 2003; Mundia & Aniya, 2006) has undesirable effects. According to Brueckner & Fansler (1983) and The World Bank (2000), the phenomenon of urban sprawl is a symptom of an economic system gone awry. This dramatic expansion of settlements at the expense of open spaces and natural resources has sparked intense interest and debate on urban sprawl (Mumphrey & Whalen, 1977; Brueckner & Fansler, 1983; Mitsch & Gosselink, 2000; Hasse & Lathrop, 2003).

Cities in Africa are not serving as engines of growth and structural transformation. Instead, they are part of the cause and a major symptom of the economic and social crises that have evolved the continent (The World Bank, 2000). The predominance of agricultural and mineral trade and the lack of manufacturing have made them more dependent on rural than on urban economic activities and more susceptible to changes in commodity market prices (Kidane- Mariam, 2003; Millennium Ecosystem Assessment, 2005). Furthermore, the gains in human well-being are not distributed evenly among individuals or social groups or among countries or the ecosystems of the world. Sub-Saharan African urbanisation is one mostly engineered by deprivation of rural areas (The World Bank, 2000; 2007). In fact, a vast majority of rural- urban migrants do so to escape the extreme deprivation of amenities. Also, according to Kaplan et al. (2009), the combination of agrarian economy and high family sizes cause rural landlessness and poverty making people seek opportunities in the city. There is therefore a net increase in city dwellings; a lot of which are uncompleted and without access roads, electricity, water and telephone facilities (The World Bank, 1991; Berkowitz et al., 2003; Millennium Ecosystem Assessment, 2005; The World Bank, 2007). In these urban centres, the gap between the advantaged and the disadvantaged is increasing (Baharoglu & Kessides, 2000; White et al., 2001; National Development Planning Commission, 2005; Ravallion et al., 2007a, b).

According to the records, 7.8 % of the population of Ghana lived in urban centres in 1921. By 1960, urban population had risen to 23.1 % and further to 32 % in 1984 and 43.8 % in 2000 (Ghana Statistical Service, 2002a) and estimated to exceed 50 % in 2008 (UN-HABITAT, 2008). The rapid urbanisation puts enormous pressure on the utilisation of natural resources

and the provision of infrastructure and amenities like water supply, sanitation, waste management, transportation, and housing. The adverse impact of rapid urbanisation has also exacerbated the incidence of unplanned changes in landuse in the larger cities, especially Accra and Kumasi (Brook & Dávila, 2000; The World Bank, 2007). Poor residents of these cities tend to concentrate in the outskirts, in non-functional government structures and land and along urban wetlands as informal settlements and slums.

2.4 URBANISATION AND THE ENVIRONMENT A significant literature exist on ecosystems within urban areas (The World Bank, 1991; Zedler & Leach, 1998; Shutes et al., 1999; Rees, 2003; Janssen et al., 2005; Ravallion et al., 2007a, b); ecology of urban areas, treating urban systems as ecosystems (Botkin & Beverage, 1997; Ehrenfeld, 2000; Keddy, 2000; Capello & Faggian, 2002; Millennium Ecosystem Assessment, 2005; Berkowitz et al., 2003; Janssen et al., 2005; Secretariat of the Convention on Biological Diversity, 2005; Ramsar Convention Secretariat, 2007), urban planning (Mabogunje, 1990; Davey, 1993; Pauleit & Duhme, 2000; Abbaspour & Gharagozlou, 2005; Obeng-Odoom, 2007), and impacts of urban areas on global environmental change e.g. heat island effect (Berkowitz et al., 2003; Pongrácz et al., 2006; 2010; Huang et al., 2009; Takebayashi & Moriyama, 2009), and pollution (Berkowitz et al., 2003; Hidalgo et al., 2008). The diversity and trends of these studies reflect the themes of interest and concern of the scientific community.

Urban areas have major influences on the environment in terms of concentration of consumption and waste discharge, concentrated industrialisation and intensive use of energy and resources. The relationship between urban economic growth and development and the ecological effects of this growth is often insidious and manifest when they are almost irrepressible. Cities have therefore been perceived as parasites on the environment and spitting out wastes to the detriment of “natural” ecosystems (UNU-IAS Urban Ecosystems Management Group, 2004). With the development and rise of the modern environmental movements in the 1960s and 1970s, it became fashionable to consider everything to do with cities as bad and everything to do with wilderness as good. Cities were considered polluted, dirty, artificial, and lack wildlife; therefore, urban environments are bad. While the wilderness is unpolluted and full of wildlife therefore, it is good (Botkin & Beverage, 1997). Even in recent times, Rees (2003) described urban areas as the human equivalent of the livestock feedlot: a spatially limited area characterized by a large population of humans living at a high density and supported by biophysical processes mostly occurring somewhere else.

Cities may not be an ecological disaster, because they generate almost 80 % of the global economic output (Pauleit & Duhme, 2000). According to the United Nations Department of Economic and Social Affairs (2005), cities with high population densities use substantially less land, energy and water per person than low-density cities or suburban or rural areas with large and widely spaced individual houses. Urban agglomerations tend to have larger populations and more likely to be the site of large facilities and higher-level administrative functions and create comparatively densely settled areas involved in the manufacturing and service industry (Berkowitz et al., 2003; Millennium Ecosystem Assessment, 2005; United Nations Department of Economic and Social Affairs, 2005). More land is thereby freed for nature. Urbanisation is therefore not in itself inherently bad for ecosystems.

In contrast, most rural areas are made up of large tracts of land that have been converted to agricultural monocultures and individual houses widely spaced out in the landscape that do not support wildlife. Dense urban settlements may therefore be less environmentally burdensome than rural and suburban sprawl (Brueckner & Fansler, 1983; Capello & Faggian, 2002; Millennium Ecosystem Assessment, 2005). Also, urban centres facilitate human access to and management of ecosystem services through, for example, the scale and proximity economies of piped water systems. Botkin & Beverage (1997) put it very well in their ‘Cities as Environments’ publication that, urban implosion has dual benefits: improving the lives of most people and helping to conserve rural land and wilderness. The environmental sustainability of a city should therefore be in terms of the future health of the environment within the city. That is, the environment in which the city’s inhabitants dwell is influenced , for better or worse, by the city’s own growth and change (William, 1990; Fedeski & Gwilliam, 2007).

Urban settlements are agglomerations of people and their activities. Human activities in urban environments therefore constitute a natural part of urban ecosystem dynamics. Collins et al. (2000) described a number of conditions that shaped the urban ecosystem. First, compared with rural ecosystems, urban vegetation ecosystems are highly patchy and the spatial patch structure is characterized by a high point-to-point variation and degree of isolation between patches. Second, disturbances such as fire and flooding are suppressed in urban areas, and human-induced disturbances are more prevalent. Third, because of the higher temperatures in urban systems, in temperate climates there are longer vegetation growth periods. Fourth, ecological successions are altered, suppressed or truncated in urban green areas. The uniqueness of urban vegetation ecosystems can therefore not be overemphasised. The urban spatial diversity and opportunities are often harnessed in landscaping and aesthetics. Therefore, as people increasingly

live in cities, the cities act as both human ecosystem habitats and drivers of local ecosystem change. The challenge therefore should be how to foster urban systems that contribute to human well- being and reduce ecosystem service burdens at all scales (Millennium Ecosystem Assessment, 2005).

Cities are increasingly being recognised as ecosystems by themselves (Botkin & Beverage, 1997; Ehrenfeld, 2000; Berkowitz et al., 2003). The ecosystem approach is used in the analysis and implementation of the objectives of the Convention on Biological Diversity (CBD). The principles amount to a strategy for the integrated or holistic management of urban resources through modern scientific adaptive management practices (Secretariat of the Convention on Biological Diversity, 2005). Following this Convention, the Global Partnership on Cities and Biodiversity was formed to support cities in the sustainable management of their biodiversity; to assist cities to implement practices that support national, regional and international strategies, plans, and agendas on biodiversity, and to learn from existing initiatives. The Curitiba Declaration on Cities and Biodiversity adopted in 2007 reaffirms the resolve of mayors of cities to improving the lives of urban residents to meet the objectives of the Convention on Biodiversity (Secretariat of the Convention on Biological Diversity, 2007). This has come about because, historically planners and decision-makers tended to neglect ecosystem services and relations between ecosystems and human well-being, at least until a local or international crisis has forced such concerns onto the policy agenda. In order to grow sustainable cities, we need a renewed emphasis on the positive aspects of urban environments. It is possible for those who are interested in wildlands and those who are interested in urban environments to both achieve their goals. This can be achieved by combining environmental scientific knowledge with spatial planning.

2.5 WETLANDS AND URBANISATION The spread of urbanisation leads to wetland ecosystem fragmentation and exploitation (Ramsar Convention Secretariat, 2007). Landuse changes and over-use of water leads to wetland loss and degradation in many parts of the world and at rates faster than other ecosystems (Millennium Ecosystem Assessment, 2005). This is jeopardising the current and future provision of their services for human well-being. As the wetland area increasingly gets urbanised, the environment and in particular, flora and fauna are affected (Tables 2.2 and 2.3). Looking at these pressures on wetlands, society cannot afford a hands-off approach to management. Rather, wetland managers must clearly state the goals of management towards maintaining functions of the wetlands for a sustainable urban environment.

Table 2.2. Anthropogenic effects on urban wetlands (Zedler & Leach, 1998) Component Problem Hydrology altered hydroperiod due to dams, flood control channels, excess runoff, decreased water quality due to contaminants and eutrophication; Habitat loss, fragmentation and degradation; Pests feral animals and increased predation on native animals; invasions of aggressive plants and negative impacts on native vegetation; Infrastructure Maintenance or dredging for shipping canals, expansion of marinas, filling for highway expansion, replacement or installation of utilities; Visitors pressure for recreation (e.g., bird-watching, canoeing, bike trails) and vandalism;

Table 2.3. Likely effects of urbanisation on wetland ecology (Ehrenfeld, 2000) Organism Effect Flora: large numbers of exotic species present; large and numerous sources for continuous re-invasion of exotics; large amounts of land with recently disturbed soils suitable for weedy, invasive species; depauperate species pool; restricted pool of pollinators and fruit dispersers; chemical changes and physical impediments to growth associated with the presence of trash; small remnant patches of habitat not connected to other natural vegetation; human-enhanced dispersal of some species; trampling along wetland edges and periodically unflooded areas; Fauna: species with small home ranges, high reproductive rates, high dispersal rates favoured; large predators virtually non-existent; “edge” species favoured over forest-interior species; absence of upland habitat adjacent to wetlands; absence of wetland/upland ecotones; human presence disruptive of normal behaviours.

Because of these disturbances, wetlands have taxonomic groups with the highest proportion of threatened species (Millennium Ecosystem Assessment, 2005). According to Williams (1990), tropical wetland vegetation species tend to have narrow ecological requirements, with habitats that are accordingly highly fragile. This is even worst considering the rate of urbanisation and limited capacities of environmental agencies to monitor and control pollution in the developing countries. The driving force behind the establishment of the Ramsar Convention was the continuing destruction of wetlands and the impact of this destruction on populations of waterbirds (Ramsar Convention Secretariat, 2007). Yet, more than four decades

later, the “degradation and loss of wetlands is continuing more rapidly than for other ecosystems” (Millennium Ecosystem Assessment, 2005). The declining condition of inland waters is putting the services derived from these ecosystems at risk. The increase in pollution and degradation of wetlands has reduced the capacity of these wetlands to filter and assimilate waste. It is established that inland water ecosystems are in worse condition overall than any other broad ecosystem type (Millennium Ecosystem Assessment, 2005; Ramsar Convention Secretariat, 2007).

Urban land has competing uses for man. The qualities that make a location good for a city may often also make it an important location for biological conservation. For example, river mouths, confluences, riparian areas, deltas are good sites for cities because of the access to transportation, leisure and water for all purposes. However, these locations are modified and polluted after settlement with the ecosystem cleaning services taking place mostly beyond urban boundaries. The water requirements of aquatic ecosystems adjacent the expanding human settlement and freshwater demands increase pressure on the ecosystem. This leads to changes in flow regime (Reinelt et al., 1998), transport of sediments and chemical pollutants, modification of habitat, and disruption of migration routes of aquatic biota (Millennium Ecosystem Assessment, 2005). Landuse change also affects evapotranspiration, infiltration rates, and runoff quantity and timing (Owen, 1999; Berkowitz et al., 2003; Fedeski & Gwilliam, 2007; Roy, 2009). Changes of particular importance for human well-being are the contrasting reductions in the overall quality and quantity of available runoff versus concentrated peaks of runoff causing flooding. Also, the urban rainwater with absorbed pollutants from the air falls on road surfaces, parking lots, vehicles, corrosive railings, brick, and copper roofs thereby resuspending pollutants in the runoff. Wetland plants attenuate the stormwater flow, bind sediment, and directly accumulate metals (Shutes et al., 1999; Shutes, 2001; Fritioff, 2005). Vegetation also influences urban environmental conditions and energy fluxes by selective reflection and absorption of solar radiation and may also influence air quality and human health (Small, 2001). Information on the abundance, quality and spatial distribution of urban vegetation is therefore very vital for designing systems for the provision of ecosystem services, park and natural area management and urban planning.

Urbanisation is a relatively permanent landuse change. In general, the first order hydrologic impacts associated with urbanisation result from increasing impervious area (Choi et al., 2003; Reinelt et al., 1998; Brody et al., 2007). The hydrologic changes are associated with changes from major classes of landcover and landuse e.g. from forest to urban system (as the

case is in Kumasi). This leads to decreased infiltration and an increase in runoff with the associated increase in pollution loads, decrease in temporal reliability of environmental systems and disturbance of wetland biota (Reinelt et al., 1998; Zedler & Leach, 1998). Reduced infiltration can also lead to longer lifespan of pools with stagnant water, thus providing increased breeding opportunities for mosquitoes (Millennium Ecosystem Assessment, 2005; Ramsar Convention Secretariat, 2007). Urban wetlands are also the destination for contaminants from roads, gutters, storm drains, and occasional wastewater spills.

Whilst the conservation of ecosystems services is undisputedly good for mankind, the urban dweller rarely plays a significant part in conserving it. There is continued lack of understanding and appreciation of the complex processes involved because, many of these services take place at some distance from the urban consumers (Millennium Ecosystem Assessment, 2005). Also, the effect of the urban dweller’s actions and inactions and the control and management of these ecosystems go beyond the city boundaries. It is unfortunate however, that, the people most adversely affected by the loss of ecosystem services tend to be the least influential economically and politically (such as the urban poor, future generations, and residents living far from where the decisions are being made) (Millennium Ecosystem Assessment, 2005).

2.5.1 Urban Wetlands Vegetation Plant species found in wetlands are considered as wetland vegetation. Because migration of oxygen in waterlogged soils and into root zones is slow and proceeds at an unacceptable or intolerable rate for most organisms, these areas are often populated by plants adapted to poor oxygen conditions in the root zone (Lyon, 1993; Haslam, 2003). The major factors believed to influence wetland biodiversity include hydrology, geomorphology, substratum, chemistry, adjacent terrestrial system, grazing, anthropogenic influences and individual plant tolerance or response to the stresses of flood and drought (Harper et al., 1995; Owen, 1999; Casanova & Brock, 2000; Fortney et al., 2004; Maltby & Barker, 2009; Miller & Fujii, 2010). Conversely, wetland species play a key role in the water quality, structure and functioning of stream ecosystems (Barrat-Segretain et al., 1998; Shutes, 2001; World Meteorological Organisation, 2006; Sassa et al., 2010). Streamside vegetation also controls the supplies of allochthonous and autochthonous organic matter, filters the nutrients and pollutants, helps stabilise banks and influences the water temperature and light through shading (Zedler & Leach, 1998; Shutes, 2001; Baarta et al., 2010).

Urban wetlands often have unusual physical features and conditions which are the result of human interventions. These include disturbed soil profiles e.g. through agricultural activities, mounds of soil from old ditch-digging or channel-widening activities, heterogeneous materials e.g. from waste dumping or stream deposition, hard concrete surfaces etc. (Reinelt et al., 1998; Ehrenfeld, 2000; Millennium Ecosystem Assessment, 2005). These forms of disturbances create spatial and temporal heterogeneity in wetland ecosystems and act as important mechanisms in maintaining species richness (Pickett et al., 1989; Chaneton & Facelli, 1991; Barrat-Segretain et al., 1998; Fortney et al., 2004; Ot’ahel’ová et al., 2007; Grimm et al., 2008). Biological diversity in an urban setting could therefore be a direct reflection of the spatial diversity provided by the urban landscape.

Apart from the edaphic and hydrologic conditions, landuse is also said to affect the vegetation. In a study on the watersheds of Wisconsin lakes, U.S.A., agricultural development explained more of the variation in macrophyte community species richness and abundance than did urban landuse and that agriculture and urban land uses differ as causal agents (Sassa et al., 2010). That is, urbanisation and agricultural developments create different perturbation gradients that appear to affect aquatic plant species differently. With these diverse ranges of influencing factors and consequences, attempts at wetland restoration have therefore been a great challenge for many wetland ecologists because of the different and sometimes combined and synergistic effects of causal agents of degradation. For example, Baarta et al. (2010) attempted to predict macrophyte development in response to restoration measures. Modelling the status quo showed that completely isolated water bodies are over grown with macrophytes resulting in an intensified terrestrialisation process. In contrast, complete reconnection to the main channel caused a loss of macrophyte abundance and species in most aquatic habitats.

2.5.2 Wetlands, Urbanisation and Agriculture Urban agriculture is defined as an enterprise located within or on the fringes of a town, a city or a metropolis, which grows or raises, processes and distributes a diversity of food and non-food products, (re-)using largely human and material resources, products and services found in and around that urban area, and in turn supplying human and material resources, products and services largely to that same urban area (Mougeot, 1999; FAO, 2007). According to Egziabher et al. (1994), urban farming is a basic urban function as ancient as cities themselves. Asia is leading the way in the sector, with highly complex and efficient systems for the production and marketing of agricultural products.

The idea of urban agriculture got some boost as city planners envisaged an idea of a sustainable city and tried to achieve it through spatial planning that included agriculture and wild areas. For example, Sir Ebenezer Howard (1850-1928) in “Garden Cities of Tomorrow” describes a utopian city in which man lives harmoniously with the rest of nature. His ‘park town’ or ‘garden city’ concepts describe a city that is surrounded and interconnected by greenbelts, forming a system of countryside and urban landscapes. His idea was to combat urban overcrowding by the establishment of garden cities of about 30,000 inhabitants, self-contained in terms of employment, food production, recreation etc. (Mumford, 1961). With such a concept, variety of habitats ranging from fairly natural ones to highly modified ones are therefore maintained (Niemelä, 1999). In developing countries’ literature however, urban agriculture or green areas are more of an evolving economic activity than a result of spatial planning (Armar-Klemesu & Maxwell, 1999; Egziabher et al., 2004; Faithful & Finlayson, 2005; Erenstein et al., 2006; Foeken, 2006; Foeken & Owuor, 2008; Magigi, 2008) and governments are introducing institutional and other policy changes that recognize or tolerate, manage, and promote the activity.

Urban agriculture may be practiced for food (Egziabher et al., 1994; Armar-Klemesu & Maxwell, 1999; Mougeot, 1999), as a livelihood and asset strategy (Armar-Klemesu & Maxwell, 1999; Erenstein et al., 2006; Foeken & Owuor, 2008), income (Foeken, 2006) and has the potential to significantly contribute to the city’s sanitation needs (Lydecker & Drechsel, 2010). It is also known that, an increasing proportion of the target group for poverty alleviation in developing countries is now found in urban areas (Adams et al., 2004; National Development Planning Commission, 2005; Ravallion et al., 2007a, b). Many of these urban poor find employment or improve the quality of their daily diet through agriculture in the open urban spaces. In times of economic crisis, urban agriculture sustains livelihood of most people and assists with resolving or coping with diverse development challenges.

Urban agriculture is spurred by a complex web of factors still to be understood, not the least of which are urban poverty and food insecurity (Mougeot, 1999; Adams et al., 2004; Ravallion et al., 2007b). Even though Pacione (2009) considers the urban populace to be less employed in agriculture, the 1988 census ranked urban agriculture as Dar es Salaam’s second largest employer, after small traders and labourers. Urban agriculture occupied 11 % of the population aged 10 or more (Egziabher et al., 1994). However, few papers from literature overtly condemn urban agriculture under any form. All near opposition papers are concerned with the issue of urban planning (Magigi, 2008), public health (Afrane et al., 2004; Matthys et

al., 2006; Obuobie et al., 2006; Klinkenberg et al., 2008) and the environment (Adam et al., 1999; Keraita et al., 2003; Faithful & Finlayson, 2005; Erenstein et al., 2006; Mireri et al., 2007).

According to Magigi (2008), urban agriculture is a big economic activity in Africa because of the unguided land development, poor policy and legislative enforcement, out-dated policy inherited from colonial period and inadequate community involvement in spatial landuse planning processes. This assertion may be true because urban land values are usually very high and such lands cannot be used for agriculture if they are not originally part of the urban planning scheme. In Ghana, urban landuse plans do not include allocation of land for agriculture. Therefore, lands used for urban agriculture belongs to either government institutions or private developers or are open areas designated as wetlands along streams (Adam et al., 1999; Brook & Dávila, 2000; Oboubie et al., 2006). Due to the hilly landscape of Kumasi, most streams run through inland valleys unsuitable for construction and are therefore prime sites for cultivation of vegetables, sugar cane and taro (Figure 2.3).

Historically, agriculture has been the main cause of wetland loss (Egziabher et al., 1994; FAO, 1998; Eppinka et al., 2004; Millennium Ecosystem Assessment, 2005; Ramsar Convention Secretariat, 2006). This is through direct pollution by fertilizers, water abstraction for irrigation, drainage of excessively waterlogged places or conversion to paddy farms, and mining of peatlands, to mention a few. Whilst wetland conversion to agriculture is a common phenomenon, the permanent wetland loss from agriculture to urban infrastructure or housing through the urbanisation process has not been given the needed attention (Mathias & Moyle, 1992; Thibault & Zippererb, 1994). These shifts are very much evident in peri-urban areas of most developing country cities. However, the livelihood of most peri-urban communities depends to a large extent on the productivity of the natural systems, in particular wetlands (Brook & Dávila, 2000; Keraita et al., 2003). Therefore, to safeguard and possibly enhance livelihoods of many urban and peri-urban communities who subsist on wetlands, it is imperative that the benefits of the natural wetland ecosystems be recognised when planning and implementing development projects (Silvius et al., 2000; Grimm et al., 2008).

Figure 2.3. Major vegetable farming areas in the Kumasi Metropolis (Oboubie et al., 2006)

2.5.3 Wetlands, Urbanisation and Floods The urban development landscape in most developing countries ranges from rich neighbourhoods with infrastructure to poor slum and or informal settlements. These poor city neighbourhoods are more prone to flooding and other natural disasters than wealthy neighbourhoods. These unfavourable initial conditions and disasters then serve as both drivers and consequences of development patterns (Millennium Ecosystem Assessment, 2005; Adelekan, 2010). Without adequate spatial planning, governance, regulation, and public spending, increasing numbers of the urban population are living in wetlands and other ecologically sensitive areas. This development pattern is making some suburbs of a large number of African cities vulnerable to natural disasters (Adelekan, 2010). It is predicted that future flood damages will depend greatly on settlement patterns, landuse decisions, quality of flood forecasting, warning and response systems (ActionAid International, 2006; Millennium Ecosystem Assessment, 2005; Adelekan, 2010).

Flooding is a natural phenomenon (ActionAid International, 2006). Very often in nature, watersheds get flooded. This natural phenomenon has however become problematic because man is increasingly drawing nearer to these natural systems. The watershed flood frequency

and scale has therefore been intensified by human activities. With improved life styles, industrialisation and urbanisation, watersheds are constantly being inundated with expensive capital goods and infrastructure. The costs of damage due to floods have therefore continually risen (United Nations Inter-Agency Secretariat of the International Strategy for Disaster Reduction, 2005; World Meteorological Organisation, 2006; McClean, 2010). Flood research has, therefore, intensified and different methods have been applied in urban flooding in different places with different price tags.

For example, in 1951, a flood along the main stem of the Kansas River was estimated to have affected mostly urban areas because about 90 % (approximately US $479,000,000) of the damage experienced occurred in urban areas. This great loss was used to justify the construction of a stair-step series of lakes in the Kansas River Basin to act as storm water collection sinks (Koilmorgen, 1953). Since then, the flood control plans for the Kansas River Basin alone have been so extended and elaborated that they bear a price tag of billions of dollars. More shattering than the rising costs are the rising number of acres of land that is to be flooded by building more and more dams. These flood control mechanisms contravene both the spirit and purpose of resource conservation. Another good intention with a wrong approach was the building of a series of levées and dredging of channels as a control method of the December 1955 flooding of Santa Cruz. This only resulted in extreme siltation and loss of control of the river. Another flood of Santa Cruz in January 1982 proved that the solution to flooding in urban areas needs a more integrated approach than artificial embankments (Griggs & Paris, 1982). In practice therefore, floods can never be eliminated entirely. However, the vulnerability to floods can be mitigated by integration of physical developments with behaviour and actions especially by residents in flood prone areas (United Nations Inter- Agency Secretariat of the International Strategy for Disaster Reduction, 2005; APFM, 2006; Egyir, 2008; Akwei, 2010; McClean, 2010).

The scale and frequency of floods has been on the increase in some suburbs of Kumasi, Ghana. For instance, residents of Aboabo and Atonsu all in the Aboabo Basin are frequent victims of floods (Egyir, 2008; Akwei, 2010). However, most of the wetland research conducted in Kumasi has been on vegetable farming and water quality in river basins of Kumasi (Keraita et al., 2003; Obuobie et al., 2006; Keraita et al., 2008a, b; Lydecker, & Drechsel, 2010). Egyir (2008) and Akwei (2010) worked on physical flood control and coping strategies to flooding in the Aboabo Basin. The watershed hydrology and causes of annual floods have not been determined by any research but public opinion attributes the floods to

obstructed water ways and climate change (Akwei, 2010; Ghana News Agency, 2010). How much and which of the annual floods is attributable to each of these factors is not known. According to Bhaskar & Singh (1988), Odemerho (1993), and the United Nations Inter- Agency Secretariat of the International Strategy for Disaster Reduction (2005), the problems of floods in third World cities are compounded by the rapid rate of urban expansion into urban fringe areas often unaccompanied by the development of storm drainage systems.

The effectiveness of wetlands for flood abatement may vary depending on the size of the wetland, type and condition of vegetation (Ot’ahel’ová et al., 2007), slope, and location of the wetland in the flood path and the saturation of wetland soils before flooding (Bhaskar & Singh, 1988). Urbanisation destroys these qualities of wetlands and aggravates flooding because large parts of the ground are covered with roofs, roads and pavements and building drains. As a result, even quite moderate storms produce high overland flows quickly filling up river channels (Griggs & Paris, 1982; United Nations Inter-Agency Secretariat of the International Strategy for Disaster Reduction, 2005; McClean, 2010). Therefore with increased climate and landscape variability, most floods are getting out of hand and cannot be predicted accurately.

Early flood control and protection approaches have been based on structural solutions like embankments, bypass channels, dams, reservoirs, storm drains etc. This approach resulted in changes in flow regimes, separation of rivers from flood plains and in most cases increased flood regimes (Koilmorgen, 1953; Griggs & Paris, 1982). There is now a shift from flood control to flood management (Integrated Flood Management) with the recognition that floods are a natural river phenomenon (World Meteorological Organisation, 2006; Adelekan, 2010). Flood management is more holistic and carried out through a public policy framework. This integrated flood management approach is, however, not developed in Ghana. Food management is reactionary through the National Disaster Management Organisation (NADMO) in collaboration with the District Assemblies. In most places, mitigation measures depend on resources available to the District Assembly and political expediency.

2.6 URBAN POVERTY AND WETLANDS

2.6.1 Urban Poverty in Africa The post-independence experience of African countries in environmental management has been characterized by a series of overlapping crises. Most countries are experiencing very rapid urbanisation, environmental degradation, and increasing poverty (Kidane-Mariam,

2003) and a shift or relocation of poverty to the urban areas (IPCC, 2007). These urban poor are usually easy to locate in the urban landscape. According to the UN-HABITAT (1996), lack of access to land by these poor, facilitates the development and expansion of slum and squatter communities within urban areas. This situation is not only typical of Africa. For example, at end of World War II, cities housed less than a third of the poor in the United States and this had risen to more than half by the end of the century and these poor people are spatially concentrated in extremely poor urban neighbourhoods (Kaplan et al., 2009). This means a relocation of population, poverty and environmental pressures to urban areas. The challenge now is for a concerted and synergised approach to dealing with this urban poverty- environmental nexus.

The link between poverty and environment is complex and goes in both directions. The poor depend heavily on a number of environmental services for their livelihoods and are very vulnerable to the loss or degradation of environmental conditions (Adams et al., 2004). At the same time, poverty and unequal access to resources often compel people to use available natural and environmental resources unsustainably. For these reasons, Griebenow & Kishore (2009) suggested that strategies to address poverty needed to address environmental problems as well, and conversely strategies to improve the environment needed to address poverty.

However, in these already financially depraved slum or informal settlements, the lack of appropriate institutional structures drives ecological and environmental trends towards the negative. The condition of the environment and human settlements in the urban areas of developing countries has therefore drawn the attention of the academic community, international development agencies, regional intergovernmental organisations, nations, and civic societies at various geographical scales and is well captured in Goal 7 of the MDGs. Almost all African countries have therefore embarked upon the preparation and implementation of broad and specific policies and strategies of environmental management towards meeting this goal and to eliminate urban slums and reduce poverty (African Peer Review Mechanism Secretariat, 2005; National Development Planning Commission, 2005). The complex linkages between poverty and environment are therefore supposed to be mainstreamed in each country’s poverty reduction strategy (Griebenow & Kishore, 2009).

2.6.2 Urban Poverty and the Environmental Kuznets Curve The cause-effect relationship between environmental degradation, poverty and disasters is complex and has been the subject of many analyses (Beckerman, 1992; Kidane-Mariam, 2003; Stern, 2004; Sánchez-Rodríguez et al., 2005; Constantini & Monni, 2008; International 38

Recovery Platform, 2009). It is widely believed that poverty and other economic conditions determine the priority placed on environmental issues (Killeen & Khan, 2002; Rietbergen et al., 2002; Esty et al., 2005; Gunter et al., 2005; Rahman, 2006). That poor people directly extract environmental resources thereby destroying habitats. However, many poor people live closely with natural resources and rely heavily on low-input production systems gives. They therefore have the motivation and advantage in the development of new environmentally friendly products and services that improves their lives (Rietbergen et al., 2002).

Ecologists wary of the effects of increasing wealth on the environment, arguing that, economic growth and conservation are incompatible goals (Czech, 2003). Economists, on the other hand, expect economic growth to be a cure for global environmental challenges; that wealthier countries have the luxury of investing more to conserve and protect ecosystems (Cropper & Griffiths, 1994; Adams et al., 2004), fundamental aspects of human well-being, and the quality of life (Mills & Waite, 2009). According to Costantini & Monni (2008) and the Secretariat of the Convention on Biological Diversity (2007) environmental resources can be sustainably harnessed to play an active role in reducing poverty, achieving higher living standards and increasing human development levels. The difficulty, however, is to define that optimum or ideal use of the environment which would be sustainable and at the same time, achieve the economic and developmental goals envisaged by policy makers to reduce poverty.

Environmental degradation tends to increase as structure of the economy changes from rural to urban or agricultural to industrial, but it starts to fall with another structural change from energy intensive industry to services and knowledge based technology-intensive industry (Munasinghe, 1999; Magnani, 2001; Panayotou, 2003; Dinda, 2004; Costantini & Monni, 2008; Wagner, 2008). At the latter stage environmental degradation starts to decrease because people tend to make preventive environmental expenditures, donations to environmental organisations or choose less environmentally damaging products. Thus, rich people value the clean environment and make conscious efforts to preserve it. This systematic relationship between income and change in environmental quality forms an inverted-U-shaped curve called the Environmental Kuznets Curve (EKC) (Dinda, 2004).

The turning points of EKCs vary for different pollutants, environmental indicators, variables used in the model, country, and the policy environment (Magnani, 2000, 2001; Stern, 2004). For most of the pollution indicators, the estimated turning point lies within the income range of 3,137 (Panayotou, 1993) to US$ 908,178 (Stern and Common, 2001). No studies have been published on EKC and wetlands to know the estimated turning point. The closest study is that 39

of McPherson & Nieswiadomy (2005) on threatened species with the turning point between 10,000 and US $15,000 (US $1995 purchasing power parity) per capita income.

Even though the EKC has shown that income–environmental degradation relationship exists, there is still no consensus on the trajectory (Panayotou, 2003; Barbier, 2004; Dinda, 2004; Wagner, 2008). For example, Munasinghe (1999) suggests that, developing countries could take alternative routes that do not reach high levels of environmental degradation by identifying and addressing the economic imperfections that might exacerbate growth-related environmental harm. That is, higher levels of development could be attained at lower environmental cost. Also, the relationship between income and environmental care depends on the distribution of benefits and costs of polluting activities (Magnani, 2001). The driving force for flattening of the EKC may therefore not depend on incomes but distribution of costs and benefits of pollution and the policy interventions that are put in place to protect the poor and weak from bearing the burden of pollution.

As human populations increase, the marginal per capita value of wetlands first increases because of increased wetland scarcity per capita. Thus a higher human population implies that there is less wetland per capita available and therefore, on the average, fewer but more precious wetlands. But the marginal value of wetlands increases with human development only to a point because wetland functions begin to be lost through the urbanisation process (Mitsch & Gosselink, 2000). A planned urbanisation programme and good policy environment might decrease encroachment on wetlands and other natural areas thereby securing environmental quality. Therefore Panayotou (1997) and Eppinka et al. (2004) are of the view that, rather than economic growth, a more secured property rights, better enforcement and effective environmental regulations will help ‘flatten’ the EKC. Property rights will reduce the externalities associated with the use of wetlands as environmental goods.

2.7 CLIMATE CHANGE, URBANISATION AND WETLANDS A major non- climate attribute affecting the ecosystem and its vulnerability to climate is the increasing urbanisation of the developing world. Close to half of the world’s population now lives in urban areas. Increasing urbanisation means that the impacts of climate change on human settlements, if they occur, will increasingly affect urban populations (IPCC, 2007; Adelekan, 2010). A significant part of the activities leading to the achievement of Goal 7 of the MDGs (environmental sustainability) will therefore be played out in the urban environment (Secretariat of the Convention on Biological Diversity, 2007). The hazards to 40

which some cities are exposed are likely to become either more severe or more frequent, and so the increased risk should be anticipated, if not predicted (F edeski & Gwilliam, 2007; IPCC, 2007). According to Niasse et al. (2004), the physical and socio-economic characteristics of West Africa predispose it to be among the most vulnerable regions to climate change worldwide.

A changing climate can modify all elements of the water cycle: change both the frequency and intensity of precipitation, evapotranspiration, soil moisture, groundwater recharge, snowmelt and therefore runoff (Fedeski & Gwilliam, 2007). It is therefore believed that the most widespread potential impact of climate change on human settlements is flooding (Fedeski & Gwilliam, 2007; IPCC, 2007; International Recovery Platform, 2009; McClean, 2010). Riverine and coastal settlements are believed to be particularly at risk, but urban flooding could be a problem anywhere storm drains, water supply, and waste management systems are not designed with enough system capacity or sophistication to avoid being overwhelmed (IPCC, 2007). Also, there is likely to be increased cost of losses due to flooding because of the abundance of more precious and capital goods and services in urban areas (McClean, 2010). Urbanisation itself explains much of the increase in runoff relative to precipitation in settled areas and contributes to flood situations (IPCC, 2007). It is estimated that, by the year 2050, more than 70 % (and 90 % by the 2080s) of people in settlements that potentially would be flooded are likely to be located in a few regions: west Africa, east Africa, the southern Mediterranean, south Asia, and southeast Asia (Niasse et al., 2004; Mille nnium Ecosystem Assessment, 2005; IPCC, 2007).

Due to climate change and urbanisation, large and growing numbers of people are at high risk of adverse ecosystem changes, a worsening trend of human suffering and economic losses from natural disasters (Nelson et al., 2009). There is also increasing demand for ecosystem services because of increasing consumption per person, multiplied by a growing human population (Millennium Ecosystem Assessment, 2005). Effects of climate change will be felt most directly through changes in the distribution and availability of water, thereby increasing pressures on the health of wetlands. Restoring wetlands and maintaining hydrological cycles is of utmost importance for addressing climate change, flood mitigation, water supply, food provision and biodiversity conservation (Ramsar Convention Secretariat, 2007; IPCC, 2007). According to the Ramsar Convention Secretariat (2008), the management of these wetlands should be supported by indigenous and traditional knowledge, recognition of cultural

identities associated with wetlands, stewardship promoted by economic incentives, and diversification of the support base for livelihoods.

Wetlands are often biodiversity hotspots and deliver a wide range of ecosystem services relating to climate change mitigation (Berkowitz et al., 2003; Janssen et al., 2005; Nelson et al., 2009). Many types of wetlands play important roles in sequestering and storing carbon (Joosten & Clarke, 2002). However, this is not yet fully recognized in national and international climate change response strategies, processes, and mechanisms (Ramsar Convention Secretariat, 2007). Because of the high amounts of carbon stored, wetlands are particularly vulnerable to climate change impacts and human disturbances of wetland systems can cause huge emissions of stored carbon. Therefore, recognising the potentially serious implications of climate change on wetlands, Contracting Parties to the Ramsar Convention have been called to manage their wetlands in such a way as to increase their resilience to climate change and extreme climatic events and to ensure that climate change responses such as revegetation, forest management, afforestation and reforestation and their implementation do not lead to serious damage to the ecological character of wetlands (Ramsar Convention Secretariat, 2007).

Despite these emerging international developments on climate change and urbanisation, cities have the power to establish local environmental policies and regulations, such as landscaping guidelines and pollution/emissions standards to mitigate the effects of climate change. They can also integrate biodiversity considerations into other local policies, landuse planning, and the development and operation of infrastructure, such as transportation and water management systems (Secretariat of the Convention on Biological Diversity, 2007). Even with these opportunities, urban areas are still understudied in the analysis of global environmental change, with a majority of research placing emphasis on the contributions of emissions of greenhouse gases and the heat island effect to global climate change (Hidalgo et al., 2008; Takebayashi & Moriyama, 2009). Much less attention has been devoted to the study of the impacts of global environmental change on urban areas and the people who live in them. Particularly critical are the conditions in poor countries (Sánchez-Rodríguez et al., 2005).

2.8 REMOTE SENSING AND URBAN LANDSCAPE PLANNING AND MANAGEMENT The effects of the gradual global dominance of urban settlements extend well beyond the urban ecosystem. Therefore, as natural areas shrink and fragment through the urbanisation 42

process, humanity needs to conserve biological diversity and ecological integrity of urban environments for a sustainable future. Meeting the challenge of conserving regional ecological integrity in urban and urbanizing landscapes will depend on effective urban management and spatial planning. However, the fragmented nature, isolated and frequently disturbed conditions of natural areas exacerbate the difficulties inherent in spatial planning and urban landuse (Mazzotti & Morgenstern, 1997). This is because the distribution of land, water and forest resources on the earth’s surface is rarely equitable. Therefore the processes of human ecosystem differentiation, resource allocation and settlement have frequently had spatial dimensions that are usually characterised by patterns of territoriality and spatial heterogeneity on many scales (Berkowitz et al., 2003). In managing these spatially heterogeneous and scarce resources, there is the need for frequent inventory for monitoring and evaluation to inform policy and reduce conflicts. Large-scale ground-based monitoring of city space, forests, agricultural lands, wetlands, etc. are however, time-consuming, labour-intensive and are typically constrained by poor accessibility (Lyon, 1993; Smith, 1993; Vincent & Saatchi, 1999).

In recent times, urban landcover provided by satellite and airborne sensors is an impo rtant complement to in situ measurements of physical, environmental and socioeconomic variables (Small, 2001; Small & Lu, 2006). Satellite- based earth observation offers the possibility to provide synoptic measurements of land cover over very large geographic areas and over long periods of time (Brinson, 1993; Lyon, 1993; Xu et al., 2000; Harvey et al., 2001; Ozesmi & Bauer, 2002; DeKeyser et al., 2003; Mundia & Aniya, 2006; Small & Lu, 2006; Islam et al., 2008). The development and application of these technologies has been progressing steadily. Remote sensing and geographic information system (GIS) technologies have been applied in wetland studies (Thenkabail & Nolte, 1996; Lunetta et al., 1999; Töyrä et al., 1999; Thenkabail et al., 2000; Harvey & Hill, 2001; Ozesmi & Bauer, 2002; Choi, 2003), urban areas (Smith, 1993; Xu et al., 2000; Pongrácz et al., 2006; Altobelli et al., 2007), and spatial planning (Pauleit & Duhme, 2000; Abbaspour & Gharagozlou, 2005).

The application of remotely sensed observations to studies of the urban wetland vegetation has, however, been rather limited. This is because data of specific spectral and spatial resolutions for the discrimination of vegetation types are expensive and hard to come by. For these reasons, very few remote sensing studies have been carried out in tropical wetland vegetation in Africa specifically Ghana (Thenkabail & Nolte, 1996; Thenkabail et al., 2000; Mundia & Aniya, 2006). Whilst Landsat data may have its shortcomings for studies of the

built urban environment, it may be sufficient to detect significant spatial and temporal variations in urban vegetation (Small, 2001). This research therefore forms one of the early studies on urban wetland vegetation changes in Kumasi, Ghana through the application of remote sensing technologies.

Up-to-date landuse inventory is mandatory when making plans for solving problems associated with the rapidly growing areas to make them more sustainable and increase the quality of urban life (United Nations Department of Economic and Social Affairs, 2005). To meet these urban spatial planning needs, the Buffer Zone Policy of Ghana advocates for a buffer of all natural landscapes and waterways in urban areas and integration of natural areas into urban planning (Water Resources Commission, 2008). This policy is belated and most of the buffer areas have already been encroached upon. To curb encroachments on natural areas, the participatory approach as has been applied in city planning in Brazil could be used in Ghana (United Nations Department of Economic and Social Affairs, 2005). In doing this however, it is recommended that stakeholders give full recognition to the significance of wetlands in spatial planning so that the values they represent can properly inform landuse and investment priority-setting and the adoption of necessary safeguards (The Secretariat of the Ramsar Convention, 2007). The decentralised administration in Ghana which devolved spatial planning to the district assemblies offers an opportunity to involve all stakeholders in the planning process.

The establishment of KUMINFO in the KNUST and the Remote Sensing Applications Unit (now called CERGIS) in the University of Ghana were some of Ghana’s earliest formal attempts to propagate remote sensing and GIS applications. Also, Ghana's development agenda, driven by the Growth and Poverty Reduction Strategy (GPRS II) which is informed by our commitments to Millennium Development Goals (MDGs), New Partnership for African Development (NEPAD) and above all, the underlying obligations set out in the Constitution of the Republic of Ghana, recognises the planning and management of resources for development. However, the land tenure system and the system of land delivery in Ghanaian cities have significant implications for spatial development and urban sprawl (UN-HABITAT, 2008a). This research will therefore provide the ecological knowledge to supplement such efforts at improving the urban environment in Kumasi, Ghana.

3 POLITICAL ECOLOGY OF WETLANDS IN KUMASI 3.1 INTRODUCTION In all the regions of the world where the absolute number of poor has increased, a majority of them are in urban areas (Baharoglu & Kessides, 2000; Devas, 2005). In Ghana, the data on poverty shows that, urban poverty as a proportion of total poverty is also increasing (White et al., 2001). Using the expenditure approach, the ratio of urban to rural poverty in the Ghana 1991/92 survey years was 1:35 (Ravallion et al., 2007a). Urban poverty to some extent reflects active rural–urban migration due to the uneven distribution of facilities and opportunities in developing countries like Ghana (Baharoglu & Kessides, 2000).

Apart from government acquired lands, land is privately owned in Ghana. The official city layouts are therefore prepared is consultation with landowners. In such layouts, wetlands and natural areas or parks are delineated and cannot be built upon. Most of these layouts are not serviced, such that, the distance to amenities influences the rent on the land and plays the initial role of spatial resource allocator since it influences location choices of potential land buyers (Capello & Faggian, 2002). This leads to residential areas reflecting spatial differences in wealth of residents, types of houses and availability and density of infrastructure. The poor immigrants tend to settle in the areas that are not part of the formal layout of the city and erect temporal structures.

A poverty profile of the city will therefore distinguish these poor from the non-poor, identify their characteristics, facilitate the process of identifying the location of the poor and highlight groups that may be particularly vulnerable (Baharoglu & Kessides, 2000; Baker & Schuler, 2004). Data is, however, often difficult to obtain in developing countries for poverty profiling. Whilst data is not easily available, the physical location and characteristics of the poor are undeniably obvious in Ghanaian urban housing landscape. The wetlands of Kumasi are prime areas for these informal settlements.

This chapter describes the physical and socio-economic characteristics of Kumasi, specifically, inhabitants of these informal settlements along the wetlands areas. The political ecology in terms of land acquisition and demolitions will be discussed. The aim is to identify plausible development pathways and the choices and actions for (re)shaping these informal settlements for a sustainable future.

3.2 METHODOLOGY

3.2.1 Physical and Socio-economic Survey of Wetland Communities of Kumasi This involved the use of primary and secondary data. The secondary data are mainly from published works on Kumasi and its environs. The primary data were collected during field visits through the use of surveys, interviews, observations and photographs at the different study suburbs. All governmental and non-governmental organisations that are relevant to this research or work directly or indirectly with the wetlands were identified. Table 3.1 is a list of stakeholder groups identified and employed in this research. These included: the Kumasi Metropolitan Assembly (Environment and Sanitation Unit), Friends of Rivers and Water Bodies, sub-chiefs, Town and Country Planning Department (TCPD), Hydrological Services Department (HSD) and the National Disaster Management Organisation (NADMO). Informal visits were first made to the identified organisations and the researcher and research ideas introduced and relevant personnel suitable for the research identified. Official introductory letters were then sent to these identified personnel and where necessary, the heads of departments to introduce the research. The specific involvement of the organisations or personnel in this research was specified and sought. From the list of stakeholders, various data collection strategies as shown in the table were employed to solicit the needed information.

3.2.1.1 Wetland communities and the associated anthropogenic activities

With the aid of maps, satellite imagery and the reconnaissance survey, houses, social and economic activities within 100 m from the stream channels based on the 100 m buffer range set by the Ministry of Lands and Natural Resources (Water Resources Commission, 2008) were selected. From this, households were randomly selected and a structured questionnaire administered to one adult member of the household (Appendix 1). Afterwards, respondents who had interest and knowledge about the area were selected for a focus group discussion to probe further, summarise and draw conclusions from the information collected from the questionnaire survey and interviews. In all, 128 respondents answered the structured questionnaires.

All anthropogenic activities within the 100 m buffer for each wetland were identified and recorded. Stakeholders whose activities have direct impact on the wetland e.g. modifications in water the channel, water quality, wetland vegetation etc. were identified and interviewed.

Table 3.1. Stakeholder list and the respective data collection tools used in the survey of wetlands in Kumasi, Ghana Data collection Stakeholder group Stakeholder units # of respondents method

Wetland associated Dakwadwom 20 Structured communitie s questionnaire and Ahensan 21 focus group Aboabo 21 discussions

Kwadaso Estates 20

Atonsu 20

Spetinpom 26

Wetland Associated Agriculture 5 Interviews Businesses Car washing bays 5

Carpenters 5

Car fitting shops 5

Churches 5

Traditional Traditional Elder of 1 Interviews Authorities Dakwadwom

Traditional Elder of 1 Ahensan

Traditional Elder of 1 Aboabo

Governmental and KMA, HSD, TCPD, 5 Interviews non-Governmental NADMO, Friends of organisations Rivers & Waterbodies

3.2.1.2 Traditional leaders, governmental and non-governmental agencies Each suburb of Kumasi usually has a chief or traditional leader/elder that oversees to traditional issues in his/her jurisdiction. The traditional leaders of Aboabo, Dakwadwom and Ahensan were identified and interviewed on traditional wetland management, land tenure, landuse planning and future of the wetland. Interview sessions were also held with the listed governmental and non-governmental organisations. Information on research, laws and policies, and activities on wetlands in Kumasi was sought.

3.2.2 Data Analysis Various univariate analyses were used to assess the socio-economic forces that drive wetland ecosystem and to indicate the degree to which socioeconomic, demographic and environmental variables contribute to wetland alteration in Kumasi. The results are presented in charts, graphs and tables.

3.3 RESULTS

3.3.1 The Physical Characteristics of Kumasi

3.3.1.1 Location, history, demography and administration of Kumasi Kumasi is the second largest city in Ghana and the capital of the Ashanti Region. It lies within latitudes 6°38' and 6°45' north and longitudes 1°41' and 1°32' west on a landscape between 250 and 350 m above sea level. Kumasi used to be known as the Garden City of Africa. This was because, during the colonial era, planning interventions in Kumasi were in the form of establishing green belts and recreational parks within the layouts of residential areas. Due to the undulating topography and numerous streams that transect Kumasi, the green belts and recreational parks were invariably situated along the major stream channels. The green belts idea was borrowed from various ‘garden city’ plans popular in England. They were seen as both sanitary measures and as a way of separating residential areas especially the European, from non-European zones. The provision of a green belt of about 300 m from buildings surrounding residential areas was designed to protect Europeans residing in these areas from mosquito borne diseases. After independence, these green belts were designated as recreational space and named ‘Nature Reserves’ or ‘Parks’ on zoning maps (Schmidt, 2005; KMA, 2006). Today, this garden city status is long lost and these green belts are largely either non-existent, their borders slowly being infringed upon for urban infrastructure and housing, subsistence agriculture or are used as dumping grounds (Brook & Dávila, 2000). The suburbs e.g. Dakwadwom, Ahensan, Aboabo and Atonsu situated in or close to these ‘vacant’ land areas are frequently under floods and therefore the subject of this research.

The Kumasi District as delineated in Figure 3.1 is metropolitan because it has a population over 250,000 (Ghana Statistical Service, 2005). The Kumasi Metropolitan Assembly (KMA) is made up of the following Sub-Metropolitan areas: Asawase, Asokwa, Bantama, Kwadaso,

Figure 3.1. Map of Ghana with an insert showing limits of the Kumasi Metropolitan Assembly and the study sites [sites are labelled in letters: Atonsu Sawmill (A), High school junction (B), FRNR farms (C), KNUST Police Station (D), Family Chapel (E), Spetinpom (F), Golden Tulip Hotel (G), Kaase Guinness (H), Bekwai roundabout (J) and Dakwadwom (K)]

Manhyia, Nhyiaeso, Oforikrom, Old Tafo, Suame and Subin. There is the Metropolitan Planning Board which integrates the development planning and management of the Metropolis; monitors and evaluates development planning and management; integrates the development planning proposals from the sub-metropolitan district councils and advise the authority thereon. The KMA is bounded by the Bosomtwe, Ejisu-Juaben, Kwabre East, Afigya-Kwabre, Atwima-Kwanwoma and Atwima-Nwabiagya Districts.

The urban population of the Ashanti Region (which is 51.3 %) is only second to the Greater Accra Region (87.7 %). The Kumasi Metropolis accounts for about a third of the region’s population. The population of Kumasi grew slowly from about 3,000 in 1901 to about 180,000 in 1960 (Figure 3.2). Since then, there has been a relatively rapid growth of Kumasi exceeding a million in 2000 with a current growth rate of 5.9 % and a density of 5,319 individuals/km2 (Ghana Statistical Service, 2002). Between 1984 and 2000, the population had increased over 200 %. Due to its fast physical and demographic growth, Kumasi now extends beyond the administrative boundaries of the KMA into the neighbouring districts (KMA, 2006).

1.4e+6

1.2e+6

1.0e+6

8.0e+5

6.0e+5 population

4.0e+5

2.0e+5

0.0 1900 1920 1940 1960 1980 2000

year Figure 3.2. Population growth trajectory of the city of Kumasi (Ghana Statistical Service, 2005)

3.3.1.2 Climate of Kumasi The climate of Kumasi is tropical, of the wet sub- equatorial type, and strongly influenced by the tropical continental and tropical maritime air masses. The migration of the Inter-Tropical Convergence Zone (ITCZ) determines the seasons. Rainfall is weakly bi-modal. The major rainfall period is between May and June and the minor rainfall period is between September and October. The dry north-east trade winds (Harmattan) reach Kumasi around mid- November to early December. Relative humidity on a normal day, in Kumasi, ranges from a maximum of 100 % at 0900 GMT to 60 % at 1500 GMT. On very dry days it ranges from 40 to 84 % and 85 to 100 % on very wet days. The average minimum and maximum temperatures in Kumasi are about 21.5 °C and 30.7 °C respectively.

3.3.1.3 Hydrology of Kumasi The Metropolis is located within the Pra basin which is divided into five sub- basins by a series of ridges running generally in a north-south direction. These sub-basins are Kwadaso, Subin, Aboabo, Sisai and the Wiwi. The Suatem River flows through North Suntreso and joins the Kwadaso River at Dakwadwom to form Daban River which flows to Kaase. The Subin River flows from the Kumasi Zoo also to Kaase. The Wiwi River flows from Nsenie and merges into the Sisai River at Atonsu. The Sisai, Subin and Daban Rivers finally join the Oda River, a tributary of the Pra near Asago, south of Kumasi. In the north-western sector of the Metropolis are several small tributaries of the Owabi River. The Metropolis is therefore drained by a relatively dense network of rivers whose natural drainage runs generally from north to south, exhibiting dendritic patterns.

However, estate developments, agricultural encroachment and indiscriminate waste disposal in the rivers have impacted negatively on the drainage systems and have consequently brought some of these water bodies to the brink of extinction and increased the incidence and spread of floods and floodwaters. Some portions of these rivers have also been converted to storm drains e.g. the portion of the Aboabo River from the Airport roundabout to the High School Junction. For those that are still flowing, pollutio n is a serious problem.

3.3.1.4 Landuse and vegetation of Kumasi The Kumasi Metropolis falls within the moist semi- deciduous South- East Ecological Zone. About 80 % of the total land coverage of the Metropolitan area is planned (KMA, 2006). The major land uses that make up the Metropolis are residential, industrial, commercial, educational, civic and culture, transportation and open spaces. Among these land uses, the predominant ones are residential, commercial, educational and industrial. The residential

areas take up to 44 % of the total land area in Kumasi. About 29 % is taken up by the different land uses such as educational, civic and culture, transportation and open spaces.

Patches of vegetation reserves within the Metropolis house the Kumasi Zoological Gardens and the KNUST Botanical Gardens. Others are the Otumfuo Children’s Park at Dakwadwom, Kumasi Children’s Park at Amakom and the Kumasi Golf Course at Nhyiaeso. Aside these, there are other patches of vegetation cover scattered over peri-urban areas of the Metropolis e.g. Atasomanso, Breman and Kagyase and Owabi. Predominant local trees are Ceiba, Triplochlon and Celtis. Due to urban sprawl, most trees have been cut to make way for buildings, roads and other infrastructure. All lands adjacent to the main rivers have been declared by TCPD as green belts. These green belts are meant to serve as natural flood plains of the river but encroachment has significantly reduced their flood storage capacity. Common vegetation in these wetland areas are shrubs, grasses and food crops and vegetables like plantain, sugar cane, palm trees, okro, carrot, cabbage etc. associated with urban agriculture.

3.3.1.5 Relief, geology and soils of Kumasi The Metropolis is undulating and lies within the plateau of the south-west physical region which ranges from 250-300 m above sea levels. The area is underlain with precambium rock. There are six detailed soil types in the metropolis but the predominant soil type of the metropolis is forest ochrosol which dominates the entire region. The soil within the basins is silty clay loam which is moderately slow to slowly permeable resulting in fairly high runoff rates. This rich soil type makes it possible for foodstuff to be grown in the Metropolis and its periphery.

3.3.2 Socio-economic Characteristics of Wetland Communities in Kumasi The results of various interviews, focus group discussions with various stakeholders and the responses of 128 residents of various wetland associated communities are represented in this section. The specific suburbs being studied are Aboabo, Spetinpom, Atonsu, Ahensan, Dakwadwom and Kwadaso Estates. Large portions of these suburbs are informal or slum settlements with haphazardly arranged houses built from wood or block. There is little or no drainage network. Most houses directly release their domestic waste water through small channels to the open which finds its way into surrounding streams. Most of the lands used for these houses are designated as “natural areas” according to the landuse plan of the city. However, the land owners sell out small portions of these natural areas for various uses leading to a gradual habitation of such zones. Alleys are very narrow or virtually non-existent because of encroachments and erection of illega l structures along the alleys. 52

These communities also have high population densities with very little social infrastructure. Local entrepreneurs take advantage of the dearth in services to provide substandard facilities to these already economically depraved community. There are several private primary schools and churches mostly housed in wooden structures. Drug and chemical stores, kiosks and artisanal shops abound in these suburbs. Also, in these wetland areas, there are no culverts or bridges linking communities on either side of the water channel. Wooden foot bridges are constructed across the streams and rivers through local efforts. There is no thoroughfare in most areas so vehicula r movement in such communities is limited.

Figure 3.3 shows the occupational distribution by gender and location of the respondents considered in this study. Although this study did not seek to investigate gender distribution in these suburbs, majority of respondents (62 %) were male. Respondents who were porters or involved in food processing were all female whilst trading, farming, artisanal works and formal sector employment had both sexes involved. Trading (35 %) and artisanal work (28 %) form the major occupations of these respondents. The traders are involved in petty trading like running kiosks, petty and buying and selling of various second-hand goods in the city. Majority of these traders either do not own the businesses or have ownership with a capital base less than $100. Most of the petty traders live in Aboabo. The artisans include masons, plumbers, carpenters, shoemakers, auto mechanics, tailors and hairdressers. Most of these artisans live in Kwadaso Estates, Dakwadwom and Spetinpom. The porters were only in Aboabo, Ahensan and Spetinpom and were all immigrants who have been in the metropolis for less than 10 years.

Various types of houses were identified within 100 m from the water channel (Figure 3.4). These were permanent block structures in the form of multifamily residential facilities commonly called compound houses, single family self- contained structures (detached or semi-detached), L-shaped detached or single room structures and wooden structures. The wooden structures were found in Aboabo, Ahensan, Atonsu and Dakwadwom. Occupants were all less than 10 years in these facilities. When the economic conditions of these residents improve, most of them transform the wooden structures to more permanent block structures. The predominant housing types for respondents from Aboabo are compound, wooden or L- shaped detached houses. The houses in Aboabo unlike those of Kwadaso Estates are mostly multifamily permanent compound structures with relatively large numbers of residents even at the household level. Those in Kwadaso Estates are mostly single family accommodations.

Figure 3.3. The occupation of respondents from flood prone suburbs of Kumasi

Figure 3.4. Residential types and homeownership of respondents in the different suburbs of Kumasi

Majority of the respondents (62 %) either own the houses or are part of the extended family and do not pay rent in the houses they lived in. The largest proportion of home owners is those in wooden and permanent single room structures in Ahensan and Atonsu. Most of the multifamily residential facilities like the compound, detached and semi- detached houses have tenants who are completely separate households. There is also a relatively high tenant turnover in these areas. Most people who did not know of the annual floods usually moved out after one or two flood experiences. The relationship between length of stay and prevalence of floods was assessed using the Pearson Correlation. There was no relationship between number of years of stay in an area and presence or absence of floods (r = -0.135). Access to land for construction of houses by wetland residents is very varied in the different suburbs (Figure 3.5). The commonest forms of access to land for building a house were family inheritance, direct purchase from the chief or other landowner or squattering. Respondents in Aboabo were either squatters or living in rented houses. Most of the respondents from Atonsu and Dakwadwom inherited the land or the houses they lived in.

Various reasons have been ascribed for use of wetland areas for building houses (Figure 3.6). The commonest reasons were that wetland areas are cheap, close to other parts of the city and are “no man’s land”. These are usually areas marked out in the landuse plans as natural areas and not to be developed. However, the landowners later bypass the city authorities and sell such areas at very low prices. The city authorities do not also actively protect or use such areas therefore creating the impression that it is a no man’s land. For these reasons, the wetland areas in Kumasi quickly develop into informal settlements harbouring squatters and slum communities.

After staying in these environments for a while some residents relocate. Respondents have different motivations to move out of their current home (Figure 3.7). Some respondents in Aboabo and Ahensan love living in those suburbs and are not making any efforts to move out. In all the suburbs, respondents over 75 % of the time, completely agree to move out if given money or an alternative shelter elsewhere. Forcefully moving people out of these suburbs is not popular. For all the suburbs, the flood water level is not a first priority to cause them to move out.

100 90 80 70 60 50 40 30 20 Proportion of respondents (%) 10 0 squatter squatter squatter squatter squatter squatter from chief from chief from chief from chief from chief from chief family inheritance family inheritance family inheritance family inheritance family inheritance family inheritance from another person from another person from another person from another person from another person from another person rental accommodation rental accommodation rental accommodation rental rental accommodation rental rental accommodation rental accommodation rental Aboabo Ahensan Spetinpom Atonsu Dakwadwom Kwadaso Estates Suburb and sources of land for construction of houses

Figure 3.5. Respondents’ sources of land for construction of houses in the different suburbs of Kumasi

100

0 Proportion of respondents agreement in with reason (%) it is cheap it is cheap it is cheap it is cheap it is cheap it is cheap close to family close to family close to family close to family close to family close to family in/close to city in/close to city in/close to city in/close to city in/close to city in/close to city "no mans's land" "no mans's land" "no mans's land" "no mans's land" "no mans's land" "no mans's land" use stream water use stream water use stream water use stream water use stream water use stream water close to amenities to close close to amenities to close close to amenities to close close to amenities to close close to amenities to close close to amenities to close economic activities economic activities economic activities economic activities economic activities economic activities Aboabo Ahensan Spetinpom Atonsu Dakwadwom Kwadaso Estates Suburbs and reasons for choice of suburbs

Figure 3.6. Reasons for choice of wetland area for construction of houses in the flood prone suburbs of Kumasi

100 90 80 70 60 50 40 30 20 10 0 when forced when when forced when when forced when forced when when forced when forced when proportion of respondents agreement in with reason (%) given money to move given money to move given money to move given money to move given money to move given money to move do nothing. love I it here do nothing. love I it here do nothing. love I it here do nothing. love I it here do nothing. love I it here do nothing. love I it here new land/plot elsewhere new land/plot elsewhere new land/plot elsewhere new land/plot elsewhere new land/plot elsewhere new land/plot elsewhere when I make some money when I make some money when I make some money when I make some money when I make some money when I make some money when floods the too get much when floods the too get much when floodsthe too get much when floods the too get much when floods the too get much when floods the too get much I am already preparing to move to preparing already I am I am already preparing to move I am already preparing to move I am already preparing to move I am already preparing to move I am already preparing to move if provided an alternative shelter alternative an provided if if provided an alternative shelter alternative an provided if if provided an alternative shelter alternative an provided if if provided an alternative shelter shelter alternative an provided if if provided an alternative shelter Aboabo Ahensan Spetinpom Atonsu Dakwadwom Kwadaso Estates suburbs and reasons to move out

Figure 3.7. Factors or instruments that will motivate residents of the various flood prone communities to relocate

3.3.3 Community Efforts at Wetland Management in Kumasi Wetlands within the Kumasi Metropolis have been used for various purposes (Figure 3.8). The commonest wetland uses are urban agriculture, grazing grounds, housing and as waste disposal sites. In Aboabo, Atonsu, Dakwadwom and Kwadaso Estates, waste disposal was commonest wetland use. The wetlands in Ahensan and Spetinpom were mainly used as grazing grounds and for housing respectively. The only wetland area with significant recreational use was that at Dakwadwom. Here, Friends of Rivers and Waterbodies had an active educational programme on wetlands through the establishment of the Otumfuo Children’ s Park.

Even though these wetlands are not used sustainably, the associated communities believed that they should be protected (Figure 3.9). Majority of respondents (59 %) were in favour of protecting the wetland area. Only a few respondents (3 %) were indifferent to their protection. There are various local laws, customs and beliefs that help in the protection and management of wetlands. These are however, not being adhered to and respondents believe several factors are responsible for this (Figure 3.10). Respondents in Aboabo and Ahensan did not believe that political influence contributed to communities not managing the wetlands. Rather, it is the lack of political will by stakeholders to protect the wetlands. In Spetinpom and Kwadaso Estates, respondents were of the view that corruption and the lack of political will to enforce the laws were the two important factors debilitating wetland management in Kumasi. In all the suburbs, poverty was also viewed as a very strong factor (more than 50 % yes) accounting for wetlands not being protected.

Figure 3.8. The predominant uses of wetlands in the various suburbs of Kumasi

Figure 3.9. Perception of wetland protection in Kumasi

100

0 Proportion of respondents agreement in with reason (%) poverty poverty poverty poverty poverty poverty corruption corruption corruption corruption corruption corruption political influence political influence political influence political influence political influence political influence cost demolition of cost demolition of cost demolition of cost demolition of cost demolition of cost demolition of lack of political will lack of political will lack of political will lack of political will lack of political will lack of political will loss of cultural values cultural of loss loss cultural of values loss of cultural values cultural of loss loss cultural of values loss of cultural values cultural of loss loss of cultural values cultural of loss changes in government in changes government in changes changes in government in changes government in changes changes in government in changes government in changes Aboabo Ahensan Spetinpom Atonsu Dakwadwom Kwadaso Estates Suburbs and reasons for continual destruction of wetlands

Figure 3.10. Reasons for the continual destruction of wetlands in the various suburbs of Kumasi

3.4 DISCUSSION Similar to the situation in most developing countries, urbanisation of Kumasi is characterised by very high population growth and construction, merging up the city with the peri-urban regions. This fast urban growth is however not matched with growth in infrastructure and services, not even in the planned suburbs. The resource poor suburbs also harbour mostly rural-urban migrants and the poorest of the population. In recent times, Kumasi has received large influxes of rural youths from mostly northern Ghana to find menial jobs. A majority of these low skilled migrants end up working in the informal sector as porters, petty traders, seasonal labourers (Yeboah & Appiah-Yeboah, 2009). The porters and farmers identified in this study are likely to be people of this origin. After a little training, most of these immigrants pick up artisanal jobs like carpentry, masonry, barbers/hairdressers, tailoring/dressmaking etc. Their economic conditions usually determine their choice of urban space and or housing.

The suburbs considered in this tudy have large immigrant populations and thus, not only provide important basket of social networks for first time immigrants, but also as political hotspots. For these reasons, residents from such suburbs have been able to vehemently resist all efforts by the KMA to demolish their illegal structures. The relatively high population density and communal spirit of these residents increase their collective bargaining power. In these suburbs, the only ‘free’ areas are often the wetland areas. The immigrants therefore find these areas convenient for erecting temporal dormitories with active assistance from old residents. It is however, surprising that most people will continue to live in such structures or their improved forms in the same locations after several years of living there and an improved economic standing.

It is important to state that, majority of the wetland associated inhabitants are natives. These natives continue to live in such neighbourhoods due to the strong historical, social and economic ties developed over the years. For an immigrant also, early social networks and the cost of accommodation might be determinants of their choice of such locations. It is not surprising that these immigrants first live in makeshift shelters made of wood and may later move to more permanent single room block structures. Aboabo, Ahensan and Dakwadwom might be suburbs providing these initial favourable conditions to attract the high immigrant squatter population. Ownership of such structures might be through a form of transfer of title or a rental arrangement with the new tenant whilst the economically better off tenant moves out. Due to the high the demand for these neighbourhoods by immigrants, landowners and chiefs also then take advantage of these illegalities to lease out or sell the plots to these 63

tenants giving them title to the land. For any given environmental challenge, it is expected that there is a tipping point at which an individual finds that particular environment unsuitable for habitation. These suburbs however, are among the densest in Kumasi. The pull factors motivating people to move in and to continue staying there are stronger than the dissuading environmental pressures in these neighbourhoods.

According to the theory of the EKC, increased incomes or improved economic conditions would lead to improvement in the environment or in this case, a desire for better housing environment away from the annual floods. The findings of this study do not support this theory either. Rather, improved economic conditions led to improved housing facilities in the same environment. For example, wooden structures were converted to more permanent block structures. This is still true even when flood water levels are increasing. A model is proposed here to describe the relationship between the EKC and the decision to move or stay in these flood prone suburbs in Kumasi (Figure 3.11). It shows that the initial trajectory of the EKC curve realistically captures decision-making by poor resident in a flood prone suburb. In beginning, the victim is only interested in making ends meet. However, at the critical income threshold, the person starts to invest in improving his/her living space. The trajectory therefore does not lead to a condition of zero environmental degradation, as would be expected, but rather a tolerable level through participation in communal efforts like building of foot bridges, network of walkways etc. In another dimension, this new poor resident usually has almost 100 % tolerance to the floods. Such a person will only move when forced. With increasing income, different motivations and needs come to the fore. Consequently, flood tolerance never drops to 0 % in these communities. There are a few who are ready to move out ‘when the water level increases’. But such a subjective target is never met and these people continue to live there. The decision to move or stay in these suburbs is therefore an abrupt one independent of the environmental and economic conditions.

Homeownership in Ghana is a cultural and societal requirement or an expectation as one grows older; as having your house is a sign of success. This is aggravated by a national housing deficit and a very poor mortgage system. Most people therefore acquire the land and build their houses over several years. Land prices are very high in urban areas and the prices reflect the quality of the potential houses and services and infrastructure availability in the neighbourhood. Poor people therefore go for the cheapest available lands that are usually wetland areas and therefore not part of the official city layout scheme. It is these cheap

Figure 3.11. A proposed model to describe the decision to move or stay in the flood prone suburbs of Kumasi

unapproved houses built in wetland areas that bear the full brunt of the annual floods. From the list of pressures or initiatives in the model above, these flood prone inhabitants will rather be moved more by external pressure than their own volition. The will to continue to stay or the decreased personal initiative to move out of the wetland may be aggravated by the acquisition of property rights through the purchase of such lands. According to Lant (1994), the core of the conflict in wetland destruction is a question of property rights. However, does the owner of the wetland have the right to alter that wetland for his or her own purposes? Does the public have the right to the ecological functions and values of a private wetland? Do individual rights acquired through the purchase of wetlands override national and local government laws? Does the owner of a wetland have the right to alter it for private purposes at the public expense? How can the externalities from such alterations be internalised? When can such an individual right be revoked for the public good?

In demarcating the wetlands as natural areas, it is unfortunate that the KMA (through the TCPD) does not make any effort at assigning and enforcing property rights to such areas. If the city wishes to acquire property rights to such areas, it will have to pay a compensation (buy the rights) to the landowner. By delineating these wetland areas as ‘Nature Reserves’ or ‘Parks’, the TCPD gives de facto property rights to the public. This will imply that, the landowner has no right to sell such lands. The land tenure system in Ghana, however, does not recognise such de facto ownership. The proposals by Panayotou (1997) and Eppinka et al (2004) for more secured property rights will not work in the Ghanaian context where wetlands delineated for public good or environmental protection is not paid for. Most of the lands are privately owned and therefore the city (public) does not automatically obtain rights to the wetlands because it has been delineated as such. Whilst it may be expensive for the government to pay compensation to designate these wetland areas as public lands, it is also widely known that government has not been good managing existing public lands. Therefore, distribution of benefits and costs of maintaining wetlands as advocated by Magnani (2001) may be a better option to management of such spaces. The Otumfuo Children’s Park in Dakwadwom is an example which could be replicated in other wetland areas. Names, uses and attributes should be assigned to the wetlands and benefits and costs accrued be shared between the landowners and city authorities for management of the wetland area. Hereafter, better enforcement and effective environme nta l regulatio ns will then be needed to mainta in the wetland as such.

4 FLOODS AND ENVIRONMENTAL MANAGEMENT IN KUMASI, GHANA 4.1 INTRODUCTION Ghana is very much in tune with global environmental concerns. Over the years, it has signed several conventions and protocols (see Table 4.1) in an effort to strengthen international cooperation and linkages, and make the environment a better place for its citizenry. Also, Ghana’s Constitution, Chapter six, Article 41 (k) enjoins the citizens to protect and safeguard the environment (Government of Ghana, 1992). Apart from the articles in the Constitution, there are specific laws and policies on the environment. Some of those that are of relevance to wetlands include the Beaches Obstructions Act (1897), Rivers Act (1903), Town and Country Planning Act (1945), Land Planning and Soil Conservation Act (1953), Volta River Development Act (1961), Land Registry Act (1962), Lands Miscellaneous Provision Act (1963), Ghana Water and Sewerage Corporation Act 310 (1965), Abandoned Property (Disposal) Act (1974), Irrigation Development Authority Act (1977), Mercury Act (1989), Small Scale Gold Mining Act (1989), Control and Prevention of Bush Fires Act (1990), Towns Act (1992), Environmental Protection Agency Act 490 (1994), Lands Commission Act 483 (1994), National Development Planning Commission Act 479 (1994), Local Government Act 462 (1994), Water Resources Commission Act 522 (1996), Environmental Assessment Regulations LI 1652 (1999), Wetlands Management (RAMSAR sites) Regulations (1999), Fisheries Act (2002) and Ghana Meteorological Agency Act (2004).

These laws and regulations have historically been applied to various contexts and use of wetlands. However, the Environmental Protection Agency Act 490 (1994), Wetlands Management Regulations (1999) and the Buffer Zone Policy are currently the most important and of direct use in the protection and management of wetlands in Ghana. These laws and policies were enacted based on the current Constitution of Ghana and have various institutions mandated to regulate, direct and implement them. Ghana may be very sufficient in terms of laws and policies. However, the human resources and institutional framework under which the laws and policies are being implemented are equally important. This chapter therefore assesses the laws and policies that apply to the environment in Ghana. Particular emphasis will be placed on wetland management and floods and the stakeholders and institutional framework under which they operate.

Table 4.1. Ghana’s international commitments to the environment Date of Accession/ Convention/Protocol Ratification Convention on the Continental Shelf, Geneva, 1958 29 April, 1958 Convention on Fishing and Conservation of the Living Resources of 29 April,1958 the Seas, Geneva, 1958 African Convention on the Conservation of Nature and Natural 17 May, 1969 Resources, 1969 Convention on International Trade in Endangered Species of Wild 14 November, 1975/ Fauna and Flora, Washington, 1973 12 February, 1976 United Nations Convention on the Law of the Sea, Montego Bay, 10 December, 1982/ 1982. 7 June, 1983 Convention on the Conservation of Migratory Species of Wild 19 January, 1988 Animals, Bonn, 1979 Convention on Wetlands of International Importance Especially as 22 February, 1988 Waterfowl Habitat, Ramsar, 1971 Convention for Cooperation in the Protection and Development of the 20 July, 1989 Marine and Coastal Environment of the West and Central African Region, Abidjan,1981 Protocol Concerning Cooperation in Combating Pollution in Cases of 20 July, 1989 Emergency, Abidjan, 1981 Vienna Convention for the Protection of the Ozone Layer, Vienna, 24 July, 1989 1985 Montreal Protocol on Substances that Deplete the Ozone Layer, 24 July, 1989 Montreal, 1987 London Amendment to the Montreal Protocol on Substances that 24 July, 1992 Deplete the Ozone Layer, London, 1990 Convention on Biological Diversity, Rio de Janeiro, 1992 29 August, 1994 United Nations Framework Convention on Climate Change, New 6 September, 1994 York, 1992 United Nations Convention to Combat Desertification in those 27 December, 1996 Countries Experiencing Serious Drought and / or Desertification, Particularly in Africa, Paris, 1994 Copenhagen Amendment to the Montreal Amendment on Substances 30 September, 2000 that Deplete the Ozone Layer, Copenhagen, 1992 Stockholm Convention on Persistent Organic Pollutants 30 May, 2003 Montreal Amendment to the Montreal Protocol 8 August, 2005 Beijing Amendment to the Montreal Protocol 8 August, 2005

4.2 ENVIRONMENTAL MANAGEMENT AND THE NATIONAL ENVIRONMENTAL ACTION PLAN

Traditionally, Ghana’s environmental challenges have been the depletion of her resources through deforestation, desertification and soil degradation (Okley, 2004). Recently, urbanisation and industrialisation have also come to the fore (UN-HABITAT, 2009). In March 1988, the Government of Ghana initiated a major effort to manage these environmental challenges. The exercise culminated in the preparation of a strategy and a National Environmental Action Plan (NEAP), to define a set of policy actions, related investments, and institutional strengthening activities to make Ghana's development strategy more environmentally sustainable and to improve the surroundings, living conditions and the quality of life for all generations of Ghanaians. The NEAP provided a framework for interventions deemed necessary to safeguard the environment. For areas related to wetlands, the policy actions included the establishment of the proposed Water Resources Commission; adoption of the proposed Water Law; adoption of policy framework for extraction of water for the different uses; integrated planning of river basins and the control of waste discharge into water bodies; and establish watershed protection areas. Institutions and detailed working documents to these objectives have all been produced.

The NEAP also includes a National Environmental Policy to provide the broad framework for the implementation of the Action Plan. The policy aims at ensuring sound management of resources and the environment and to avoid any exploitation of these resources in a manner that might cause irreparable damage to the environment. The policy identified among others, the following issues to be dealt with: population pressure, land ownership and tenure, landuse planning, agricultural land use, water resources assessment, problems of water management, water pollution, watershed protection, urbanisation and urban degradation.

4.2.1 The Environmental Protection Agency The Agency started as the Environmental Protection Council (EPC) in 1974 and was responsible for all activities and efforts aimed at protecting and improving the quality of the environment. The EPC functioned as an advisory body until 30th December 1994 when it was transformed into an Agency by the Environmental Protection Agency (EPA) Act 490. By this Act, the EPA became a corporate body mandated to implement the NEAP with powers to sue and be sued. The EPA by law is therefore the leading public body for protecting and improving the environment in Ghana. The objectives of the EPA, among others of interest to this research, are to:

mainstream the environment into the development process at the national, regional, district and community levels; ensure that the implementation of environmental policy and planning are integrated and consistent with the country’s desire for effective, long- term maintenance of environmental quality; and guide development to prevent, reduce, and as far as possible, eliminate pollution and actions that lower the quality of life. It performs these with the following guiding principles, inter alios, partnerships, pollution prevention and control, ecosystem management, environmental justice, environmental education, compliance and enforcement. The EPA has power to issue environmental permits and pollution abatement notices in order to control waste discharges and emissions and to prevent or reduce noise pollution. It is supposed to provide standards and guidelines in relation to air, water and other forms of environmental pollution. It also has authority to ensure that developers of commercial entities comply with environmental impact assessments.

4.2.2 Water Resources Commission The major laws that guide the regulation and management of water resources in Ghana are contained in the Water Resources Commission Act (Act 522 of 1996) and the Water Use Regulations (L.I. 1692 of 2001). This Legislative Instrument categorises and prioritises water uses in Ghana and spells out procedure for acquisition of water use permits. Water, like all other natural resources in Ghana, falls within the provisions of Article 269 of Ghana's Constitution, and is therefore vested in the President on behalf of and in trust for the people of Ghana. There is no private ownership of water in Ghana, but that the President, or anyone so authorised by the President, may grant rights for water use. The Water Resources Commission was set up and given the powers by the President for the regulation and management of water resources and for the coordination of policies in relation to them (Government of Ghana, 1996). The Commission is the focal point in fostering coordination and collaboration among the various actors involved in the water resources sector.

The Commission’s management of water resources is underpinned by the National Water Policy (Ministry of Water Resources, Works and Housing, 2007). The Policy recognises the perceived abundance of water resources in Ghana. However, the production and utilisation of water resources for consumptive and non-consumptive uses is not at optimal level and therefore calls for an efficient and effective management of available water resources. The Policy is based on principles of Integrated Water Resources Management (IWRM). In

particular, it encourages the sustainable exploitation, utilisation and management of water resources to ensure full socio-economic benefits to the present and future generations, while maintaining biodiversity and the quality of the environment.

Wetland ecosystems in Ghana constitute about 10 % of the country’s total land surface and are grouped as marine/coastal, inland and man-made. Recognising the scale of distribution and importance of wetlands, the Ministry of Lands and Forestry (1999b) integrated wetland issues into the National Landuse Policy. The Policy recognises wetlands as environmental conservation areas and precludes the following practices: physical draining of wetland water, draining of streams and water courses feeding the wetlands, human settlements and their related infrastructural developments in wetlands, disposal of solid waste and effluents in wetlands and mining in wetlands. The Government of Ghana sees its role in wetlands management as best performed through partnership and co-operation with local people, governmental and non-governmental organisations, and the private sector.

Ghana became a signatory to the Ramsar Convention in 1988. A major obligation of signatories to the Convention is the implementation of the principle of ‘wise use’ of their wetland resources. Therefore, in consultation with stake-holders, the “Managing Ghana’s Wetlands: A National Wetlands Conservation Strategy” document was prepared to promote the participation of the local communities and other stakeholders in the sound management and sustainable utilisation of Ghana’s wetlands and their resources (Ministry of Lands and Forestry, 1999a). The document provides a national definition and distribution of wetlands, a management strategy for wetlands and a policy framework for incorporating wetlands management into the day- to- day activities of Government, organisations, traditional authorities, communities and individuals. Accordingly, the Government of Ghana has overall responsibility for perpetuating wetland areas, and also to administer a range of social, economic and environmental programmes which impact on wetland management throughout the country whilst local communities directly manage the "wise use" of wetland resources in their localities.

Also, there is the Buffer Zone Policy for the protection, regeneration and maintenance of the native or established vegetation in riparian buffer zones in Ghana (Water Resources Commission, 2008). This policy is a further attempt at managing Ghana’s river basins in accordance with the IWRM approach and to harmonize traditional and existing public institutional standards on buffer zones in Ghana. The Policy adopts a 3-tier buffer zone as the norm for effective erosion control and halting of water quality degradation caused by 71

nutrients, animal or human wastes, sediments and runoff. It however, gives provision for variation of this optimum buffer “design” from the prescribed values to accommodate local requirements. The buffer zones are also to be permanently reserved by legislative instruments with the consent of local authorities and incorporated into the local landuse plans. Local authorities must take steps to minimize the risk of conversion of buffer zones to uses incompatible with sustainability of their functions and exclud e exploitation or occupation inimical to the purposes of designation of the areas as buffer zones. Government is also to ensure that all designated riparian buffer zones along rivers, streams, and around lakes, reservoirs or other surface water bodies are adequately vegetated and sustainably managed to restore and maintain their ecological integrity. This is expected to provide socio-economic benefits to local communities and in fulfilment of Ghana’s overall water, environment and landuse policies, the MDGs and the Growth and Poverty Reduction Strategy (National Developme nt Planning Commiss io n, 2005).

4.2.3 Lands Commission Ghana has basically two types of land ownership: state and private. State lands are lands compulsorily acquired by the government through the invocation of the appropriate legislation, vested in the President and held in trust by the State for the entire people of Ghana. On the other hand, private lands in most parts of the country have communal, family or individual ownership. Stool or skin lands are a feature of communal land ownership in most Akan traditional groups in southern Ghana and in most traditional groups in northern Ghana respectively. However, scattered all over Ghana are a number of traditional groups, which do not recognize a stool or skin as symbolising private communal land ownership. Also sandwiched between the public and private lands are vested lands, which are a form of split ownership between the state and the traditional owners. Fundamentally, land ownership is based on absolute allodial or permanent title from which all other lesser titles to, interests in, or right over land derive. Normally, the allodial title is vested in a stool, skin, clan, family, and in some cases, individuals.

Because of this diversity of land tenure regimes, the Lands Commission was established under article 258 of the 1992 Constitution and came into being through the Lands Commission Act (Act 483 and 767 of 1994 and 2008 respectively). The Commission is an integration of the various land agencies for the efficient management of public lands and other lands, advise and facilitation of good land delivery system in the country. The Commission’s work is being guided by the National Land Policy. The National Land Policy, District

Assemblies Act 462 and the National Development Planning Systems Act 480 enjoin District Assemblies in conjunction with land owners and the Lands Commission to prepare planning schemes for all land uses to facilitate dispositions of land for development. In making these planning schemes, the policy states that wetlands are environmental conservation areas and the following uses considered incompatible with their ecosystem maintenance and natural productivity are strictly prohibited: physical draining of wetland waters, damming of streams and water courses feeding the wetlands, human settlements and their related infrastructural development in wetlands, disposal of solid waste and effluents in wetlands and mining in wetlands.

4.2.4 Local Government and Environmental Management For the purposes of local government, Ghana is divided into districts (Government of Ghana, 1993). The Local Government Act (Act 462 of 1993) makes the District Assembly the highest political authority in any district with deliberative, legislative and executive powers. Among others, the district assembly is responsible for the development, improvement and management of human settlements and the environment in the district. The assemblies are empowered to enact the appropriate bye-laws for planning and management of resources in their geographical area. In accordance with the NEAP, the responsible committee for environmental issues at the local level is the District Environmental Management Committee (DEMC) of the district assembly. The number and composition of this committee varies from district to district. In general, the DEMCs are made up of about 15 members. These include the regional programme officer of the EPA, five assembly members (one as chair), at least two of whom should be women, two representatives of environmental NGOs, officers from decentralised institutions like the Department of Parks and Gardens, TCPD, Department of Agriculture, Forest Services Division, District Health Directorate, the District Education Service, Ghana Water Company or Community Water and Sanitation Division and National Council on Women and Development. The EPA's Regional Programme Officers and their staff are expected to provide support and assist with training of DEMCs and to act as a conduit for dissemination of information from EPA to the districts. The DEMC enforces environmental standards and regulations set by the EPA and other national regulatory agencies within its jurisdiction.

4.3 WETLANDS, FLOOD PREVENTION AND MANAGEMENT

4.3.1 Stakeholders’ Capacity for Managing Wetlands and Floods The following stakeholders were identified to be involved in wetlands and or flood management in Kumasi: KMA, HSD, TCPD, NADMO, Assemblyman, Chiefs and NGOs (Friends of Rivers and Waterbodies).

There were varied perceptions about the responsibilities of stakeholders for the management of wetlands (Figure 4.1). The commonest stakeholders mentioned were NADMO and the KMA, especially, in Atonsu and Dakwadwom. Also, at all the study suburbs, chiefs and land owners were rated high in the management of floods and wetlands. In Ahensan, respondents completely agreed that the chiefs and landowners were responsible for the wetlands. In Spetinpom, all the institutions listed as being responsible for managing wetlands and floods were ranked relatively high by community members.

Apart from the Assemblyman, the other stakeholders were given the opportunity to assess each other on wetlands protection and management and flood prevention and management. The outcome of these assessments is presented in Table 4.2. Among the different stakeholders, NADMO and the KMA were considered the most financially resourced to manage floods and wetlands. NADMO however lacked the technical expertise at local levels. The chiefs also lacked the technical and coordinating capacity in wetlands protection and management and flood prevention even though they were considered very much respected and could sometimes mobilise the financial resources.

Figure 4.1. Assessment of stakeholders’ responsibility for the management of floods and wetlands in Kumasi 74

4.3.1.1 Kumasi Metropolitan Assembly (KMA) According to the Local Government Act (Act 462 of 1993), the KMA is the local authority of the metropolis. This study revealed that decentralised structures for environmental management in Kumasi are well in place. Flooding issues had been incorporated in the Medium Term Development Plans of the Assembly (spanning 2006-2009). According to the Assembly, the floods are caused by residential dwellings and other structures built in waterways, dumping of refuse in gutters and drains and the inability of existing culverts to receive large volumes of water whenever there is a heavy downpour. Some mitigating strategies such as drain constructions or enlargement of existing ones, desilting or dredging of streams, clearing of waterways (demolition of obstructing structures) and public education were being carried out. Several houses along waterways have therefore been marked for demolition. This was an on-going exercise in all the flood prone suburbs like Aboabo, Ahensan, Anloga, Asafo, Asokwa, Atonsu, Breman, Daban and Oforikrom. According to the KMA, plans were also in place to curb further building on waterways through strict enforcement of the building regulations.

Respondents from the communities had varying perceptions about the demolition exercises carried out by the KMA (Figure 4.2). Almost 50 % of the respondents in Aboabo and Ahensan had houses in their neighbourhood marked for demolition and virtually none in Dakwadwom. On the contrary, whilst there is a general support by respondents for demolition in the other suburbs, no respondent in Aboabo supported it. Most support for demolition of houses on waterways came from Atonsu.

4.3.1.2 Town and Country Planning Department (TCPD) The TCPD is a government agency responsible for planning and management of the orderly development of human settlements; providing planning services to public authorities and private developers and provision of layout plans to guide orderly development in Ghana. It formulates goals and standards relating to the use and development of land in areas of rapid urban growth. This Department is present in the Metropolis as one of the decentralised government institutions. In performing these functions, the TCPD plays a key role in delineating and protecting wetlands and waterbodies and the prevention of floods in built up areas.

Apart from Spetinpom and Kwadaso Estates where the TCPD is recognised as responsible for floods and wetland management, all the other suburbs virtually did not know the TCPD.

4.3.1.3 National Disaster Management Organisation (NADMO) NADMO is responsible for the coordination of the management of disasters and similar emergencies. It ensures that the country is prepared to prevent disasters and manage them well when they occur. It does this through training, education and awareness creation, coordinating the activities of various stakeholders in the management of disasters and rehabilitates persons affected by disasters. NADMO is very well decentralised to regional and district offices. Both Regional and Metropolitan offices of NADMO are found within Kumasi.

Among all the institutions, NADMO is the most popular with the respondents as being responsible for floods and wetlands management. This is especially so in Atonsu and Dakwadwom where respondents virtually assigned all floods and wetlands management responsibilities to NADMO.

Table 4.2. Stakeholder capacity assessment on wetlands and flood prevention and manage me nt in Kumas i Rating (increasing Stakeholder Capacity from left to right) Comment 1 2 3 4 Personnel X The KMA knows what to do and how to do Technical/logistics X it. What is not available are the finances and KMA Financing X logistics to carry out their plans. The KMA Coordination X has other priorities and only reacts in times of flood. Personnel X NADMO has a budget that is usually Technical/logistics X increased in times of emergency. The NADMO Financing X organisation is however very politicised and Coordination X lacks the qualified technical personnel. Personnel X This organisation is well staffed in Kumasi. Technical/logistics X They however do not coordinate with other TCPD Financing X stakeholders in their landuse planning Coordination X mandate. Personnel X This is very much an obscure organisation in Technical/logistics X function and visibility in the KMA. They are HS D Financing X relatively under-resourced and much Coordination X marginalised. Personnel X They are very important in landuse decisions Technical/logistics X in Kumasi. Sale of land is a very big source Chiefs Financing X of finance for this institution. Their decisions Coordination X lack technical backing from the various experts. Personnel X Has the personnel and capacity to coordinate NGO (Friends of Technical/logistics X wetland management issues. They are well Rivers and Financing X known and resented in some of the suburbs Waterbodies) Coordination X of the KMA Generally, the Metropolis has the technical capacity but does not coordinate its flood Total and general assessment 7 3 4 10 and wetland management efforts with the little logistical and financial resources available.

Figure 4.2. Spatial distribution of houses marked for demolition and attitude towards demolition of houses in waterways by the Kumasi Metropolitan Assembly

4.3.1.4 Hydrological Services Department (HSD) The HSD is responsible for assessing the country’s water resources (quantity, quality and distribution in time and space), the potential for water related development, and the ability to supply actual and foreseeable demand. The HSD also assesses the environmental, economic and social impacts of existing and proposed water resources management practices, projects and the adoption of sound policies and strategies. Above all, they assess the impacts of other non-water sector activities such as urbanisation, waste management, forest harvesting etc. on water resources thereby providing security for people and property against water–related hazards such as floods, storm surges and droughts.

Among respondents, this is the most unknown stakeholder for the management of floods and wetlands. The HSD and its functions are not known in Aboabo, Atonsu and Dakwadwom. Like the other institutions, it was only in Spetinpom that the HSD was known and considered, more than 50 % of the time, responsible for floods and wetlands management.

4.3.1.5 Chiefs and Assemblymen Chiefs or the chieftaincy institution is a traditional leadership arrangement which is guaranteed in the Constitution of Ghana. The Constitution also recognizes customary law as a source of law in Ghana and explicitly states that the management of stool lands are to be in accordance with the relevant customary laws and usage. Majority of the land of the Metropolis is customary or stool lands and customary law vests the powers in the appropriate stool (chief) to manage such lands on behalf of and in trust for their people.

The assemblymen on the other hand, are a product of the local government system representing decentralised decision making bodies at the community level. Officially, they represent the community in the KMA. In most communities, the assemblymen work very closely with the chiefs as community leaders to coordinate community development activities and relief efforts in emergency situations.

Apart from Atonsu and Dakwadwom, the roles of chiefs in wetland and flood management are very much recognised by all the stakeholders. Respondents in Ahensan and Spetinpom, almost 100 % of the time, think chiefs are responsible for the management wetlands and floods. The Assemblyman was rated higher than the chief in solving flood issues and the provision of relief. The rating of an assemblyman in any community depends on the dynamism and efforts the person puts in solving community problems. The assemblyman and the chief will usually be the first line of contact for person(s) or institution(s) coming to the

community or bringing assistance in times of flood. Similarly they collaborate and mobilise communal labour for various flood interventions.

4.3.1.6 Non-governmental organisations (Friends of Rivers and Waterbodies) Several non-governmental organisations (NGO) have interest in water and watershed management in Ghana. One local NGO that has risen to prominence in the Kumasi Metropolis is the Friends of Rivers and Waterbodies. The Friends of Rivers and Waterbodies is an amalgam of volunteer scientists, businessmen, economists, opinion leaders and other professionals as well as students. Volunteers are mobilised and empowered to execute tasks in consonance with traditional and scientific values, to provide water and protect waterbodies from pollution, siltation and abuse. A playground in Dakwadwom called Otumfuo Children’s Park, was established by the Friends of Rivers and Waterbodies on one side of the wetland. The NGO has also undertaken several wetland related activities in the community. Respondents from Dakwadwom therefore rated NGOs high for management of floods and wetlands. All the other suburbs rated the NGOs relatively low.

4.4 DISCUSSION In spite of the long list of institutions and environmental laws, Ghana’s urban environment is one of the worst in the sub region. The urban environmental issues described above are serious and damning to Kumasi and Ghana as a whole. According to the FAO (1985) few sewage treatment plants existed in Ghana. Since this publication, the situation has rather become worst because of the increase in urbanisation and little accompanying infrastructure development (Brook & Dávila, 2000; Keraita et al., 2003; Fobil & Atuguba, 2004; UN- HABITAT, 2009). Therefore, most of the raw domestic effluents are discharged directly into water bodies. Information on the extent of pollution caused by domestic wastes on water bodies exists only for the main rivers and reservoirs that serve as drinking water sources for communities. This is well captured in a recent United Nations Human Settlements Programme report that, Ghana’s environmental challenges and priorities are urban planning and management, housing/urban development and the environment (UN-HABITAT, 2008b). These environmental issues are the result of conflicts between human and environmental interests due to urbanisation that has gone haywire. Also, the environmental framework based on which one can evaluate these issues will be difficult to define because Ghana has no urban policy. This discussion will therefore be compared to best practices elsewhere and an attempt made at relating it to local conditions.

Appending signatures to conventions and treaties seem not to be a problem for Ghana irrespective of the obligations and commitments required. This is because the translation of the obligations of conventions into local laws and policies for implementation and enforcement has not been followed through to the letter. The divide between policy and implementation is evident in the contrast between the number of national and international obligations and the unsanitary conditions and the increasing degradation of wetlands. For example, Ghana ratified the Ramsar Convention in 1988, demonstrating her willingness and commitment to reverse wetland loss and degradation. However, it took more than a decade afterwards to fulfil the requirements of the Convention by recognising wetlands in the National Land Policy (Ministry of Lands and Forestry, 1999b) and to develop a strategy for management and sustainable utilisation of wetlands and their resources (Ministry of Lands and Forestry, 1999a). Now, Ghana has to also fulfil the operational objectives of the Ramsar Strategic Plan 2009-2015 which among others, required all parties to complete their national inventories and have a web-based database; put in place a national wetland policy and management plans founded on sound scientific research and identify priority sites for restoration. Since becoming a member of this Convention, research on wetlands that could help meet this operational objective has barely been conducted. Most wetland research has been on the Volta and Densu Rivers because of their hydroelectric use and provision of drinking water to Accra respectively.

Even though the environmental governance system is supposed to be decentralised according to the Constitution and Local Government Acts, the extent of decentralisation is not the same across the country. Whilst the older and richer districts have the various decentralised government agencies, most of new or poor districts are still under the oversight of the regional level agencies. There are, therefore, differences in availability of expertise and even representation on the DEMC. For example, the EPA, TCPD, Forestry Service Division and the Land Commission are not present in most districts in Ghana. These differences give multi- dimensionality to the already complex problems associated with the urban environment. Also prominent in most districts that are fortunate to have these agencies, e.g. the KMA, is the conflict of interests among these agencies. For example, Forestry Service Division, EPA, Department of Agriculture, District Directorate of Health Services, have overlapping functions and sometimes try to outmanoeuvre each other into approving or disapproving some environmental issues. This has been aggravated by high staff turnover, frequent changes in the institutions' mandates, the complexity of the institutional framework and weak coordination.

The local government system practised in Ghana is very much in line with international best practices of participatory governance and bottom-up approaches. Also, according to all the laws and policies, environmental action in Ghana should be bottom- up. In the spirit of the Constitution and policies of Ghana, the use of the term ‘Government’ at the local level on environmental issues refer to the DEMC. This is because environmental protection and management is a constitutional mandate of every citizen. This mandate is exercised through local level environmental management and decision making through the DEMC. This is expected to create ownership and empowerment for the sustainability of environmental programmes by taking cognisance of local needs and aspirations (Okley, 2004). It is however, unfortunate that the DEMCs are either non-existent or non-functional in most assemblies. At the KMA, the EPA has overshadowed this Committee.

Unless land that has been acquired and protected for the public good, the system of land ownership in Ghana does not augur well for the environment. Urban land value is always on the ascendency. For example, wetland areas gain value due to urbanisation and improved construction engineering feats. Like all other jurisdictions, there is market failure in the environmental services of these wetlands. Wetland uses for residential and commercial developments therefore often out-compete their use for environmental services as a landuse priority. Urban wetland areas are therefore very popular sites for fuel stations in Kumasi and other urban areas of Ghana. Rather than protecting, the EPA’s regulations for siting such businesses promote the use and destruction of urban wetland areas. Also, most of the institutions, especially, the DEMC, are non-functional when it comes to enforcing and implementing the planning schemes for these urban areas.

The rate of urbanisation of Kumasi has overwhelmed the authorities. Settlements are constructed with layouts from landowners before the TCPD develops or approve the chief’s layout for the suburbs. In these instances, the wetlands are usually long destroyed by the time these state institutions start to work. This is because the constitutional powers of governments to protect the public welfare and the political acceptance of existing environmental regulations are generally difficult to enforce (Lant, 1994). Albeit different geographical, economic and social context, Sims & Schuetz (2009) showed that local regulations and bylaws slowed wetland conversion to residential uses in Massachusetts. In the case of Kumasi and Ghana at large, a strong DEMC and TCPD working in collaboration with the Chiefs and other landowners could be a panacea to the wetland loss and environmental management.

From the described conditions of Kumasi, the future environment will largely consist of human-dominated ecosystems, managed to varying degrees, and become harder and harder to maintain in the poor suburbs. The victims of this looming crisis are both the state and the persons at these wetland sites. Already plenty resources are spent yearly in supporting flood victims in Kumasi. Also, the capacity rating of flood and wetland stakeholders has implications on their level of responsiveness to the affected communities. With the limited resources, there is the need to strengthen stakeholder collaboration and coordination. Importantly, though NADMO has been providing relief items and materials to flood victims, it is usually for immediate and short time relief. It is rather more important for these stakeholders to focus on the reduction of socioeconomic vulnerability of community members and the promotion of the culture of preparedness of both the authorities and community members. This improves our collective resilie nce to floods.

From the information gathered, a general model for management of floods and wetlands within the Kumasi Metropolis has been proposed in Figure 4.3. Whilst an integrated and collaborative approach between stakeholders is advocated, the model takes cognisance of the strengths and weaknesses of stakeholders in addressing the various flood management elements. All these elements have to be treated with equal importance. What differ are the scale and the dimensions of the issues. Based on the findings of this research, different levels of emphases are proposed for the different dimensions or issues. The proposed order in decreasing significance of the issues is: social, technical, institutional, environmental and legal. With the limited financial resources available to the KMA, flood management within the metropolis should focus on the social and technical elements to achieve better impacts.

The current level of economic development of Ghana is also a contributory factor to ineffective implementation or non- enforcement of environmental laws. Most obviously, law enforcement has financial implications and therefore will depend on the wealth of the country. Developed countries may be more capable and willing to comply with their environmental obligations than poorer countries because implementation places relatively less burden on the national cake. On the other hand, if the cost of implementation is very low, or if implementation produces a net economic gain, a country’s level of economic well-being may not be very critical for its success. This is because, the decisive factor is the relationship between the cost of implementation and the capacity of a country’s economy to absorb them.

NADMO Chiefs Stakeholders KMA TCP HSD NGOs

Flood Solid waste Strengthen Storm water Flood preparedness Landuse Restoration or Management management education and drainage and adaptation planning protection of urban Elements law enforcement strategies greens or wetlands

DimensionsDimensions of Technical Environmental Institutional floodflood managementmanagement Legal Social

Figure 4.3. A conceptual model for management of floods and wetlands within the Kumasi Metropolis (the thickness of the line is a reflection of the importance of the issue or intervention or institution)

Apart from a nation’s economic situation, national environmental consciousness and institutions are also very necessary for the protection and management of wetlands. Ghana needs the birth of a green party. In many developed countries, rising environmental consciousness has given birth to green parties (or the vice versa) that are revolutionising the political landscape with green politics by putting environmental issues to the fore in national policy debates. The challenge now lies as to whether it is the economy or the people that give rise to green politics. If it is the economy, then Ghana needs to reach a certain level of economic independence for businesses to go green and still survive the competition. The Ghanaian economy and businesses are not yet there. If it is the people, there is the need for a critical mass of people to have a level of dissatisfaction with the environment. That also, Ghanaians are not yet there. The option left is for the developed countries to dictate the pace towards a better urban environment through strings tied to bilateral and multilateral assistance or introduction and enforcement of environment related trade barriers and conditionalities.

5 WETLAND ECOLOGY AND VEGETATION GEOGRAPHY IN THE KUMASI METROPOLIS 5.1 INTRODUCTION Human settlements, like other systems, are under constant change involving growth, decline, aesthetic changes etc. depending on the age of the settlement, demographic trends, planning and resources available. Urbanisation is currently the biggest change most developing country settlements are undergoing (Bhattacharya, 2002). It is both a driver and outcome of economic, political, cultural, social and physical processes. The ecological consequences of the increasing built environment are pollution and landscape changes resulting in increased range of spatial diversity from ‘biological deserts’ with paved or gravelled surfaces or manicured monospecific lawns to biological hotspots like urban wetlands and abandoned industrial sites and rail lines (Niemelä, 1999). This is because urbanisation and the associated environmental stress lead to fragmentation of natural vegetation and changes in their structure and floristic composition (Altobelli et al., 2007). Urbanisation also alters local climate and air quality through increased concentrations of oxidants (O 3, SO2), nutrients and chemicals (nitrates, cations especially heavy metals), decreased net radiation and average wind speed, increased cloudiness and precipitation and temperature (“heat island” effect). These changes imply that urban sites cannot be directly compared with non-urban sites, any more than sites in different climatic zones (Ehrenfeld, 2000). Therefore the wetlands of Kumasi were considered as unique domains separate and different from the adjoining peri-urban or non-urban wetlands or urban wetlands elsewhere in the country.

Most urban ecological research has focused on contributions of greenhouse gases and the heat island effect (Hidalgo et al., 2008; Takebayashi & Moriyama, 2009). Urban wetlands on the other hand, deliver a wide range of ecosystem services relating to climate change mitigation and or adaptation, water catchments, buffer against flooding, soil purification, water and air or act as carbon sinks and contribute to human well-being (Joosten & Clarke, 2002; Berkowitz et al., 2003; Janssen et al., 2005; Nelson et al., 2009). Despite their ecological and socioeconomic importance, urban wetlands have been understudied in the analysis of global environmental change and flooding. Urban wetland areas are usually characterised by vegetation patches of different sizes and nature, depending on the landuse options developed by planners during the urbanisation process or patches left out in the landuse planning. The vegetation is often affected by the hydroperiod or hydrologic changes and availability of nutrients. Also, toxic materials in the runoff can interfere with biological processes of wetland

plants, resulting in impaired growth, mortality, and changes in plant communities. For example, direct storm sewer inputs affect plant community structure and vegetation dynamics (Ehrenfeld, 2000). Plants, as individuals or as communities, could therefore be valuable indicators of water regime and other environmental conditions (Haslam, 2003).

Reinelt et al. (1998) in a study on the impacts of urbanisation on palustrine wetlands found that, changes in plant community composition may describe impact of pollution. In other studies, it was found that contrary to that of Reinelt et al. (1998), vegetation types of different floristic composition or vegetation cover or density are rather predetermined by geomorphology, soil conditions, and water regime (Federal Interagency Committee for Wetlands Delineation, 1989; Barrat-Segretain & Amoros, 1996; Coroi et al., 2004; Hejcmanovā-Nežerková & Hejcman, 2006; Sorrell et al., 2007). On a large scale, factors such as soil and climatic conditions may be the determining factors that control the distribution and abundance of vegetation. However, on a small scale, while the soil and climatic conditions may not be spatially varied, the vegetation may be very varied. These spatial differences in species distribution are pronounced in the wetlands of Kumasi. This chapter therefore seeks to identify the underlying factors that determine the wetland vegetation type and distribution in Kumasi. Knowledge of the characteristic vegetation is useful in delineating wetland boundaries and reduces incidence of conflicts associated with landuse planning and wetland encroachments. This will be an input towards restoring Kumasi to its Garden City status.

5.2 METHODOLOGY

5.2.1 Reconnaissance Survey of Wetlands in Kumasi and Selection of Study Sites The wetlands in Kumasi were identified and delineated using the definition of Joosten & Clarke (2002). Following a detailed study of drainage maps of Kumasi and use of aerial photography, physical inspection and ground truthing of these drainage channels was done using a handheld GPS. Detailed descriptions of landcover and landuse types were recorded and photographs of salient features taken. The data collected was then be used to validate the information gathered from the digitised drainage maps and aerial photography.

Based on the information from this survey, some of the major wetland sites of ecological, landscape, economic and social interests were chosen for the study. Wetlands of ecological interests included, but not limited to, areas with at least 50 m of wetland vegetation on both sides of the water channel. Areas with economic and social activities directly affecting or bordering the wetland (e.g. tie-dye industries, car wash, agriculture, settlement etc.) were

selected as sites of economic interests. These sites were selected for detailed socio-economic studies as described in earlier chapters.

5.2.2 Ecological Survey of Wetlands in Kumasi The various wetlands of interest for this study were selected as described in section 3.2 above. For each wetland area, e.g. along a stream X, three permanent transects A, B and C were marked out not less than 50 m apart based on the ecological, landscape and surrounding economic and social activities (Figure 5.1). The GPS coordinates of these transects were noted to aid future identification of these sites if markers are removed. Within the permanent transects detailed ecological studies were carried out at various sections of the transect (Coroi et al., 2004; Barrat-Segretain & Amoros, 1996).

Figure 5.1. A schematic diagram of a sample study site along a stream X showing the transects and quadrat layout

5.2.2.1 Soil sampling A graduated soil corer of diameter 5.7 cm was used to take soil core samples to a depth of 30 cm (Driessen et al., 2001, Maltby & Barker, 2009) and within 10 m from the water channel. Each sediment core was carefully studied and described. The sample was then put into labelled sterile polythene bags and stored in ice chests for further laboratory analysis. Also, as a proxy for testing soil consistency, the corer was pushed into the soil by hand at these locations to estimate the resistance to penetration by measuring the total depth of

insertion. In the laboratory, a portion of the soil sample was taken and the texture determined using the Bouyoucos hydrometer method of mechanical analysis (Beverwijk, 1967). The textural classes were based on the US Department of Agriculture soil particle size ranges.

The total organic matter was estimated from these ice cooled samples using the loss on ignition method. Each sample was first treated with excess of 10 % HCl to remove any carbonates that might be present. The water was first removed by air drying in an oven for 24 hours at 70 °C. The dried sample was ground with an agate mortar and sieved through a 1000 µm sieve to homogenise it. A known weight was then placed into a preheated furnace at

500 °C for 2 hours. After ignition, 2-3 drops of (NH4)NO3 was added and the samples put back into the furnace for 10 minutes in order to remove any unburned organic carbon. The total organic carbon in each sample was the weight lost on ignition.

5.2.2.2 Stream sediments and physicochemical characteristics Along each permanent transect the presence or absence of water at the surface of each of the sampling points was noted monthly. The total duration of surface water during the year was used to assess the effects of wetland hydroperiod on vegetation dynamics.

In each stream, temperature (bottom), turbidity, conductivity, total dissolved solids, dissolved oxygen and pH of the water were measured using a Hanna multiparameter water quality portable meter (HI 9828) during the peak rainy season and dry season. Surface sediment samples were also taken at a depth of 0-15 cm (U. S. Environmental Protection Agency, 1985) and quickly packed in air tight polythene bags to be analysed for various heavy and trace metals. In the laboratory, samples were air dried in a vacuum oven at 70 ºC for 3-4 days until constant weight and ground in an agate mortar for homogenisation. Approximately 0.2 g of dried sample was extracted overnight with 2 ml concentrated nitric acid (HNO3) and evaporated until near dryness and no nitrous vapours were released. After cooling, 5 ml of an acid mixture made of concentrated acids HF:HClO4:HNO 3 at 6:2:3 ratio was added. The digestion continued until no coloured vapours were released. The atomic absorption spectrophotometer (Varian SpectrAA-200FS) was used for the determination concentrations of Lead (Pb) Mercury (Hg), Copper (Cu), Zinc (Zn), Cadmium (Cd), Chromium (Cr), Nickel (Ni) and Arsenic (As) and photometric determination used for Phosphorus (P).

Assessment of contamination in the sediments This was done using three parameters: enrichment factor (EF), geo-accumulation index (Igeo), and pollutio n load index (PLI).

a. Determination of heavy metal enrichment factor (EF)

Iron (Fe) was chosen as the element for normalisation of the enrichment because natural sources vastly dominate its input in the environment (Schiff & Weisberg, 1999; Ghrefat & Yusuf, 2006; Sekabira et al., 2010). To obtain the natural Fe concentration in the wetland sediments of Kumasi, three sediment samples were taken from different locations in the KNUST Botanical Gardens and pooled. The KNUST Botanical Gardens was chosen because of its remoteness from known or suspected anthropogenic metal sources. World geochemical background values in average shale as reported by Turekian & Wedepohl (1961) were used for all the metals. The EF for each metal was computed as the relative abundance of species in the sediment to that found in the Earth’s crust (Simex & Helz, 1981) with the formula: EF = () ơ Where X is concentration of the metal studied. (ℎ′ ͗ͩ)

Six categories of enrichment were recognised: <1 background concentration, 1-2 depletion to minimal enrichment, 2–5 moderate enrichment, 5–20 significant enrichment, 20–40 very high enrichment and >40 extremely high enrichment (Duzgoren-Aydin et al., 2006; Harikumar et al., 2010). b. Determination of index of geoaccumulation (Igeo)

The Igeo values were calculated for the different metals using the formula of Müller (1969) expressed as: Igeo = log2(Cn/1.5*Bn).

Where Cn is the measured concentration of element n in the sediment sample and Bn is the geochemical background for the element n. The factor 1.5 is introduced to include possible variation of the background values that are due to lithogenic variations. Müller (1969) proposed seven sediment quality classes associated with the geoaccumulation index as given in Table 5.1.

Table 5.1. Classes of the geoaccumulation index used to define sediment quality Igeo value Igeo Class Quality of sediment ≤ 0 0 Unpolluted 0-1 1 From unpolluted to moderately polluted 1-2 2 Moderately polluted 2-3 3 From moderately to strongly polluted 3-4 4 Strongly polluted 4-5 5 From strongly to extremely polluted ≥ 6 6 Extremely polluted

c. Determination of pollution load index (PLI)

The PLI is an indicator of site quality. For each metal, the contamination factor (CF) was first determined using the formula, CF = metal concentration in sediments/background values of the metal. The extent of pollution by trace metals was assessed using the PLI formula 1/n developed by Tomlinson et al. (1980): PLI = (CF1 * CF2 *.....* CFn) Where CF = contamination factor and n= number of metals. PLI value >1 is polluted whereas PLI value < 1 indicates no pollution.

5.2.2.3 Vegetation sampling To fulfil the requirements of the Ramsar Convention Secretariat (2005), the systematic sampling approach described by Orloci (1978) was used. At each selected wetland site, starting from the edge of the permanent water channel, 1 m x 1 m quadrats were marked along each transect. Sampling was done at 10 m intervals. All herbaceous plants and shrubs in the quadrats were identified to species level with their respective relative abundance and cover recorded (see Appendix 2) using the Braun-Blanquet scale described in Table 5.2. Cover was determined from estimates of vertical plant shoot-area projection as a percentage of quadrat area (Wikum & Shanholtzer, 1978).

The plants were identified with the assistance of a technician and the use of vegetation manuals by Akobundu & Agyakwa (1998), Cook (1996), Haflinger & Scholz (1980a, b), Haflinger et al. (1982) and Terry (1983). Plants that could not be identified on the field were collected and later identified in the laboratory. Around each wetland area, the various anthropogenic activities up to 100 m from the water channel were noted.

Table 5.2. The Braun-Blanquet scale and the description of the values used for sampling of wetland species in Kumasi Braun-Blanquet Range of cover Midpoint of over Relative abundance scale (%) range (%) 5 75-100 very common, usually single spp. 87.5 4 50-75 dominant 62.5 common, dominance is shared by 2 3 25-50 37.5 to 3 other species. frequent, a species not sharing 2 5-25 dominance but remains significant 15.0 to the composition of the site <5; numerous 1 occasional occurrence 2.5 individuals <5; few rare, individuals infrequently seen, + 0.1 individuals low count and crown cover

Sampling was done in June-July 2010 (peak rainy season) and December, 2010-January 2011 (peak dry season). The presence or absence of water on the surface at each sampling site was noted. Variations in species composition were estimated using the importanc e value index for all study sites using the following formulae:

= 0(- *! .+$. i. density - .(+' = 0(- *! +'*/. $) 2#$# .+$. *0-. ii. frequency */' )0(- *! +'*/. .(+' = */' *1- *! .+$. iii. dominance - .(+' = ).$/4 *! .+$. × 100 iv. relative density */' ).$/4 *! '' .+$. = -,0)4 1'0 !*- .+$. × 100 v. relative frequency */' *! '' !-,0)4 1'0. !*- '' .+$.

*($)) !*- .+$. × 100 vi. Relative dominanc e= */' *($)) !*- '' .+$.

Therefore, importance value index = relative density + relative dominance + relative frequency.

5.2.3 Data Analysis Differences among sampling sites along each stream and sampling periods were found by means of analysis of similarities (ANOSIM) test at α= 0.05. The various indices of diversity (Shannon, species evenness and species richness) were used to assess the plant community of each wetland. The non-metric multidimensional scaling (MDS), cluster analysis and principal components analysis (PCA) were used to explain species richness/distribution patterns and environmental parameters that influence/determine these patterns. The BIO-ENV procedure was used to examine the extent to which the soil texture, physicochemical water parameters and the total organic matter are related or explain (using the Spearman rank correlation coefficient ρ) the distribution and type of vegetation in each wetland. The Pearson correlation analysis was used to analyse and establish associations between heavy metals of wetland sediments and other variables after the data was transformed and normalised. A PCA was carried out on the correlation matrix to reduce the dimensionality of the problem and enable an interpretation of the relationships between variables and sampling sites.

5.3 RESULTS

5.3.1 Anthropogenic Activities in the Wetland Areas of Kumasi The commonest anthropogenic activities identified within 100 m from either side of the channel are houses, schools, churches, fuel/gas filling stations, farms, car washing bays and repair shops (Figure 5.2, Photograph 1 and Appendix 3). Activities with the highest land

cover are housing and urban agriculture (vegetable farms). The wetlands along the Sisai River from Family Chapel through Ahensan to Atonsu Sawmill (Sites A, B and E) have the highest anthropogenic activities associated with them. The predominant anthropogenic activities at these sites are houses, carpentry shops, car wash and repairs centres and waste disposal sites. The car repair centres are usually a cluster of wooden shelters serving as garages which do general repairs, engine oil change, vehicle body building and spraying. A lot of wastes from these activities are washed directly into the water channel. Natural vegetation cover in these sites is as low as 30 %.

All the sites have some vegetable farms associated with them. These are commonly small plots of about 0.25 ha with carrots, cabbage, onions, green pepper and okro. The water from the channel is used to irrigate the fields to provide an all year round cultivation. Fertilizers, weedicides, insecticides and pesticides are used to increase yields and quality. The farming activity at the wetland stretching from the Bekwai to Santasi roundabouts is on a relatively large scale. Crops cultivated here include maize, oil palm, pawpaw, banana and plantain. Only about 20 % of the 100 m buffer area is natural wetland vegetation. Houses, infrastructure and public gardens or space occupy over 75 % of the total wetland area.

Figure 5.2. Landcover and Landuse activities within 100 m from the stream channel at the different study sites in Kumasi

5.3.2 Wetland Edaphic Characteristics, Heavy Metals and Stream Physicochemical Characteristics The soils in Kumasi developed from two main primary materials (Obeng, 1971). The western and north-eastern to south-eastern parts of the city have soils that developed from Cape Coast granites to form the Kumasi-Asuansi/Nta-Ofin/Bomso-Asuansi compound associations (Figure 5.3). The wetlands found around the KNUST Police Station (D), Faculty of Renewable Natural Resources farms (C), High School junction (B), Kaase Guinness (H), the Family Chapel (E), Spetinpom (F) and Bekwai roundabout (J) belong to this soil category. The soils stretching from the central to south- western Kumasi developed from the lower Birimian phyllites, greywackes, schists and gneisses primary materials. The soil groups from these primary materials form the Bekwai- Nzima/Akumadan-Bekwai/Oda complex associations. The wetlands around Dakwadwom (K), Family Chapel (E), Golden Tulip Hotel (G) and Atonsu Sawmill (A) belong to this soil category.

The site specific site edaphic conditions are described in Table 5.3. The upper layers of the soils were generally dark brown and deeper layers were yellowish-reddish brown. Mottles were very common in wetlands like Atonsu and Spetinpom wetlands. The texture types were mainly loamy from forest ochrosols and lithosols. All the sites were waterlogged from a minimum 6 months to being permanently waterlogged. The wetland at the KNUST Police Station had the highest organic matter content of 34.5% and the whole length of the corer was easily inserted into the soil. At the Atonsu, Spetinpom, Golden Tulip Hotel and Kaase Guinness wetlands, the corer could hardly penetrate up to 20 cm and these sites also had comparatively higher proportion of sand and low organic carbon.

There were spatial and temporal variations in physicochemical parameters measured (Table 5.4). During the peak rainy season, all the study sites were weakly acidic with pH between 5.4 at Atonsu to 6.4 at the Golden Tulip Hotel wetlands. The TDS and conductivity were the most varied in the selected wetland sites. The conductivity of water was directly related to the TDS (conductivity ≈ 2 TDS) for all sites. The water at Atonsu and Kaase Guinness had relatively high TDS of 411 and 423 mg/l and salinity of 0.43 and 0.52 ‰ respectively. The water at the KNUST Police Station also recorded the least salinity. The channel bottom water temperature, dissolved oxygen were less varied between sites.

Figure 5.3. Soil map of the Kumasi showing study sites [sites are labelled in letters: Atonsu Sawmill (A), High school junction (B), FRNR farms (C), KNUST Police Station (D), Family Chapel (E), Spetinpom (F), Golden Tulip Hotel (G), Kaase Guinness (H), Bekwai roundabout (J) and Dakwadwom (K)] (Source: Obeng, 1971). These observations were not different for some parameters for the dry season. The conductivity was also approximately twice the TDS for all the wetlands. For this reason, only the TDS was used in the subsequent analysis to eliminate the compounding effects of using both parameters. During the dry season however, pH was generally alkaline with a range from an almost neutral 6.88 at the KNUST Police Station to 10.6 at site the High School Junction. The biggest seasonal differences were in dissolved oxygen concentration. For example, dissolved oxygen at sites Atonsu, High

School Junction, KNUST Police Station, Kaase Guinness, Bekwai roundabout and Dakwadwom declined from 4.20, 5.50, 3.30, 4.90, 4.50 and 4.30 mg/l in the rainy season to 0.45, 2.05, 0.62, 0.16, 1.48 and 0.75 ml respectively in the dry season. The non-parametric Wilcoxon Test was used to assess if there are any seasonal differences in the physicochemical parameters at the various wetland sites (Appendix 4). The null hypothesis of no seasonal differences was rejected for pH and dissolved oxygen (α=0.05, p-values = 0.005 and 0.009 respectively).

The concentrations of metals in the different wetland sites investigated are presented in Appendix 5. Heavy metal concentrations were generally highest in Atonsu Sawmill, Family Chapel and the High School Junction. The KNUST Police Station had relatively lowest concentrations in all heavy metals except total phosphorus. The association between these metals and some of the measured environmental parameters was assessed using the Pearson correlation coefficient (Figure 5.4 and Appendix 6). The test showed a significantly positive correlation between Hg, Ni, Zn, pH and Pb. Other clusters of significant correlations are clay, Cu and Cr; dissolved oxygen and salinity. The percentage of silt in the soil was negatively correlated to pH and Cr; dissolved oxygen and salinity. To reduce the dimensionality of variables that may be of significance in this study, a PCA was carried out on the data (Figure 5.5 and Appendix 7). The first two principal components accounted for 52.9 % of the variation. The major contributors to the first principal component axis are Pb, Zn, Ni, Hg and pH and those to the second principal component axis are temperature, salinity, dissolved oxygen and dissolved solids.

The extent of sediment contamination by these heavy metals was further evaluated using the EF,

Igeo and the PLI. All these indices were found to be significantly different between the seasons (Appendices 8, 9 and 10). The FRNR farms and Spetinpom sediments that did not have significant enrichment in any heavy metal in the rainy season whilst the KNUST Police Station had very high enrichment in only phosphorus (Figure 5.6). The Family Chapel, Atonsu Sawmill and High School Junction sediments had significant enrichment in all metals except cadmium. Sediments from Family Chapel had very high enrichment of P and Pb and that from Atonsu Sawmill also has very high enrichment of Pb. Heavy metal enrichment is higher for all sites in the dry season than the rainy season (Figure 5.7). Very high enrichment was recorded at various study sites for P, Cu, As and Pb.

According to the Igeo for both seasons (Figures 5.8 and 5.9), most heavy metals were accumulated in sediments of Family Chapel, Atonsu Sawmill and High School Junction. In the rainy season ( Figure 5.8), only Pb, Cu, P and Cr caused pollution class 4 (i.e. sites were strongly polluted by these metals). In the dry season (Figure 5.9), these sites were extremely polluted by these metals. Extreme 96

pollution was reported of P and (or) Pb in the sediments of the wetlands at Atonsu Sawmill, High School Junction, KNUST Police Station, Family Chapel and Dakwadwom.

The PLI was used for an overall assessment of the extent of pollution of each site from the cumulative contributions of all the substances studied. By this criterion, all the sites considered in this study are deemed polluted (Figure 5.10). Pollution levels in the dry season for all sites were higher than that of the rainy season. The wetland sites in order of increasing levels of pollution are FRNR farms, KNUST Police Station, Spetinpom, Dakwadwom, Golden Tulip Hotel, Bekwai Roundabout, Kaase-Guinness, Family Chapel, Atonsu Saw Mill and the High School Junction.

Wetland being used as rubbish dump Cattle grazing in a wetland

House in a wetland Vegetable farm in a wetland Photograph 1. Some uses of wetlands in the Kumasi Metropolis (Photograph taken by Author in 2011)

Table 5.3. Edaphic characteristics of wetland sites sampled in the Kumasi Metropolis Depth of Soil texture Designated Organic Location of Description of dominant soil form at 10 m from water insertion of the De scription re se arch Sand Silt Clay matter wetland channel core r at 10 m code (%) (%) (%) (%) from channel Atonsu Sawmill A Deep, brown, mottled soil underlain by gleyed clay. Top 28 19 40 41 Clay loam 19.85 soil is dark and loose. Permanently waterlogged High school B Shallow soil underlain by hardpan (ferricrete). The upper 10 62 12 26 Sandy clay 3.84 junction loose soil layer is dark brown and lower hard layer is loam brown with red bands and gleyed patches. Waterlogged for about 8 months FRNR farms C Shallow soil underlain by hardpan (ferricrete). Soil is 16 38 52 10 Loam 6.09 made up of dark and red portions with gleyed patches. Waterlogged for about 6 months KNUST police D Deep soft layer of organic matter, loose dark brown, light > 50 21 46 33 Clay loam 34.50 station weight and permanently waterlogged with deeper portions containing gleyed clay. Family Chapel E Deep, grey, mottled soils underlain by gleyed clay, 22 46 38 16 Loam 5.65 subject to permanent waterlogged conditions. Upper layer is loose and dark brown. Spetinpom F Shallow grey soils underlain by hardpan ferricrete. 10 42 30 28 Clay loam 14.03 Mottling is common in upper parts exposed to air by roots of plants. Waterlogged for about 6 months Golden Tulip G Red well drained shallow soils underlain by hardpan 15 50 32 18 Loam 6.23 Hotel ferricrete with clear evidence of periodic wetness and lateral flow of water in the soil profile. Waterlogged for about 6 months Kaase Guinness H Deep, grey, mottled soils underlain by gleyed clay, 17 48 30 22 Loam 7.87 subject to permanent waterlogged conditions. Waterlogged for about 7 months Bekwai J Shallow loose soils underlain by hardpan ferricrete with 20 47 32 21 Loam 5.24 roundabout clear evidence of frequent waterlogging and lateral flow of water in the soil profile. Waterlogged for about 8 months Dakwadwom K Deep, grey, mottled soils underlain by gleyed clay, 39 33 31 36 Clay loam 6.17 subject to permanent waterlogged conditions. Permanently waterlogged Table 5.4. Physicochemical characteristics of water running through the wetlands in Kumasi 98

Rainy season (June) Dry season (January) Designated Location of re se arch Temp Salinity DO TDS Conductivity Temp Salinity DO TDS Conductivity we tland pH pH code (°C) (ppt) (mg/l) (mg/l) (µS/cm) (°C) (ppt) (mg/l) (mg/l) (µS/cm) Atonsu Sawmill A 5.4 28.84 0.43 4.2 411 822 8.6 30.31 0.33 0.45 340 679

High School B 6.3 26.14 0.10 5.2 103 205 10.6 26.98 0.11 2.05 115 230 Junction FRNR farms C 6.3 25.68 0.11 5.6 120 239 7.05 24.36 0.08 3.94 86 172

KNUST Police D 5.8 26.30 0.10 3.2 110 220 6.88 26.87 0.15 0.62 164 340 Station Family Chapel E 5.9 27.88 0.24 3.3 252 505 7.06 27.36 0.25 4.54 257 512

Spetinpom F 5.9 27.88 0.24 3.3 252 505 7.3 26.88 0.13 3.6 144 288

Golden Tulip Hotel G 6.4 27.04 0.46 4.0 211 430 7.06 28.02 0.25 0.82 258 516

Kaase-Guinness H 6.2 26.51 0.52 4.9 423 805 8.32 31.77 0.52 0.16 537 1074

Bekwai roundabout J 6.1 27.12 0.26 4.5 320 630 7.47 38.13 0.29 1.48 297 394

Dakwadwom K 6.0 26.13 0.29 4.3 305 610 7.67 27.14 0.29 0.75 298 597

Figure 5.4. A cluster diagram of the correlation matrix of the various environmental variables measured in the different suburbs of Kumasi

100

Figure 5.5. A principal components analysis plot of the measured environmental parameters (The vector length reflects the importance of the variable’s contribution to the two principal components in relation to all possible axes and if a vector touches the circle, then none of that variable’s other coefficients in the Eigen vectors will differ from 0)

101

Figure 5.6. Heavy metals enrichment in the stream sediments of the different suburbs of Kumasi in the rainy season

102

Figure 5.7. Heavy metals enrichment in the stream sediments of the different suburbs of Kumasi in the dry season

103

Figure 5.8. Heavy metals accumulation in the stream sediments of the different suburbs of Kumasi in the rainy season 104

Figure 5.9. Heavy metals accumulation in the stream sediments of the different suburbs of Kumasi in the dry season

105

Figure 5.10. Seasonal differences in pollution by metals in the stream sediments of the different suburbs of Kumasi

106

5.3.3 Wetland Vegetation of Kumasi

5.3.3.1 Vegetation abundance and diversity in wetlands of Kumasi There were no seasonal differences (dry season and rainy season) in vegetation, so the data were pooled and analysed according to the wetland areas sampled. In all, 112 species were identified (Table 5.5). All the identified species are native vegetation of the forest region of Ghana except Limnocharis flava and Ceratophyllum demersum which are exotic and invasive. To assess the species’ individual contributions to the total wetland vegetation, the importance value index was used. Using this index, species which have high relative dominance, frequencies and densities in the various wetland areas have high importance value (see Photographs 2, 3 and 4). The top five species with the highest importance values for each wetland are marked in bold in Table 5.5. Some of the important species were Coix lacryma- jobi (found at KNUST Police Station and Spetinpom), Paspalum vaginatum (Atonsu Sawmill), Thalia geniculata (Atonsu Sawmill, High School Junction, FRNR farms, Family Chapel and Kaase Guinness), Thelypteris palustris (KNUST Police Station) and Typha australis (Atonsu Sawmill and Spetinpom). Most of these important species were found in almost all the wetlands sampled, albeit with relatively low importance values at some sites. Some species like Achyranthes aspera, Ipomoea carnea, and I. aquatica were found only at one site each but with very high importance values. Paspalum vaginatum, Brachiaria deflexa, Chromolaena odorata, Sporobolus pyramidalis and the Cyperaceae were ubiquitous and formed the dominant species in at least 2 or more wetlands as manifested by their relatively high total importance values. Species like Thelypteris palustris, Coix lacryma-jobi, Typha australis, Lemna minor and Thalia geniculata each contributed relatively very high to the total importance values but were present in few wetland areas whilst Bracharia lata, Chromolaena odorata, Paspalum vaginatum and the Cyperaceae had relatively lower importance values but contributed to total importance in all the wetlands sampled.

Different number of species were identified in the different wetlands sampled (Figure 5.11). The FRNR farms had the highest number of species (86) and Spetinpom wetland with the least number of species was. Comparing the various wetlands, the Shannon diversity and species richness were very much varied than the species evenness. Overall, the Dakwadwom, Golden Tulip Hotel and Bekwai roundabout wetlands had highest species richness and Shannon diversity while Spetinpom was assessed to be poor in these indices (Figure 5.12). Dakwadwom had the highest species richness and Shannon diversity of 6.09 and 3.53 respectively. 107

The longitudinal variability in species distribution from the water channel was also investigated for each wetland area. In general, species richness and Shannon diversity increased with increasing distance from the water channel (Figure 5.13). Sampling sites between 30 and 40 m from the water channel were generally highest in species richness and Shannon diversity. Except at Kaase Guinness, species evenness did not change much with increasing distance from the water channel. This site has been channelised and there were no plants growing in the water channel.

River channel at Family Chapel silted with vegetation and rubbish

Photograph 2. State of some of the wetlands in the Kumasi Metropolis (Photograph taken by Author in 2011)

108

Table 5.5. Species identified in the different wetlands of Kumasi and their importance values (Top 5 importance value species for each site are shown in bold) Sampling sites Total Bekwai importa Identified species Atonsu High School FRNR KNUST Police Family Spetin- Golden Tulip Kaase Dakwa- round- nce Sawmill Junction farms Station Chapel pom Hotel Guinness dwom about val ue Achyranthes aspera 16.3 9.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 26.2 Aerva lanata 0.0 5.9 5.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 11.3 Aeschynomene 0.0 0.0 2.0 2.2 0.0 0.0 0.0 0.0 2.1 3.5 9.8 crassicaulis Adenia lobata 0.0 2.9 4.2 0.0 4.8 0.0 2.2 2.8 0.0 2.9 19.7 Ageratum conyzoides 0.0 1.8 3.7 1.4 0.0 0.0 2.6 0.0 4.0 2.7 16.2 Alchornea cordifolia 0.0 5.6 5.2 0.0 0.0 0.0 0.0 0.0 2.0 0.0 12.7 Alchornea laxiflora 0.0 4.3 6.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 10.8 Alternanthera sessilis 0.0 1.0 0.0 0.0 5.5 6.8 0.0 0.0 0.0 1.3 14.6 Amaranthus hybridus 2.3 2.1 2.0 1.3 5.8 1.6 4.7 0.0 3.0 1.4 24.2 Amaranthus 1.9 7.6 2.0 2.5 8.9 5.8 4.3 0.0 3.1 1.0 37.2 spinosus Andropogon gayanus 2.7 3.2 2.0 2.3 3.0 6.2 3.2 3.0 1.3 2.5 29.4 Aspilia africana 0.0 0.0 2.8 1.2 0.0 0.0 4.6 0.0 4.8 5.8 19.2 Asystasia gigantica 4.3 4.6 5.5 1.8 0.0 0.0 2.7 0.0 5.0 2.4 26.3 Axonopus 3.9 3.5 0.0 0.0 6.0 0.0 0.0 0.0 0.0 0.0 13.3 compressus Bambusa vulgaris 0.0 0.0 3.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.8 Bidens pilosa 0.0 1.7 4.4 1.8 4.4 0.0 2.8 3.7 5.6 3.2 27.6 Boerhavia diffusa 2.5 3.7 1.9 0.0 4.6 0.0 3.7 3.8 4.8 3.1 28.0 Brachiaria lata 5.4 5.0 6.6 7.4 5.4 5.5 3.5 3.7 2.7 9.3 54.5 Brachiaria deflexa 18.2 0.0 5.2 7.4 0.0 0.0 5.9 6.4 6.5 11.8 61.4 Calopogonium 6.0 4.5 3.6 1.2 4.5 0.0 6.5 7.0 5.5 9.2 48.0 mucunoides Cassia mimosoides 3.3 1.6 3.8 2.0 0.0 0.0 3.2 4.6 2.0 3.0 23.6 Cassia occidentalis 6.5 5.0 4.2 0.0 0.0 0.0 3.0 1.6 0.0 0.0 20.3

109

Cassia siamea 5.7 3.9 0.0 0.0 4.6 0.0 6.2 0.0 2.8 0.0 23.2 Centrosema 3.9 3.8 2.7 8.2 0.0 9.0 0.0 5.9 5.0 4.6 43.2 pubescens Ceratophyllum 1.0 0.0 0.0 3.8 0.0 0.0 0.0 0.0 0.0 0.0 4.9 demersum Chromolaena 0.0 6.9 7.1 0.0 6.8 9.0 12.2 10.9 13.7 9.0 75.5 odorata Cleome rutidosperma 3.3 2.0 2.3 2.5 0.0 0.0 4.0 0.0 4.4 2.8 21.3 Coix lacryma-jobi 0.0 0.0 0.0 13.6 15.3 24.1 0.0 0.0 0.0 0.0 53.0 Colocasia esculenta 6.1 2.8 0.0 0.0 4.0 5.1 1.9 0.0 2.4 1.9 24.2 Combretum hispidum 0.0 0.0 3.8 0.0 0.0 1.4 0.0 0.0 0.0 0.0 5.3 Commelina 0.0 4.6 3.9 4.0 3.7 0.0 4.0 5.6 6.7 5.1 37.6 benghalensis Commelina erecta 0.0 0.0 0.0 0.0 0.0 0.0 2.7 0.0 2.3 13.5 18.5 Commelina 0.0 0.0 5.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.3 senegalensis Cucumis sativus 0.0 2.8 2.3 0.0 2.2 0.0 2.0 1.2 0.9 0.9 12.3 Cymbopogon citratus 0.0 0.0 2.2 1.0 0.0 0.0 1.7 0.0 2.1 1.7 8.7 Cynodon dactylon 11.6 2.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 13.8 Cyperus alternifolius 3.1 2.0 2.7 6.3 0.0 3.6 4.3 4.2 5.6 4.1 36.0 Cyperus cyperoides 2.1 2.9 3.0 0.0 0.0 0.0 6.6 5.9 6.8 4.6 31.7 Cyperus esculentus 3.2 3.1 5.4 7.9 8.2 3.0 7.8 6.0 4.6 5.5 54.7 Cyperus 2.1 1.6 1.7 4.1 4.9 5.0 7.8 6.8 4.2 1.8 40.0 longibracteatus Cyperus rotundus 3.9 2.8 4.1 4.5 4.4 0.0 5.6 5.3 2.0 5.8 38.2 Cyperus strigosus 5.0 0.0 1.7 4.7 0.0 0.0 4.6 4.8 3.4 5.5 29.8 Desmodium 0.0 8.0 3.1 0.0 0.0 0.0 3.1 1.8 4.0 2.2 22.2 adscendens Echinochloa colonum 2.5 1.2 0 0 0 0 0 0 1.8 0 5.5 Echinochloa 5.9 3.1 2.2 9.2 0.0 0.0 5.4 3.2 4.5 0.0 33.6 obtusiflora Eclipta alba 0.0 0.0 2.7 8.4 0.0 0.0 3.0 2.0 2.4 2.1 20.6 Eleusine indica 6.3 5.7 2.4 4.7 10.5 0.0 4.1 7.1 3.6 3.1 47.5

110

Euphorbia 0.0 0.0 2.7 1.3 5.1 8.8 4.5 4.7 5.5 5.7 38.1 heterophylla Fleurya aestuans 3.4 2.7 3.1 1.4 4.4 0.0 2.3 4.3 3.9 3.1 28.7 Glinus oppositifolius 0.0 0.0 2.0 3.3 0.0 0.0 0.0 0.0 0.0 0.0 5.3 Gomphrena 3.4 3.6 1.0 1.1 0.0 0.0 3.7 0.0 2.1 0.0 15.0 celosioides Heliotropium 6.1 9.3 0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 16.2 indicum Hydrolea glabra 0.0 0.0 0.0 7.7 0.0 0.0 0.0 0.0 0.0 0.0 7.7 Hyparrhenia rufa 3.6 3.3 2.4 2.5 2.7 4.1 3.1 4.2 3.1 2.3 31.2 Imperata cylindrica 3.1 3.7 3.4 7.0 6.0 3.7 2.2 2.8 2.2 6.2 40.4 Ipomoea aquatica 0.0 0.0 0.0 14.0 0.0 0.0 0.0 0.0 0.0 0.0 14.0 Ipomoea asarifolia 5.4 4.6 0.0 0.0 8.0 9.8 0.0 7.0 0.0 2.4 37.2 Ipomoea cairica 0.0 1.0 2.9 0.0 0.0 0.0 5.1 3.8 4.3 5.4 22.4 Ipomoea carnea 0.0 13.7 2.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 16.2 Ipomoea involucrata 0.0 2.3 2.7 5.3 5.8 6.7 7.0 4.6 4.0 3.6 41.7 Ipomoea mauritiana 0.0 3.6 0.9 1.2 2.3 0.0 2.3 2.4 7.3 4.0 24.0 Ipomoea sublobata 0.0 0.0 0.0 0.0 0.0 0.0 4.9 5.1 5.3 3.3 18.7 Justicia flava 0.0 0.0 3.3 1.4 3.9 0.0 3.3 2.5 3.0 3.1 20.5 Lemna minor 16.9 0.0 0.0 7.6 4.4 0.0 0.0 0.0 0.0 0.0 28.9 Leptochloa 3.5 2.8 3.7 3.5 0.0 0.0 4.4 5.4 3.5 5.0 31.8 caerulescens Limnocharis flava 1.7 3.0 1.8 6.3 0.0 0.0 0.0 0.0 0.0 0.0 12.7 Ludwigia abyssinica 3.0 1.9 2.9 3.2 4.6 6.3 4.7 5.0 3.1 3.5 38.0 Ludwigia decurrens 0.0 0.0 0.0 15.9 0.0 6.0 0.0 0.0 0.0 5.3 27.2 Luffa cylindrica 2.3 0.0 1.9 0.0 0.0 3.7 2.5 4.9 3.9 2.4 21.7 Malvastrum 0.0 3.5 4.3 2.9 0.0 0.0 5.0 4.4 3.7 3.6 27.4 coromandelianum Melanthera scandens 0.0 0.0 3.1 5.1 0.0 5.5 4.7 3.9 5.2 2.8 30.3 Millettia zenkeri 1.2 1.4 3.1 2.9 2.7 5.0 5.3 2.9 5.3 2.0 31.7 Mimosa pigra 0.0 0.0 6.9 1.0 0.0 0.0 0.0 0.0 0.0 5.0 12.8 Momordica charantia 3.1 3.1 2.9 0.7 2.3 4.4 4.5 6.8 2.4 3.6 33.7

111

Neptunia oleracea 1.7 2.3 3.1 0.0 3.3 3.3 2.7 3.1 2.2 0.0 21.6 Nymphaea lotus 0.0 0.0 0.0 3.4 0.0 0.0 0.0 0.0 0.0 0.0 3.4 Ocimum basilicum 2.8 4.2 2.3 0.0 2.3 5.5 2.4 3.2 1.9 2.3 26.9 Oldenlandia 0.0 0.0 3.6 4.7 0.0 0.0 0.0 1.5 1.8 3.2 15.0 corymbosa Oxalis corniculata 0.0 0.0 2.7 0.0 2.8 2.5 3.8 4.9 2.9 3.4 23.1 Palisota hirsuta 0.0 0.0 2.9 0.0 0.0 4.0 0.0 3.3 2.7 0.0 12.8 Panicum 0.0 0.0 5.8 2.6 8.6 7.9 4.1 6.2 5.7 6.2 47.1 dichotomiflorum Panicum maximum 3.6 4.8 3.3 3.7 8.9 9.2 5.7 6.1 2.5 5.4 53.2 Paspalum vaginatum 14.9 7.9 6.9 10.3 14.9 14.7 3.8 5.0 4.9 5.9 89.2 Pentodon pentandrus 0.0 0.0 0.0 6.6 0.0 0.0 0.0 0.0 0.0 0.0 6.6 Phyllanthus amarus 3.9 6.0 3.4 5.2 4.9 0.0 4.8 5.4 3.7 2.7 39.9 Physalis angulata 2.2 4.0 3.9 6.4 4.2 5.1 5.3 5.0 4.4 3.7 44.2 Pithecellobium dulce 0.0 4.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.6 Polygonum 4.4 6.7 0.0 0.0 0.0 5.4 0.0 0.0 0.0 0.0 16.4 lanigerum Polygonum 0.0 0.0 0.0 4.6 0.0 0.0 0.0 0.0 0.0 0.0 4.6 salicifolium Rhynchospora 6.6 0.0 3.9 0.0 0.0 9.0 7.3 13.2 8.4 6.7 55.1 corymbosa Ricinus communis 4.1 5.0 2.9 0.0 6.0 0.0 0.0 0.0 0.0 0.0 18.0 Rottboellia cochinchinensis 0.0 0.0 9.6 0.0 4.5 8.2 11.5 6.1 9.4 9.3 58.5 Schrankia 0.0 0.0 3.2 0.0 0.0 0.0 0.0 0.0 3.4 6.8 13.4 leptocarpus Scirpus cubensis 1.2 1.8 3.2 0 0 0 0 0 0 1 7.2 Setaria barbata 0.0 0.0 2.6 0.0 3.7 5.9 4.4 2.7 6.8 2.6 28.7 Setaria pallide-fusca 5.2 0.0 2.3 0.0 3.8 4.3 2.7 2.7 2.2 2.0 25.4 Sida acuta 2.0 2.9 2.2 3.0 2.8 3.5 3.6 3.6 2.3 1.0 26.8 Sida garckeana 2.9 3.2 2.4 0.0 5.1 3.1 4.0 4.1 2.4 2.8 30.0 Solanum nigrum 3.3 3.4 1.4 0.0 2.8 0.0 0.0 1.5 3.4 0.0 15.8 Solanum torvum 0.0 4.3 0.0 0.0 3.1 0.0 0.0 0.0 0.0 0.0 7.3

112

Sorghum 4.6 0.0 3.1 0.0 5.7 5.9 5.3 5.9 6.4 2.5 39.4 arundinaceum Spigelia anthelmia 0.0 2.2 1.8 0.0 1.4 3.0 3.4 1.7 2.2 1.9 17.6 Sporobolus pyramidalis 15.0 4.9 4.7 8.0 5.3 9.9 3.8 3.3 6.3 6.6 67.7 Synedrella nodiflora 1.1 0.0 2.0 1.0 3.9 4.9 4.4 4.0 3.2 4.1 28.7 Talinum triangulare 2.3 1.3 2.1 1.3 4.0 4.1 1.3 4.3 2.2 2.4 25.3 Thalia geniculata 20.4 16.3 20.9 0.0 20.6 0.0 0.0 19.1 0.0 0.0 97.2 Thelypteris palustris 0.0 0.0 2.5 27.4 0.0 0.0 0.0 0.0 0.0 0.0 29.9 Trianthema 2.2 2.0 1.4 0.0 2.6 0.0 1.8 0.0 1.2 2.6 13.8 portulacastrum Typha australis 0.0 13.0 0.0 0.0 0.0 24.0 0.0 0.0 0.0 0.0 37.0 Urena lobata 3.7 4.7 0.0 0.0 2.2 5.0 3.6 2.3 2.6 1.8 26.0 Vernonia cinerea 0.0 0.9 1.3 1.0 2.9 0.0 0.0 0.0 2.2 1.7 9.9 Vernonia colorata 0.0 0.0 0.0 0.0 0.0 1.8 0.8 0.0 1.3 2.7 6.6 Total number of species 63 73 86 62 57 46 70 64 78 76 112

113

Thelypteris palustris Ipomoea carnea

Luffa cylindrica Achyranthes aspera

Thalia geniculata Coix lacryma-jobi

Photograph 3. Some high importance value wetland species in the Kumasi Metropolis (Photographs taken by Author in 2011)

114

Centrocema pubescens Typha australis

Urena lobata Cyperus alternifolius

Ipomoea aquatica Lemna minor

Photograph 4. More high importance value wetland species in the Kumasi Metropolis (Photographs taken by Author in 2011)

115

Figure 5.11. Number of species identified at the various wetland sites in Kumasi

Figure 5.12. Species richness, evenness and Shannon diversity at the study sites in Kumasi

116

Figure 5.13. Species diversity indices with respect to distance from the water channel at the different sampling sites in Kumasi

117

5.3.3.2 Determinants of wetland vegetation distribution in Kumasi An analysis of similarities (ANOSIM) was carried out to assess the differences or similarities in species within and between wetlands. There were significant differences in species composition among the wetland sites with regards to the presence or absence of surface water at the sampling sites (Figure 5.14a). Also, species changed with increasing distance from the water channel (Figure 5.14b). The sites could therefore be separated hydrologically and the distribution of the species also varied according to distance from the water channel (Global R= 0.24 and 0.37 respectively, at 1 %).

The species found from the water channel up to 10 m away were significantly different from those found from 30 m and beyond. Typical species in these two categories were Ipomoea aquatica, Thalia geniculata, Nymphaea lotus, Typha australis, Ipomoea asarifolia, I. aquatica, I. carnea, Alchornea cordifolia, Thelypteris palustris, Coix lacryma-jobi, Ludwigia decurrens, Centrocema pubescens, Limnocharis flava, Hydrolea glabra in the water channel up to 10 m and Cynodon dactylon, Desmodium adscendens, Eleusine indica, Aspilia africana, Vernonia cinerea, Hyparrhenia rufa, Synedrella nodiflora, Trianthema portulacastrum, Setaria pallide-fusca, Malvastrum coromandelianum, Ipomoea sublobata, Euphorbia heterophylla, Oxalis corniculata, Paspalum vaginatum, Eleusine indica, Sporobolus pyramidalis, Brachiaria deflexa, B. lata, Cyperus longibracteatus, C. esculentus C. alternifolius from 30 m and beyond. All the species found within the 10 m distance from the water channel were also found in sampling sites that had water on the surface irrespective of the distance. Species found about 20 m from the water channel were not significantly different from those beyond this distance. Cyperus spp were present in various clusters in the different sampling sites and independent of distance from the water channel or the presence or absence of surface water.

Having found the sites to be significantly different from each other, a hierarchical cluster analysis was used to group the sites at the different levels of similarity using the vegetation data. The clustering was done according to the two environmental factors: presence or absence of surface water and distance from the main water channel. The sites showed 5 distinct clusters at 28 % similarity (Figure 5.15a and b). Inspecting the two subfigures, the sites showed a high tendency to cluster according to the presence or absence of surface water (Figure 5.15a open triangles separate from closed triangles). The clusters were, however, not distinct when the distance from the water channel was used as a factor for distinguishing the

118

Figure 5.14. Analysis of similarities of the community composition among wetland sites in Kumasi according to (a) presence or absence of surface water and (b) distance from the water channel

119

sampling sites (Figure 5.15b distances symbols show no consistent order). In an attempt to find underlying reasons for the observed clustering, multidimensional scaling (MDS) was used. The resultant MDS plot with a stress of 0.17 is shown in Figure 5.16. Apart from site H2 which in both the hierarchical cluster and MDS plot looked like an outlier, all the other sites stick together similar to the clusters above according to the hydrological conditions of presence or absence of surface water. The sites were also further distinguished by their measured environmental parameters using the MDS (Figure 5.17) and PCA (Figure 5.18). The MDS shows that the wetlands at Family Chapel, Golden Tulip Hotel, Bekwai roundabout and Dakwadwom are quiet similar to each other than to the others. From the PCA plots, wetlands at Family Chapel, Golden Tulip Hotel, Bekwai roundabout and Dakwadwom were different from the other the wetland sites in the measured environmental parameters. A total variability of 68.1% is attributed to the PC1 and PC2 (45.7 and 22.4 % respectively). Vectors with the strongest contributions to the first two axes mentioned above are the percentage organic carbon, sand, soil consistency, pH, temperature, TDS and salinity.

The assumption from the objective is that the suite of measured environmental parameters are responsible for structuring the spatial distribution of vegetation or the vice versa such that, an ordination based on the data would always group sites the same way. In order to match biotic to environmental patterns, the vectors with the strongest contributions to the first two axes in the PCA are considered in the BIO-ENV procedure. The combination of variables that best explained the distribution of the vegetation is percentage organic carbon, sand, temperature and salinity (in decreasing order of magnitude). Looking at the measures of these important variables in the various sites, the wetland at the KNUST Police Station has the highest amount of organic carbon, the lowest salinity, and comparatively low values for percentage sand and temperature. A further hierarchical step by step analysis of these best results identified percentage of organic carbon in the soil to be the strongest abiotic factor influencing vegetation distribution in the studied wetlands.

120

Figure 5.15. A hierarchical cluster of the community composition among wetland sites in Kumasi according to (a) presence or absence of surface water and (b) distance from the water channel (all samples standardised by totals and square root transformed)

121

Figure 5.16. A non-metric multidimensional scaling plot of the community composition among wetland sites in Kumasi using the presence or absence of surface water as a factor (All samples standardised by totals and square root transformed)

Figure 5.17. A non-metric multidimensional scaling plot of the wetland sites in Kumasi using the measured environmental parameters

122

Figure 5.18. A principal components analysis plot of the sampled wetland sites in Kumasi superimposed with vector plots (circle and lines) of the measured environmental parameters. The vector length reflects the importance of the variable’s contribution to the two principal components in relation to all possible axes and if a vector touches the circle, then none of that variable’s other coefficients in the Eigenvectors will differ from 0

5.4 DISCUSSION A review of literature on soils of Kumasi shows that they are generally forest ochrosols and lithosols (Obeng, 1971; Adjei-Gyapong & Asiamah, 2000). These soils may belong to the nitisols and acrisols in the World Reference Base for Soil Resources (Driessen et al., 2001). Although a detailed soil analysis was not undertaken to enable wetland specific soil classification, Adjei-Gyapong & Asiamah (2000) states that soils bordering most of the rivers and network of streams in Ghana are either recent alluvial deposits or presently influenced by 123

the floods of these drainage channels and therefore classified as alluvial soils. These alluvial soils may accumulate autochthonous materials from dying vegetation during the dry seasons when there is reduced flow and erosive power leading to deposition of organic material at some wetlands. According to the Driessen et al. (2001), histosols have soil material with more than 20 % organic matter by weight. Following this definition, soils of Atonsu Sawmill and KNUST Police Station may be considered histosols. The relatively higher percentage of organic carbon in wetlands of Atonsu Sawmill and KNUST Police Station could be due to the low levels pH, salinity/conductivity and dissolved oxygen which all favoured net deposition/accumulation of organic material.

The pH and dissolved oxygen were also significantly different between the seasons. These conditions may have caused a decrease in biotic activity in the sites and decreased decomposition of organic matter. An initial condition of a silty- clay soil might have facilitated the anoxic conditions leading to the accumulation of organic material. The major anthropogenic activity in these study sites that could have affected the soil conditions is agriculture. However, the samples were taken from points that were not under cultivation, at least not in the recent past. The depth of penetration of the corer was closely related to the soil textural conditions. It is therefore difficult to isolate soil influence on plant species distribution and hence accumulation of organic material. Root penetrability and anchorage are dependent on soil consistency and very important for plant growth (Kim, 2006). The direct relationship between the TDS and conductivity may be due to the type and amount of inorganic and ionized substances in the water. The quantities of these substances may be such that their contribution to conductivity and TDS for all the sites is similar. This direct relationship also exists with salinity. The range of values for the salinity also shows that all the study sites are freshwater environments.

By all indices used, Kumasi is said to be polluted by the heavy metals assessed in this study. This pollution also varies with the seasons. The seasonal differences in sediment contamination by heavy metals could also be related to the variations in dissolved oxygen and pH. This is because, heavy metals concentration in sediments depends not only on their inputs but also upon the ability of the environment to retain the metals. Some factors that determine heavy metal retention are the textural characteristics, organic matter contents, mineralogical composition and properties of the adsorbed compounds and depositional environment of the sediments (Chakravarty & Patgiri, 2009). The seasonal variation in dissolved oxygen and pH could therefore explain the positive correlation between Hg, Ni, Zn, Pb with pH and Cu and

124

Cr with dissolved oxygen and the respective strong seasonal differences in concentrations of these metals. These physicochemical conditions might be necessary for the high retention or adsorption of these metals in the sediments. The significantly positive correlation between these metals also demonstrates possible co- contamination from similar sources: auto mechanic shops (car spray and mechanic shops), car wash centres, pesticides and herbicides from vegetable farms, wood preservatives from waste wood and saw dust from the carpentry shops, waste from tie-dye industries and vehicle emissions.

If these metals presumably had the same sources, then differences in Igeo, EF and CF values may be due to the difference in the magnitude of input for each metal in the sediment and/or difference in the removal or retention rates of each metal from the sediment. This may be particularly significant due to soil differences and the rainy season when erosion washes away precipitated and adsorbed metals. Also, the changes in physicochemical conditions could lead to destabilisation of the adsorbed complexes leading to the measured lower concentrations in the rainy seasons. In general, when dissolved metals from natural or anthropogenic sources come in contact with saline water they quickly adsorb to particulate matter and are removed from the water column to bottom sediments (Schropp & Windom, 1988). This could explain the high metal concentrations in the sediments at Atonsu Sawmill which has relatively higher salinity. The Kaase Guinness wetland has been modified to a storm drain and the high salinity, even though might have increased adsorption, did not reflect in high metal concentration in the sediment because the concrete bottom of the water channel.

Conversely, no correlations were noted between Cd, P and As with the other heavy metals suggesting that contamination of these substances in Kumasi might be from different sources or had a different depositional nature. Cd concentrations were generally low in all sites. This might be from natural sources such as weathering rather than anthropogenic (Sekabira et al., 2010). The KNUST Police Station had conditions to potentially contain high concentrations of heavy metals: fine-grained inorganic particles of clay and natural organic substances which could have facilitated adsorption, chelation, complex formation or direct precipitation (Gibbs, 1977). Except for P, the low levels of other metals at this site could be due to the absence of anthropogenic activities that would have introduced such substances. The phosphorus could be from fertilizers washed from the surrounding agricultural activities and domestic wastewater.

Trace amounts of heavy metals are always present in fresh waters from terrigenous sources through weathering of rocks generally at relatively low concentrations. In recent times, due to 125

the combined effects of industrialisation/transport and urbanisation, heavy metal pollution of aquatic ecosystems is becoming a potential global problem. Ghana is fast developing along these lines, but like other developing countries, lacks mechanisms and tools to detect and monitor water quality. Whilst none of these water sources is being directly used for human consumption, the measured concentrations are so high that it could easily reach humans through the food chain. For example, these wetlands are used as grazing grounds for livestock and it will be very necessary to check the concentrations of these metals in the surrounding vegetation and animals grazing in such wetlands. The Ghana EPA has effluent quality guidelines for discharges into natural waterbodies. The car wash and repair garages do not follow these guidelines. Even with the maximum permissible levels for release of effluents given by the Ghana EPA, heavy metals have the tendency to accumulate in sediments and will still pollute streams as detected in Kumasi.

Typical of tropical environments, the study sites had very diverse vegetation as expressed by the Shannon diversity and species richness (Burslem et al., 2005). However, it is difficult to attribute these levels as a standard for describing wetlands in Kumasi because of the various forms of disturbances that may be specific to different areas. The inverted U-shaped pattern of species richness and Shannon diversity might be due to fewer plants adapted to the extremes and the 30 or 40 m zone providing a more ideal intermediate environment for most species. The intermediate disturbance hypothesis may hold true for these study sites and thus explain the pattern of species richness and Shannon diversity (Dial & Roughgarden, 1988).

Even with this inverted U-shaped pattern, species distribution followed specific longitudinal environmental gradients like flooded conditions (presence or absence of water on the surface). That is, some of the species are facultative whilst others obligate wetland species. The presence or absence of water on the surface may have caused the within wetland differences in species composition and distribution. Since the landscape is not completely flat, it is usual to find small high ground quadrat plot areas within a wetland that are not flooded and hence have different species cover. It is worthy to state that, even though no two quadrat plots can be exactly the same, the edaphic differences may not be so sharp between the quadrat plots to produce distinct differences in species along the distances that were considered for this study. Even though this study did not investigate obligate and facultative wetland species, the association of important value species to flooded sites defines the very unique niches that characterise such species in the environment. The sites of obligate high importance value species could be delineated to define the wetland microcosms for each study site.

126

The most popular definitions of wetlands have the key words hydrological conditions and vegetation adapted to the flooded conditions (Keddy, 2000; Maltby & Barker, 2009). Even though the cluster analysis shows that the presence or absence of surface water best grouped species (vegetation distribution), the species found under the different hydrologic conditions were different for the different wetland sites. What then causes these differences in species among wetlands in Kumasi? That is, per the definition of wetlands, places with similar hydrologic conditions in the same area should have similar vegetation. Therefore , other environmental conditions might be contributing to the species distribution because, the much used definitions of Keddy (2000) and Maltby & Barker (2009) are not adequate in describing small scale heterogeneity in the species distribution in Kumasi. The findings of this research confirm the difficulty to delineate wetlands from a landscape (Mitsch & Gosselink, 1993; Keddy, 2000; Haslam, 2003; Maltby & Barker, 2009). A wetland is a transition zone between aquatic and terrestrial environments. The limits and characteristics of this environment may therefore not be defined by only the hydrological characteristics.

Apart from hydrology, several studies show that wetland geomorphology, substratum, chemistry, adjacent terrestrial system, grazing, anthropogenic influences and individual plant tolerance or response to the stresses of flood and drought are significant (Harper et al., 1995; Barrat-Segretain et al., 1998; Owen, 1999; Casanova & Brock, 2000; Fortney et al., 2004; Maltby & Barker, 2009; Miller & Fujii, 2010). By virtue of the percentage organic carbon and other important physiochemical parameters identified by the BIO-ENV procedure, the wetland at KNUST Police Station can be described as a peatland (Joosten & Clarke, 2002). The peatland conditions might also have been the defining conditions for the growth and establishment of Thelypteris palustris not only because of its dominance there but that was the only site in which this species occurred. Other important species that also occurred in this site were Coix lacryma-jobi, Hydrolea glabra, Ipomoea aquatica, Ludwigia decurrens, Centrosema pubescens and Pentodon pentandrus. These species could also be said to be well adapted to peatland conditions. These wetlands being in urban settings, urbanisation could also have created perturbation gradients that affect aquatic plant species (Ot’ahel’ová et al., 2007). The landuse change due to urbanisation could have resulted in alterations in the hydrology, autochthonous and allochthonous inputs and water chemistry of wetland areas leading to changes in the local plant communities (Owen, 1999).

127

6 CLIMATE, LANDUSE CHANGES AND FLOODING IN THE KUMASI METROPOLIS 6.1 INTRODUCTION Floods are part of the ecosystem and a natural hydrological feature of most river systems. They provide benefits to humans through maintenance of ecosystem functioning such as sediment and nutrient inputs to renew soil fertility in floodplains and floodwaters to fish spawning and breeding sites (Avakyan & Polyushkin, 1989; Brinson, 1993; Casanova, 2000). However, this important natural phenomenon causes disasters because human settlements are getting increasingly closer to the floodplains or river channels such that communities are more vulnerable and unprepared for these floods. Recent floods in most areas are therefore preventable or anthropogenically induced and on a spatial and temporal increase (IPCC, 2007; Millennium Ecosystem Assessment, 2005). On the average, 140 million people are affected by floods each year, more than all other natural or technological disasters put together and in addition to human lives, floods cost more than $244 billion damage from 1990 to 1999, the most expensive of any single class of natural hazard (Millennium Ecosystem Assessment, 2005; IFRC, 2010).

Naturally, climate is dynamic. However, the world has experienced more anthropogenic induced climate changes over the last 100 years. Observed changes include more frequent and extreme weather events increasing the risk of flooding especially in urban areas (IPCC, 2007; IFRC, 2010). The extent to which urban centres are vulnerable to changes in climate is influenced by a variety of factors. However, this vulnerability could be mediated through social and economic interventions and the ability of stakeholders and institutions to address the challenges. Research shows that West Africa is among the most vulnerable regions to climate change worldwide (Niasse et al., 2004; Millennium Ecosystem Assessment, 2005; IPCC, 2007). Kumasi has experienced both landuse changes and floods in recent years. The relationship between these floods and the landuse changes, climate variability or change has not been investigated. This chapter therefore seeks to identify climate and landuse changes of Kumasi and assess their influences on the current incidents of floods in some parts of the city. The adaptations by the vulnerable suburbs to these annual floods are studied.

128

6.2 METHODOLOGY

6.2.1 Rainfall and Flood Data Collection and Analysis Monthly mean rainfall data on Kumasi from 1961 to 2010 were collected from the Meteorological Services Department, Accra, Ghana. Basic descriptive statistics were used to evaluate the rainfall data. The Mann–Kendall test (ZMK) was also used to determine the trends in annual and monthly rainfall and trends in rainy days (Rodrigo et al., 2001; Modarres & da Silva, 2007; Batisani & Yarnal, 2010). When testing either increasing or decreasing monotonic trends at a p significance level, the null hypothesis is rejected for absolute values of Z > Z1-p/2, obtained from standard cumulative distribution tables as 1.96. The standard significance levels of p = 0.05 is applied. Positive values of ZMK indicate increasing trends while negative ZMK indicate decreasing trends. The slope of the linear trend (as change per year) was estimated with the nonparametric Sen’s method Q (Gilbert, 1987). The relationship between total monthly and annual rainfall and number of rainy days (R≥1 mm/day) and heavy rainfall events (R≥31 mm/day derived from an estimate of the 95th percentile of the rainfall events of 1963 [chosen at random among first 3 years of Kumasi’s rainfall data]) was assessed using correlation. This was to assess if Kumasi’s rainfall is explained more by the total number of days with rain or by individual heavy rainfall events.

The data on flooding in Kumasi was obtained from the Ashanti Regional Office of the National Disaster Management Organisation (NADMO). Extra data on the frequency and duration of floods and extent of damage due to flooding was gathered through focus group discussions, interviews and questionnaires administered to the affected communities in Kumasi as described in Section 3.2.2. During the field visits and administration of questionnaires and interviews, effort was made to identify and understand practices in dealing with floods. This involved looking out for, noting and recording of behaviours, adaptations, practices and activities related to flooding and life in slum or informal environments.

6.2.2 Assessment of Urbanisation of Kumasi Urbanisation was assessed by determining the landcover change of Kumasi. Landsat images of Kumasi (at path 194, row 055) were collected from the landsat website www.landsat.org. In order to minimise problems related to vegetation phenology, only images acquired during the dry season were used. These images were captured on 11th January, 1986 and 13th January, 2007.

129

Various image enhancement and classification techniques were performed to help with the identification of the landcover types and changes using Leica Geosystems Erdas Imagine 9.2. A hybrid of supervised and unsupervised classification techniques was performed on the satellite images. The method was evaluated for accuracy using field verified landmarks with a GPS. For each study site, the spatio-temporal changes in vegetation cover were estimated. The final results of the classified image data and change thematic layer were further processed into maps in ARC-GIS 9.3 environment.

6.3 RESULTS

6.3.1 Rainfall Patterns and Trends in Kumasi

6.3.1.1 Rainfall pattern in Kumasi The annual rainfall pattern of Kumasi is weakly bimodal (Figure 6.1). The first and major rainy season starts from a relatively dry period in January with an average rainfall of 18.9 mm and peaks in June with an average of 211.7 mm. The second starts from August with an average rainfall of 87.7 mm and peaks at an average of 168.9 mm in September and drops to an average of 23 mm in December. The average annual rainfall of Kumasi was 1372.5 mm (Figure 6.2). Even though the highest average rainfall occurred in June, the highest monthly total rainfall ever recorded was 534.5 mm in September, 2007 and the lowest and highest annual rainfalls were 891.3 and 2345.1 mm in 1982 and 1968 respectively.

The descriptive statistics of the rainfall of Kumasi from 1961 to 2010 is presented in Table 6.1. Within these 50 years, there were 15 January months and 6 December months without rainfall. All the other months always had rainfall. The months with the highest variability in rainfall distribution were those between April and October. The rainfall distribution for these months was also characterised by relatively high coefficients of variation and mostly positive values for skewness and kurtosis. The rainfall distribution for August and September was significantly peaked over the normal distribution (kurtosis of 6.98 and 5.50 respectively). All the monthly rainfall distributions had positive skew values. That is, several extreme rainfall values lie to the right of the mean. Skewness for August was the highest (2.06).

130

250

200

150

100 Average rainfall (mm)

0 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC month

Figure 6.1. Average monthly rainfall pattern of Kumasi from 1961 to 2010

2600

2400

2200

2000

1800

1600

1400 1372.5 Annunal rainfall (mm) 1200

1000

800 1960 1970 1980 1990 2000 2010 year

Figure 6.2. Historical annual rainfall pattern of Kumasi from 1961 to 2010

131

Table 6.1. Descriptive statistics of monthly rainfall pattern of Kumasi from 1961 to 2010 Std. Mann– Sen’s Variable M inimum M aximum M ean Variance Skewness Kurtosis Deviation Kendall Z slope Q January 0 111 18.92 25.85 668.05 1.52 1.97 0.00 0.000 February 1 135 55.75 40.03 1602.44 0.43 -0.92 -1.31 -0.546 March 36 310 119.12 53.61 2874.49 1.21 2.31 -2.44 -1.200 April 45 311 146.32 64.54 4165.61 0.80 0.03 0.90 0.631 May 46 307 174.79 63.93 4086.88 0.43 -0.56 -1.15 -0.715 June 41 377 211.73 83.00 6889.14 0.26 -0.66 -0.47 -0.490 July 18 446 146.62 98.14 9630.74 1.10 1.20 -0.74 -0.562 August 7 400 87.70 69.71 4859.20 2.06 6.98 0.49 0.304 Sep tember 15 535 168.89 87.10 7586.15 1.80 5.50 0.00 -0.004 October 43 336 156.12 63.55 4038.46 0.77 0.73 -0.50 -0.378 November 3 213 62.38 47.33 2240.15 1.20 1.24 -2.90 -1.145 December 0 116 26.40 26.60 707.48 1.37 1.81 -0.26 -0.025

6.3.1.2 Rainfall trends in Kumasi From the results of the nonparametric Mann-Kendall (Z) and Sen’s slope (Q) methods, it was not possible to generalise either an increasing or a decreasing trend for the given set of data (Figures 6.3 to 6.15.). Rainfall for January and September showed no changes (Z=0, Q=0). Rainfall in April and August were increasing whilst that for February, May, June, July, October and December was decreasing. In both instances above, the trends were not significant. Only rainfall for March and November had decreased significantly over the period (Z= -2.44, Q= -1.20 and Z=-2.90, Q= -1.15 respectively at α=0.05). That is, the average rainfall for these two months was decreasing at 1.20 and 1.15 mm per year respectively. The annual total rainfall was also decreasing although not significant (Z= -0.89, Q= -0.59).

132

120

100

80 Data Sen's estimate 60 95 % conf. min 95 % conf. max 40 Residual

total rainfall (mm), Jan Z= 0.00 20 Q= 0.00

0 1960 1970 1980 1990 2000 2010

-20 Year

Figure 6.3. Trends in rainfall of Kumasi for the months of January from 1960 to 2010

150

100

Data Sen's estimate 50 95 % conf. min 95 % conf. max

0 Residual 1960 1970 1980 1990 2000 2010 Z= -1.31 total rainfall (mm), Feb Q= -0.55 -50

-100 Year

Figure 6.4. Trends in rainfall of Kumasi for the months of February from 1960 to 2010

133

350

300

250 Data 200 Sen's estimate 150 95 % conf. min 95 % conf. max 100 Residual 50

total rainfall (mm), Mar Z= -2.44* Q= -1.20 0 1960 1970 1980 1990 2000 2010 -50

-100 Year

Figure 6.5. Trends in rainfall of Kumasi for the months of March from 1960 to 2010

350

300

250

200 Data Sen's estimate 150 95 % conf. min 100 95 % conf. max

50 Residual Z= 0.90 total rainfall (mm), Apr 0 Q= 0.63 1960 1970 1980 1990 2000 2010 -50

-100

-150 Year

Figure 6.6. Trends in rainfall of Kumasi for the months of April from 1960 to 2010

134

350

300

250

200 Data Sen's estimate 150 95 % conf. min 100 95 % conf. max

50 Residual Z= -1.15 total rainfall (mm), May 0 1960 1970 1980 1990 2000 2010 Q= -0.72 -50

-100

-150 Year

Figure 6.7. Trends in rainfall of Kumasi for the months of May from 1960 to 2010

500

400

300 Data Sen's estimate 200 95 % conf. min 95 % conf. max 100 Residual

total rainfall (mm), Jun 0 Z= -0.47 1960 1970 1980 1990 2000 20100 Q= -0.49

-100

-200 Year

Figure 6.8. Trends in rainfall of Kumasi for the months of June from 1960 to 2010

135

500

400

300 Data Sen's estimate 200 95 % conf. min 95 % conf. max 100 Residual Z= -0.74 total rainfall (mm), Jul 0 Q= -056 1960 1970 1980 1990 2000 2010

-100

-200 Year

Figure 6.9. Trends in rainfall of Kumasi for the months of July from 1960 to 2010

500

400

Data 300 Sen's estimate 95 % conf. min 200 95 % conf. max Residual 100 Z= 0.49 total rainfall (mm), Aug Q= 0.30

0 1960 1970 1980 1990 2000 2010

-100 Year

Figure 6.10. Trends in rainfall of Kumasi for the months of August from 1960 to 2010

136

600

500

400 Data 300 Sen's estimate 95 % conf. min 200 95 % conf. max Residual 100

total rainfall (mm), Sept Z= 0.00 0 Q= 0.00 1960 1970 1980 1990 2000 2010 -100

-200 Year

Figure 6.11. Trends in rainfall of Kumasi for the months of September from 1960 to 2010

400

350

300

250 Data 200 Sen's estimate

150 95 % conf. min 95 % conf. max 100 Residual 50 Z= -0.50 total rainfall (mm), Oct 0 Q= -0.38 1960 1970 1980 1990 2000 2010 -50

-100

-150 Year

Figure 6.12. Trends in rainfall of Kumasi for the months of October from 1960 to 2010

137

250

200

150 Data Sen's estimate 100 95 % conf. min 95 % conf. max 50 Residual

total rainfall (mm), Nov Z= -2.90** 0 Q= -1.15 1960 1970 1980 1990 2000 2010

-50

-100 Year

Figure 6.13. Trends in rainfall of Kumasi for the months of November from 1960 to 2010

140

120

100 Data 80 Sen's estimate 60 95 % conf. min 95 % conf. max 40 Residual 20 Z= -0.26 total rainfall (mm), Dec Q= -0.03 0 1960 1970 1980 1990 2000 2010 -20

-40 Year

Figure 6.14. Trends in rainfall of Kumasi for the months of December from 1960 to 2010

138

2500

2000

1500 Data Sen's estimate 1000 95 % conf. min 95 % conf. max 500 Residual

total annual rainfall (mm)total 0 Z= -0.89 1960 1970 1980 1990 2000 2010 Q= -2.34

-500

-1000 Year

Figure 6.15. Trends in total annual rainfall of Kumasi from 1960 to 2010

6.3.2 Floods in Kumasi

6.3.2.1 Scope and frequency of floods in Kumasi The first rainy peak of the year (May, June and July) has always been associated with floods in some parts of Kumasi. For the purposes of this study, the definition of flood events by NADMO as occasions when water from e.g. rainfall events covered temporarily land outside the limits of the river banks, which is not normally covered, is adopted. Table 6.2 is an extract of rainfall events that have been reported to have caused floods and documented by the NADMO Office in Kumasi for June 2009. Except for the flood that occurred on 9th June, 2009, all others occurred during/after heavy rainfall events (R≥31 mm/day). None of the antecedent rainfall events was heavy. June and July were the months with the most reported incidence of floods. The flood experience of the residents from the suburbs that are being studied is shown in Figure 6.16. All respondents in Aboabo and Ahensan had had their homes flooded before. In the other suburbs, depending on the distance of a house to the water channel and length of stay of the respondent in the suburb, he/she might have experienced floods or not. In general, residents who have lived longer than 10 years in the various suburbs claim the floods are a recent phenomenon.

The relationship between rainy days, heavy rainfall events and the total monthly rainfall was assessed by correlation. The rainfall of Kumasi was observed to be influenced by both number

139

of rainy days and heavy rainfall events (Figure 6.17). However, total monthly rainfall was found to be explained more by heavy rainfall events than the number of rainy days in the month (85 % and 72 % of variance in common between heavy rainfall events and rainy days with total rainfall respectively). A trend analysis of heavy rainfall events is presented in Figure 6.18. Annual heavy rainfall events are increasing in trend, although not significant (Z= 0.45, Q=0.00). The heavy rainfall trends for the two flood prone months of June and July are varied. Whilst the heavy rainfall events are decreasing for June (Z= -0.78, Q= 0.00), that for July is increasing (Z= 0.91, Q= 0.00).

Table 6.2. Rainfall events and reported floods in some suburbs of Kumasi in 2009 Flood conditions Antecedent conditions Flood date Amount of Flood location Date of last Amount of rainfa ll (mm) rainfall rainfa ll (mm) 09/06/2009 27.8 Aboabo, Anloga, 08/06/2009 3.1 11/06/2009 60.4 Oforikrom, Ahensan, 10/06/2009 0.1 15/06/2009 43.3 Kwadaso Estates, 13/06/2009 19.1 26/06/2009 90.3 Dakwadwom, Sisaso, 25/06/2009 17.3 04/07/2009 53.4 Atonsu 03/07/2009 12.8 09/07/2009 72.7 08/07/2009 27.6

Figure 6.16. Respondents’ flood experience in the various suburbs of Kumasi

140

R2=0.72

R2=0.85

Figure 6.17. Relationship between different rainfall events and total monthly rainfall for Kumasi from 1961 to 2010

141

15 Data Sen's estimate 10 95 % conf. min 95 % conf. max 5 Residual

0 Z= 0.45 1960 1970 1980 1990 2000 2010 Q= 0.00 -5 total heavy rainfall events in theyear

-10 Year 10

6 Data Sen's estimate 4 95 % conf. min 95 % conf. max 2 Residual

0 Z= -0.78 1960 1970 1980 1990 2000 2010 Q= 0.00 heavy rainfall events in June -2

-4 Year 7

4 Data Sen's estimate 3 95 % conf. min 2 95 % conf. max 1 Residual 0 Z= -0.13 1960 1970 1980 1990 2000 2010 heavy rainfall eventsJuly in Q= 0.00 -1

-2

-3 Year

Figure 6.18. Trends in heavy rainfall events in Kumasi from 1961 to 2010

142

6.3.2.2 Adaptations to floods in Kumasi All the suburbs considered have experienced floods at one time or the other. The location of respondents’ houses influenced their estimation of the number of days the flood waters take to recede in the suburbs (Figure 6.19). The predominant number of days the people live under flood conditions was more than 20 days in a year for all the suburbs. For some suburbs like Ahensan, Atonsu and Dakwadwom, it took more than one month for the water to recede. The estimation of the loss was usually astronomical and varied depending on the interest and capacity of the person or institution making the estimate. These losses have however been increasing year after year.

Several coping strategies have been identified in the various suburbs (Figure 6.20). Each strategy is used in combination with the others. The most popular strategies are fetching or pumping out of flood waters and the creation of channels to facilitate the flow of the water. Respondents who do nothing in these flood situations (about 25% of the time) are usually the very young and the aged. Respondents of Kwadaso Estates and Spetinpom have relatively higher educational backgrounds and, almost 75% of the time, more likely to complain to city authoritie s.

There are also individual and community adaptations to floods (Figure 6.21). The action a resident took depended on previous experiences of flood and length of stay within the community. Most new rent paying tenants of these suburbs do not to know that the houses are flood prone. The first time experience of flood is usually that of shock and annoyance at the non-disclosure that the house is flood prone. Coping strategies of such persons to the flood are usually egocentric e.g. victims attempt to retrieve their rent (rent is usually paid 2 to 5 years in advance); move valuables and or relocate till the conditions improve; seal vents or build embankments around apartment (Photographs 5,6,7 and 8), complete relocation of property; and to complain to the city authorities. Most of these individual efforts often result in conflicts between the person and the landlord or neighbours. Some of the conflict issues were disagreements on the direction of a new drainage channel created by the individual; perception of individual not participating in communal efforts; or perception of the individual being proud. Residents who have lived in these suburbs for a while respond differently to the floods: they are more accommodating of the landlords and employ more communal adaptations to the floods e.g. participation in the construction of new drainage channels; (re)dredging of streams; landlords build embankments around houses; construction of networks of raised walkways to the various houses or compounds.

143

Figure 6.20. Adaptations of residents to floods in the various Figure 6.19. Recession time of water at the various flood prone suburbs of Kumasi suburbs of Kumasi

144

Suburb is flooded

individual community

New resident Old resident and homeowners

Move residence to new location Move things to higher grounds Retrieve their rent Construct drainage channels in the Relocate till water subsides neighbourhood First year Complain to authorities Dredge river/stream channel (Y) Fetch/pump water out of room or Build embankments e.g. with sand house bags or ridges Seal water ways and build Fetch/pump water out of room or embankments around room or house house raised walkways of block and wood Do nothing conflict

Moved out or started to behave like Construct drainage channels in the an old resident neighbourhood Dredge river/stream channel Next year Build embankments around house (Y+1) Build houses on stilts Improve on our houses Do nothing

Moved out or started to behave like Things will improve here an old resident God will take care of things. It will not always be bad Subsequent Construct bigger drainage channels years in the neighbourhood Dredge river/stream channel (Y+1+2+…n) Build higher embankments around house Do nothing

Figure 6.21. Flood response model of new and old residents of flood prone suburbs of Kumasi

In most of these flood prone suburbs, the chiefs and assemblymen harness local labour and capacity into building walkways, footbridges, drains etc. The public institutions usually provide some assistance in times of flood. NADMO may provide relief items depending on the availability of funds and number of people who have been affected. This is also the time that the KMA and NADMO mostly carry out their public education on floods. The major flood management strategy of the KMA is the engineering approach. For example, a storm drain is being constructed on the Aboabo River spanning through Anloga, Asokwa and Ahensan to Atonsu. In other places, the KMA does desilting, dredging and removal of structures within the waterways to improve flow of the water.

6.3.3 Landcover Change in Kumasi Landcover change was used as a proxy to assess urbanisation of Kumasi. To do this, it was assumed that there are two possible landcover types: the land is either covered with vegetation or built-up and bare areas. The landcover change from vegetated to built-up and bare areas is shown in Figure 6.22. In 1986, the land area currently under the jurisdiction of the KMA was about 80 % covered by vegetation. The remaining land area was either built up or bare e.g. public spaces, construction sites, roads, newly cleared farmlands etc. Land areas adjacent streams were largely vegetated (Schmidt, 2005; KMA, 2006). The major vegetated continuums in 1986 were in the eastern and the southern corridors stretching from Asokore Mampong to Deduako and the north western corridors from Ohwim to Tanoso. Typical built- up or bare areas were Central Kumasi stretching towards Suame, Airport, Atonsu, Santasi and Kwadaso. Built-up or bare areas beyond these suburbs were settlements located along the major routes leading out of Kumasi. By 2007, the vegetated areas within the KMA had reduced from 80 % to 50 %. The remaining major vegetation patches were at KNUST, Dakwadwom, Nyankyereniase and the south eastern land area beyond Emena. Vegetation cover along stream banks had also reduced. That is, the bare and built-up areas had been growing at about 1.9 % per annum between 1986 and 2007.

146

Figure 6.22. Landcover changes within the jurisdiction of the Kumasi Metropolitan Assembly between 1986 and 2007

147

Photograph 5. House being built on stilts in a flood prone suburb in the Kumasi Metropolis (Photograph taken by Author in 2011)

Photograph 6. Raised walkways in the compound of a flood prone house in the Kumasi Metropolis (Photograph taken by Author in 2011)

148

Photograph 7. Embankment built around a flood prone house in the Kumasi Metropolis (Photograph taken by Author in 2011)

Photograph 8. Embankment built around the main door to an apartment in a flood prone suburb of the Kumasi Metropolis (Photograph taken by Author in 2011)

149

6.4 DISCUSSION The seasonal rainfall cycle of Ghana shows a progressive shift from an all-year-round or two distinct rainy seasons from the coast to a single rainy season in the north. The results for Kumasi located around latitude 6°N show two distinct rainy seasons with a lower second peak. This is because the rainfall in Kumasi is dependent on convection following the movements of the ITCZ as the dominant mechanism producing rain (Barbé et al., 2002). After September, the ITCZ is receding and there is less penetration of moisture laden winds to cause rainfall, hence the weaker second peak. The high variations in rainfall of the rainy seasons may be due to inconsistencies in quantity and temporal distribution in the rainfall as shown by the skewness and kurtosis of the rainfall data. Positive skew values mean that the larger proportions of the rainfall for those months are a result of extreme rainfall events (Barbé et al., 2002; Modarres & da Silva, 2007; Batisani & Yarnal, 2010). The statistical results for July, August and September which are the peaks of the rainy season and floods in Kumasi are therefore not surprising and therefore due to heavy rainfall events as proven by the correlation analysis. Even though this analysis did not consider time between rainfall events, the antecedent conditions, as in date of last rainfall event, could have affected the probability of occurrence and duration of floods.

Apart from January and September, all months experienced a change in rainfall, albeit, most of them not significant. March and November are the only months with significant decrease in rainfall at a rate of 1.2 and 1.15 mm/annum respectively. According to Ghana’s Initial National Communication to the IPCC in 2000, Ghana’s annual rainfall was found to have decreased by 20% in the past 30 years (Ghana Environmental Protection Agency, 2001). Even though total rainfall for Kumasi may be decreasing (albeit, not significant), the change is less than the national figure reported by the EPA. However small it is, this change in can modify some elements of the water cycle in Kumasi: the frequency and intensity of precipitation, evapotranspiration, soil moisture, groundwater recharge and runoff (Fedeski & Gwilliam, 2007; Ramsar Convention Secretariat, 2008). Due to the relatively short period of the available data (50 years), it is difficult to draw conclusions on trends which may only end up being natural variability in rainfall. Because of this shortfall in the data, these results may therefore have to be validated by other climate models to draw stronger conclusions on climate change and rainfall trends in Kumasi.

The rate of growth of bare and built up areas in Kumasi could be a proxy measure of the rate of urbanisation. The built-up and bare areas within the KMA have been growing at a rate of

150

1.9 % per annum in the past two decades. The total population of Kumasi more than doubled within this period. The associated anthropogenic removal of vegetation and soil compaction, e.g. for housing and road constructions resulting from this growth, could decrease infiltration of rainwater and increase overland flow to cause the recent floods. It is now known that the most widespread serious potential impact of climate change on human settlements is flooding (Dazé, 2007; Fedeski & Gwilliam, 2007; IPCC, 2007; Wilby, 2007; International Recovery Platform, 2009; McClean, 2010). However, from the results obtained, flooding as being experienced in Kumasi is may not be the result of climate change in June and July but wrong urban spatial planning and management. The growth of Kumasi is not accompanied by development of sustainable urban drainage systems and waste management systems with enough system capacity or sophistication to avoid being overwhelmed by the increased overland flow. These are aggravated by multiple stresses and low adaptive capacities arising from endemic poverty and weak institutions (United Nations Inter-Agency Secretariat of the International Strategy for Disaster Reduction, 2005; Satterthwaite, 2003, 2008).

Even though the floods occurred during or after heavy rainfall events, the relatively long time it takes for the flood waters to recede may indicate that not all the floods are flash floods or due to heavy rainfall events. These suburbs are sited on lands, hitherto, considered as wetlands. This is evidenced by the prevalence and dominance of vegetation adapted to growing in wetlands (see section 5.3.3). The experiences of these flood prone neighbourhoods are, therefore, consequences of the spread of physical urban developments into wetlands. Surface compaction due to urbanisation may increase overland flow and produce flash floods as described by Griggs & Paris (1982) and the United Nations Inter-Agency Secretariat of the International Strategy for Disaster Reduction (2005). However, the rising of the water table during the rainy season (peaking in June) creates a network of effluent streams in large parts of these suburbs. The heavy rainfall events recorded each year trigger the flow and duration of these effluent streams. Effluent streams, rather than flash floods, best explain the delay of the flood waters in drying out from the neighbourhoods. Also, the dense networks of raised walkways and drainage channels could only be an adaptation to effluent stream flows rather than flash floods. From the above arguments, these events in those neighbourhoods should not be considered as floods contrary to the KMA’s definition of floods. This is because, even though the areas are inundated, the water is very much expected.

It is widely reported that a large and growing number of people are at high risk of adverse ecosystem changes because of the worsening trend of human suffering and economic losses

151

from floods (ActionAid International, 2006; Fedeski & Gwilliam, 2007; IPCC, 2007; Nelson et al., 2009). The ecosystem change as a result of the unmanaged urbanisation of Kumasi may increase the losses to floods. The scale of the risk, extent of vulnerability and losses from these flood events is much influenced and increased by the poor quality of housing and infrastructure in Kumasi, the lack of planning and enforcement of wetland management plans and the low level of preparedness among the city’s population and key emergency services. The flood prone suburbs have relatively higher human population and densities (Ghana Statistical Service, 2005) and higher demand for wetland services per person, therefore, making the human and ecological impacts of flooding more serious.

Also, these flood prone suburbs are predominantly inhabited by the urban poor and new rural- urban immigrants. Even though there is no spatial poverty profile of the residents of Kumasi, the physical location and characteristics of the poor are undeniably obvious in the housing landscape: an agglomeration of dilapidated structures without social infrastructure and amenities. It is however surprising that estimated economic losses from the floods keep rising year after year irrespective of the scale of the floods. Because the type and quality of houses identified in these suburbs are among the worst in the city, these reported increases in losses may be due to improvement in the economic conditions of these slum residents leading to increase in acquisition of precious and capital goods and services which may easily be destroyed by the floods. However, after living in such environments for several years, the residents adapt very much and losses are supposed to be very minimal except for some off- season large scale flash floods. Also, there is usually the perception and anticipation of help or a craving for public sympathy by these victims and therefore the tendency to exaggerate their losses. Either way it is difficult to independently verify these claims or accurately estimate the value of properties destroyed by floods in Kumasi.

In recent years, the notion of adaptation has become a significant component of climate change literature and international negotiations on climate (Niasse et al., 2004; World Meteorological Organisation, 2006; Intergovernmental Panel on Climate Change, 2007; Ireland, 2010). However, small scale adaptations by individuals and communities in informal settlements have not been given attention. What is reported in this study are individual and communal coping strategies rather than proactive and concerted efforts by city authorities at managing floods. After several years of stay, residents might have developed strong economic and social ties in the community and are therefore more reluctant to move out than the new community entrants. The choice to remain therefore transcends the push effect of annual

152

floods. A key consideration in the adaptation debate should therefore be on the small scale adaptation efforts and coping strategies of residents of mostly informal settlements who are most hit by floods. In addition, most of the flood victims have problematic relationships with the KMA because their houses have either been marked for demolition or they have no building permits. Yet the KMA is to act with all stakeholders to reduce these risks. In the light of the increasing urbanisation, the actions of the KMA must be to increase its efforts at not only re-engineering the city but also dialogue and enforcement of spatial plans and building regulations to reduce the proliferation of these informal settlements and preserve the demarcated buffer zones of streams within its jurisdiction.

153

7 GENERAL DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS 7.1 GENERAL DISCUSSION The state of the urban environment in Ghana is undeniably unacceptable. On the one hand, chiefs, landowners and local government agencies do not coordinate their efforts in spatial planning, and on the other, urban residents are overwhelmed with their own waste. The result is that waste is disposed of in any available “free” space, mostly wetlands. Urban wetland communities therefore bear the full brunt of filth and floods. Cleanliness may be deemed a virtue. However, this virtue needs to be reinforced and promoted through conscious efforts by all stakeholders. Whilst residents of Kumasi may be consciously interested in maintaining a clean environment, it behoves the responsible stakeholders to provide the much needed public services to enable residents carry out this civic responsibility. Waste disposal in streams and wetland areas could therefore be a thing of the past if the appropriate spatial planning and waste management approaches are implemented.

The consequences of urbanisation in developed and developing countries have been very much described. The KMA does not have to make all the mistakes that have been described or experienced by other cities elsewhere. There is still a window of opportunity for the KMA to avert this doom by harnessing the interest of the current King of Ashanti (Asantehene), Otumfuo Osei Tutu II, in its environmental management agenda. The Asantehene is very much interested in saving waterbodies. Because of the land tenure system and the authority the Asantehene wields in Kumasi, efforts to reverse the dire wetland situation could be made easy. The clearing of waterways through the demolition of houses has been vehemently resisted in most areas. Political interferences into these activities have sometimes made the situation more complicated. However, the involvement of the Asantehene in these activities could provide the much needed impetus that the KMA needs in carrying out their corrective environmental management and Garden City agenda.

The recent experiences of urban Kumasi make restoration to its former Garden City status not only desirable, but necessary. More and more people are experiencing floods and the results of this research show that, total annual rainfall of Kumasi is more from heavy rainfall events than rainy days. If this trend continues, the city has two options to curb the incidence of flash floods: (i) either adopt the engineering approach by increasing and expanding the drainage systems or (ii) improve the conditions of the wetlands to play its natural role of absorbing and

154

slowing down the storm waters. Whilst none of these options will single- handedly solve the problem, the economics of these options favour the latter. Urban storm drainage systems are expensive to construct and maintain. Improving wetland areas will not only curb the incidence of floods but is also an ecologically sound approach for a city that is growing and increasingly producing more pollutants.

This research has shown that the wetlands of Kumasi are a rich store of plant biodiversity. It is surprising that, a city that is undergoing such a rapid landcover change still has wetland patches that have maintained plant diversity comparable to undisturbed environments. That is, despite the pollution, encroachment and fragmentation of the wetlands, good fragments worthy of protection still exists. Whilst most of the wetlands have relatively high diversity indices, the importance value indices show that they are predominantly vegetated by 1-3 species. There seems to be a form of vegetation specialisation based on the soil and water regime of each wetland. Conservation objectives for these wetlands should therefore be for their ecological services rather than their store of biodiversity. For example carbon sequestration should be the conservation objective for the wetlands at the KNUST Police Station and Atonsu Sawmill; aesthetic values for Golden Tulip Hotel, Bekwai roundabout, Dakwadwom, Atonsu Sawmill; and FRNR Farms; and flood mitigation objectives for those at KNUST Police Station, Kaase Guinness. The diverse ecological potentials of these wetlands should be harnessed for sustainable urban environment of Kumasi. The success of the new Garden City of Kumasi will depend on how this rich diversity is integrated into the spatial planning efforts. Whilst the presence of heat island effect has not yet been determined in Kumasi, the increasing use of air conditioners, vehicles and the reduction in vegetation cover make it plausible. When these urban green spaces and streams are improved, they could mitigate the heat island effect and also provide places for relaxation to escape from the pressures of the city.

Poverty and vulnerability have always been used together in literature on slums and informal settlements. Even though these suburbs generally house the poor of Kumasi and are exposed to the annual floods, they cannot necessarily be described as vulnerable. Most of these residents have accepted the annual flooding as the status quo. Whilst anyone can be vulnerable to non-seasonal flash flood events, the prevalent form of inundation in these suburbs is due to effluent streams from the annual rise in the water table. This type of inundation by its nature and frequency is very much expected. The inhabitants may therefore be frequently exposed to floods but cannot be described as vulnerable because they expect it

155

and prepare for it. This may explain their reluctance to relocate. Poverty (expressed by their inability to afford the relatively high rental charges at other suburbs) may be the determinant in the initial choice to settle in that location. However, these residents soon tap into other forms of human capital (social, experiential, cultural, spiritual capitals) to reduce their vulnerability to the floods. These forms of capital need time to mature, improve and increase in value. Any efforts by the city authorities to relocate these wetland inhabitants should therefore involve the early identification and removal of factors that provide resilience to floods. These may, however, be deeply rooted in the social fabric and, therefore, difficult to remove by the time the inertia ridden city authorities decide to take action.

This study also shows that human environmental behaviour and choices are very fluid and unpredictable. Different people have different tolerance levels which are related to their individual value judgements. The Ghanaian society is still very much communal. Even in urban settings, people easily build relations in their neighbourhood. These bonds are Ghanaian social values that influence highly the outcomes and decision making processes of these urban residents. The trajectory of the EKC would apply in very individualistic societies where persons, or at most the household, take decisions independent of the influence of their neighbours. Where this independence is lost, the trajectories of flood tolerance or environmental pollution with increasing wealth deviates from what has been proven by Dinda (2004), Stern (2004) and Mills & Waite (2009). What is observed is that, after staying in these communities for a while, income gradually loses its power as a major decision making factor in residentia l choices.

Kumasi is growing faster than the inhabitants and planning authorities can cope with. Most inhabitants are still dependent on agriculture, natural resources and fresh foodstuff. With increasing urbanisation, these are drawn from increasingly distant environments to satisfy these inhabitants. The distance from which goods are acquired and to which waste can be sent is directly related to the spatial developmental stage of the city. Distances and costs to dispose of wastes are increasing for both the residents and waste management companies. The distance is also a function of energy (or financial resources). The more the availability of energy, the further these wastes could be transported or better processed into less harmful forms. Because of the high costs of managing wastes, streams serve as sources of “free energy” to transport the waste when it rains. Disposal of waste in wetlands and streams may therefore be an adaptation by urban residents to reduce this energy function in their lives.

156

These are conscious decisions residents take considering the cost of floods and aesthetic values of the environment being lost through such waste disposal methods.

Furthermore, the placing of waste containers in wetlands by the KMA could also be a reinforcing factor to the wrongful use of these wetlands as dump sites. The KMA sees these wetlands as the remaining islands of free public spaces for these purposes in an already unplanned and overdeveloped neighbourhood. In most of these poor suburbs, there is a high default rate of fee payment for collection or disposal of waste. Therefore, domestic waste is secretly deposited in the immediate surroundings of these containers. These are all washed into the surrounding streams choking them and increasing the siltation rates. These shallow rivers therefore easily overflow their banks during the heavy rain events flooding the neighbouring low lying areas. Therefore, whilst the KMA is working in one dimension to reduce the incidence of floods by removing structures along the waterways, they should also to be blamed, in part, for increasing the incidence of floods by siting waste bins in wetland areas.

7.2 CONCLUSIONS This study sought to contribute knowledge on the state of wetlands, incidence of flooding and stakeholder needs for the management of wetlands in Kumasi, Ghana. Data was collected in the chosen study sites and suburbs from January, 2010 to December 2011. After various analyses, the following conclusions were drawn in answer to the research questions and hypotheses that were set for the study: All the wetlands are riverine and associated with Rivers Suatem, Subin, Aboabo, Sisai and Wiwi which flow through the jurisdiction of the KMA. The size and quality of these wetlands are being decreased through adjacent anthropogenic activities.

Whilst there are some species common to all the wetlands, each wetland is unique in species composition. In each wetland, vegetation distribution followed a hydrological gradient. That is, the degree of soil water saturation was found to be the main determinant of species composition for each site (within wetland differences). However, no one obligate wetland species could be used to characterise all wetlands in Kumasi.

The percentage organic carbon was the strongest factor that explained the observed spatial heterogeneity of wetland vegetation in Kumasi (between wetland differences). Whilst there were seasonal differences in water conditions (pH and dissolved oxygen), there was no seasonal change in the vegetation. The wetland vegetation species diversity

157

was not found to be influenced by adjacent landuse but by the hydrology and edaphic conditions of the wetland. The hydrology and edaphic conditions at the KNUST Police Station make the wetland have the highest amount of carbon sequestered among the wetlands investigated. Typical species at the KNUST Police Station are Thelypteris palustris, Coix lacryma-jobi, Hydrolea glabra, Ipomoea aquatica, Ludwigia decurrens, Centrosema pubescens and Pentodon pentandrus.

All the wetlands were found to be polluted by the heavy metals investigated. The heavy metals Hg, Ni, Zn, and Pb probably come from the same source whilst Cu and Cr from another. Major anthropogenic activities that could be the sources of these heavy metals are the auto mechanic shops (car spray and repair shops), car wash centres, pesticides and herbicides from vegetable farms, wood preservatives from waste wood and saw dust from the carpentry shops, wastes from tie-dye industries and vehicle emissions.

The total annual rainfall is decreasing but not significant. Only rainfall for March and November is significantly decreasing over the period considered (Z= -2.44, Q= -1.20 and Z=-2.90, Q= -1.15 respectively at α=0.05). Annual heavy rainfall events are also decreasing in trend, although not significant. The analyses show that total monthly rainfall of Kumasi is explained more by heavy rainfall events than by the number of rainy days in the month. However, there is no significant change in trends of heavy rainfall events for the two flood prone months (June and July). Floods in Kumasi are therefore not attributable to changes in climate.

There are two types of floods that affect the wetland communities studied: flash floods and ‘effluent stream floods’. The flash floods arise from overland flow due to heavy rainfall events. The effluent stream floods are due to a rise in the water table during the peak rainy season of June-July causing a relatively longer surface inundation from groundwater.

The built-up and bare areas within the KMA have been growing at a rate of 1.9 % per annum between 1986 and 2007. This landcover change leads to increase in surface runoff because of reduced infiltration, decrease in concentration time and time to peak discharge and consequently increase in peak discharge. This is the major cause of the increased flash floods in Kumasi.

158

There was no relationship between number of years of stay in an area and presence or absence of floods (r = -0.135). The inhabitants are adapted to life in these flood prone neighbourhoods. Adaptations include: dredging of streams, erection of embankments around houses, move property to higher grounds, build a network of raised walkways and building houses on stilts. With these adaptations, residents plan to move out only when they are provided with an alternative house elsewhere. These are adaptations rather to the long lasting seasonal effluent floods than to flash floods.

The behaviour of the flood prone residents does not follow the EKC trajectory. Even when incomes of these residents increase, they do not move out, not even when they are forced. The social and experiential capitals developed by staying in these suburbs far outweigh the financial or economic losses due to the floods. The opportunity cost of moving out is therefore too high for these flood prone residents to bear. The duration of stay in a flood prone environment rather than income, is the stronger determinant in the decision making process of a resident to stay or relocate. Increased income will flatten the curve but will not bring flood tolerance to 0.

The stakeholders contacted for this study considered wetlands important and should be protected. The stakeholders most important for the management of floods and wetlands are chiefs and the KMA. The responsible authorities should manage wetlands and floods through the enforcement of landuse plans, demolition of houses on waterways and setting up public parks. The social, institutional and technical dimensions of flood and wetland manageme nt will achieve better outcomes for all the stakeholders.

7.3 RECOMMENDATIONS Based on the findings of this research, the following actions are recommended: The KMA should use the presence of effluent streams and the identified typical wetland important vegetation species for each wetland as defining factors to delineate the wetland boundaries. This exercise should particularly target the relatively large stretches of wetlands at the KNSUT Police Station, FRNR farms, Atonsu, Family Chapel, Dakwadwom, Spetinpom and Golden Tulip Hotel. After delineating the wetland boundaries, all structures within the wetlands should be demolished. These recovered wetlands should be given special names and revegetated to recover their natural system functions. In doing this, the KMA should liaise with the Asantehene (King of Ashanti) to solicit his support to reduce opposition.

159

The wetlands at the KNUST Police Station and Atonsu should be protected for their peat storage value. These sites could be further assessed for their potential to earn carbon credits for the city in the on-going climate change mitigation schemes. For example, the opportunities in the Reducing Emissions from Deforestation and forest Degradation (REDD+) programme being advocated by the United Nations could be tapped to facilitate the achievement of the Garden City status. This will also pave the way for Kumasi to meet the demands of the Millennium Cities Initiative and the Convention on Biological Diversity’s Global Partnership on Cities and Biodiversity.

Relevant academic institutions and faculty should develop and actively focus their research on urban ecology. The research should include the use of vegetation for urban waste water treatment, wetland vegetation for storm water management and carbon sequestration. Site specific rainfall input should be studied to assess the wetland hydrodynamics and its influence on temporal variation in vegetation. Also, scientific input from this research and others should be the backbone for the management of the delineated wetland areas to conserve biodiversity and provide ideal environments for human relaxation.

The KMA should endeavour not to site waste containers in wetlands to reduce the direct introduction of waste into stream channels. Where it is necessary to do so, a wall should be built around such a facility to contain any overflows. Also, waste collection in these poor neighbourhoods should be free so that residents are encouraged to use the designated dump sites. Anthropogenic activities around the wetlands should also be regulated to reduce the direct anthropogenic destruction and pollution of these habitats. In particular, car wash and repair shops should either be relocated or their waste be managed so that it is not directly introduced into the wetlands.

Considering the limited financial resources available to the KMA, an engineering based urban drainage system is not feasible. The approach to managing floods in Kumasi should be a combination of the reconstruction of the wetlands as described above and social interventions. The social flood management elements in each suburb should be identified and prioritised for action to increase resilience and reduce the scope and incidence of floods. In this regard, the KMA and NADMO need to recognise and give the needed support to chiefs and assemblymen of these neighbourhoods and relevant NGOs who are better placed to achieve these objectives.

160

The DEMC of the KMA should be strengthened to carry out their objectives and mandate. The powers vested in this Committee and its apolitical nature will make their actions more acceptable to the people. When operational, the DEMC should collaborate with the chiefs and assemblymen of the flood prone suburbs in enforcing community based flood and wetland management plans.

161

8 REFERENCES Abbaspour, M. & Gharagozlou, A. (2005). Urban planning using environmental modeling and GIS/RS: A Case Study from Tehran. Environmental Quality Management, 14(4): 73-85 ActionAid International (2006). Climate change, urban flooding and the rights of the urban poor in Africa- key findings from six African cities. ActionAid International, Hamlyn House, London N19 5PG, UK Adam, M. G., Sarfo-Mensah, P. & Warburton, H. (1999). Urban blight and farmers plight: peri -urban agriculture around Kumasi, Ghana. Land, 22: 131-148 Adams, W. M., Aveling, R., Brockington, D., Dickson, B., Elliott, J., Hutton, J., Roe, D., Vira B. & Wolmer, W. (2004). Biodiversity conservation and the eradication of poverty. Science, 306: 1146–1149 Adelekan, I. O. (2010). Vulnerability of poor urban coastal communities to flooding in Lagos, Nigeria. Environment and Urbanization, 22: 433-450 Adjei-Gyapong, T. & Asiamah, R. D. (2000). The interim Ghana soil classification system and its relation with the World Reference Base for Soil Resources. In: FAO (2000). Quatorzième Réunion du Sous-Comité ouest et centre Africain de corrélation des sols pour la mise en valeur des terres. Abomey, Bénin, 9-13 Octobre, 2000 Afrane, Y. A., Klinkenberg, E., Drechsel, P., Owusu-Daaku, K., Garms, R. & Kruppa, T. (2004). Does irrigated urban agriculture influence the transmission of malaria in the city of Kumasi, Ghana? Acta Tropica, 89(2): 125-34

African Peer Review Mechanism Secretariat (2005). Country Review Report of the Republic of Ghana. Midrand, South Africa. http://www.nepad.org

Akobundu, I. O. & Agyakwa, C. W. (1998). Handbook of West African Weeds. 2nd Ed. International Institute of Tropical Agriculture, Oyo Road, P.M.B. 5320, Ibadan, Nigeria

Akwei, E. (2010). Flooding in Aboabo Basin and adaptation: local responses to a global phenomenon. BSc thesis submitted to the Kwame Nkrumah University of Science and Technology, Ghana Altobelli, A., Bressan, E., Feoli, E., Ganis, P. & Martini, F. (2007). Improving knowledge of urban vegetation by applying GIS technology to existing databases. Applied Vegetation Science, 10: 203-210

Armar-Klemesu, M. & Maxwell, D. (1999). Accra: urban agriculture as an asset strategy, supplementing income and diets. In: Proceedings of International Workshop on Growing Cities Growing Food: Urban Agriculture on the Policy Agenda, La Habana, Cuba, October 1999

Avakyan, A. B. & Polyushkin, A. A. (1989). Flood control experience in the USA. Power Technology and Engineering, 23(1): 53-58 Baarta, I., Gschöpf, C., Blaschke, A. P., Preinera, S. & Heina, T. (2010). Prediction of potential macrophyte development in response to restoration measures in an urban riverine wetland. Aquatic Botany, 93(3): 153-162

162

Baharoglu, D. & Kessides, C. (2000). Urban Poverty. In: Poverty Reduction Strategy Papers - A Sourcebook. The World Bank Baker, J. & Schuler, N. (2004). Analyzing Urban Poverty: A summary of methods and approaches. World Bank Policy Research Working Paper 3399 Barbé, L. Le, Lebel, T. & Tapsoba, D. (2002). Rainfall variability in West Africa during the Years 1950–90. Journal of Climate, 15: 187- 202

Barbier, E. B. (1994). Valuing environmental functions: tropical wetlands. Land Economics, 70(2): 155-173

Barrat-Segretain, M. H. & Amoros, C. (1996). Recovery of riverine vegetation after experimental disturbance: a field test of the patch dynamics concept. Hydrobiologia, 321: 53- 68 Barrat-Segretain, M., Bornette, G. & Hering-Vilas-Boas, A. (1998). Comparative abilities of vegetative regeneration among aquatic plants growing in disturbed habitats. Aquatic Botany, 60: 201–211

Batisani, N. & Yarnal, B. (2010). Rainfall variability and trends in semi- arid Botswana: Implicatio ns for climate change adaptation policy. Applied Geography, 30: 483 –489 Beckerman, W. (1992). Economic growth and the environment: whose growth? Whose environment? World Development, 20: 481–496. Berkowitz, A. R., Nilon, C. H. & Hollweg, K. S. (Eds) (2003). Understanding Urban Ecosystems: A New Frontier for Science and Education. Sringer-Verlang New York, Inc. Beverwijk, A. (1967). Particle size analysis of soils by means of the hydrometer method. Sedimentary Geology, 1: 403-406 Bhaskar, N. R. & Singh, V. P. (1988). Planning flood control projects in urban areas. Water Resources Management, 2: 123- 140 Bhattacharya, P. (2002). Urbanisation in developing countries. Economic and Political Weekly, 37(41): 4219-4228 Biney, C. A. (1990). A review of some characteristics of freshwater and coastal ecosystems in Ghana. Hydrobiologia, 208: 45-53

Blake, B. & Kasanga, K. (1997). Kumasi Natural Resources Management Research Project Report. Natural Resources Institute. The University of Greenwich. UK and University of Science and Technology, Kumasi, Ghana Botchie, G., Akabzaa, T., Gyasi, E. & Sarpong, D. B. (2007). Land tenure, land use and environment in Ghana. ISSER Technical Publication Series 72. Legon, Ghana 78pp Botkin, D. B. & Beverage, C. E. (1997). Cities as environments. Urban Ecosystems, 1: 3–19

Brinson, M. M. (1993). A hydrogeomorphic classification for wetlands. Wetlands Research Program Technical Report WRP-DE-4. Army Engineer Waterways Experiment Station, Vicksburg, MS

163

Brody, S. D., Highfield, W. E., Ryu H. & Spanel-Weber, L. (2007). Examining the relationship between wetland alteration and watershed flooding in Texas and Florida. Natural Hazards, 40: 413–428

Brook, R. M. & Dávila, J. D. (eds.) (2000). The peri-urban interface: a tale of two cities. School of Agricultural and Forest Sciences, University of Wales and Development Planning Unit, University College London. 251 + vii pp. Brueckner, J. K. & Fansler, D. A. (1983). The economics of urban sprawl: theory and evidence on the spatial sizes of cities. The Review of Economics and Statistics, 65(3): 479- 482 Burslem, D. F. R. P., Pinard, M. A. & Hartley, S. E. (eds)(2005). Biotic Interactions in the Tropics: Their Role in the Maintenance of Species Diversity. Cambridge University Press, 580pp Capello, R., & Faggian, A. (2002). An economic-ecological model of urban growth and urban externalities: empirical evidence from Italy. Ecological Economics, 40: 181–198 Carter, J. L. & Vincent, H. R. (2001). After site selection and before data analysis: sampling, sorting, and laboratory procedures used in stream benthic macro invertebrate monitoring programs by USA State Agencies. Journal of the North American Benthological Society, 20(4): 658-682 Casanova, M. T. & Brock, M. A. (2000). How do depth, duration and frequency of flooding influence the establishment of wetland plant communities? Plant Ecology, 147: 237–250 Chakravarty, M. & Patgiri, A.D. (2009). Metal pollution assessment in sediments of the Dikrong River, N.E. India. Journal of Human Ecology, 27(1): 63-67

Chaneton, E. J. & Facelli, J. M. (1991). Disturbance effects on plant community diversity: spatial scales and dominance hierarchies. Vegetatio, 93: 143-155

Choi, J., Engel, B. A., Muthukrishnan, S. & Harbor, J. (2003). GIS based long term hydrologic impact evaluation for watershed urbanization. Journal of the American Water Resources Association, 39(3): 623-635 Cofie, O. O., Kranjac-Berisavljevic, G. & Drechsel, P. (2005). The use of human waste for peri-urban agriculture in northern Ghana. Renewable Agriculture and Food Systems, 20(2): 73–80

Cohen, B. (2004). Urban Growth in Developing Countries: A review of current trends and a caution regarding existing forecasts. World Development, 32: 23–51

Cohen, B. (2006). Urbanization in developing countries: current trends, future projections, and key challenges for sustainability. Technology in Society, 28: 63–80 Cook, C. D. K. (1996). Aquatic Plant Book. 2nd, revised edition. Amsterdam: SPB Academic Publishing Cooper, P. (2009). What can we learn from old wetlands? Lessons that have been learned and some that may have been forgotten over the past 20 years. Desalination, 24: 11-26 Cornish, G. A. & Kielen, N. C. (2003). Wastewater irrigation – hazard or lifeline? Empirical results from Nairobi, Kenya and Kumasi, Ghana. In: Scott C.A. Faruqui, N.I. and L. Raschid- 164

Sally (eds) Wastewater Use in Irrigated Agriculture: Coordinating the Livelihood and Enviro nme nta l Realities. IWMI, IDRC Coroi, M., Skefﬁngton, M. S., Giller, P., Smitha, C., Gormally, M. & O’Donovand, G. (2004). Vegetation diversity and stand structure in streamside forests in the south of Ireland. Forest Ecology and Management, 202: 39–57

Costantini, V. & Monni, S. (2008). Environment, human development and economic growth. Ecological Economics, 64: 867 – 880 Cropper, M. & Griffith, C. (1994). The interaction of population and growth and environmental quality. The American Economic Review, 84(2): 250–254 Czech, B. (2003). Technological progress and biodiversity conservation: a dollar spent, a dollar burned. Conservation Biology, 17: 1455–1457 Danso, G., Drechsel, P., Wiafe-Antwi, T. & Gyiele L. (2002). Income of farming systems around Kumasi, Ghana. Urban Agriculture Magazine, 7: 5–6 Danso. G., Drechsel, P., Wiafe-Antwi, T. & Gyiele, L. (2006). Income generated by farming systems in and around Kumasi. In: van Veenhuizen, R. (Ed.). Cities Farming for the Future: Urban Agriculture for Green and Productive cities. RUAF Foundation, IDRC and IIRR. p. 205-207 Davey, K. J. (1993). Elements of Urban Management. World Bank

Dazé, A. (2007). Climate Change and Poverty in Ghana. CARE Internatio na l. 20pp.

DeKeyser, E. S., Kirby, D. R. & Ell, M. J. (2003). An index of plant community integrity: development of the methodology for assessing prairie wetland plant communities. Ecological Indicators, 3: 119–133 Devas, N. (2005). Metropolitan governance and urban poverty. Public Administration and Development, 25: 351–361 Dial, R. & Roughgarden, J. (1988). Theory of marine communities: the intermediate disturbance hypothesis. Ecology, 79: 1412–1424 Dinda, S. (2004). Environmental Kuznets Curve Hypothesis: A Survey. Ecological Economics, 49: 431– 455

Drechsel, P. (2006). Learning from rapid urban growth and reflection in Kumasi. Case study 25: Ghana. In: Hague, C. et al. (Eds.). Making Planning Work. A guide to approaches and skills. Intermedia te Technology Publishing, p. 74- 75 Driessen, P., Deckers, J., Spaargaren, O. & Nachtergaele, F. (Eds.) (2001). Lecture notes on the major soils of the world. Natural Resources and the Environment. World soil resources reports series 94. ISBN 925-104637-9

Duzgoren-Aydin, N. S., Wong, C. S. C., Aydin, A., Song, Z., You, M. & Li, X. D. (2006). Heavy metal contamination and distribution in the urban environment of Guangzhou, SE China. Environmental Geochemistry and Health, 28: 375-391

165

Egyir, S. N. (2008). Use of digital elevation models for watershed management in Aboabo, Kumasi. BSc thesis submitted to the Kwame Nkrumah University of Science and Technology, Ghana

Egziabher, A. G., Lee-Smith, D., Maxwell, D. G., Memon, P. A., Mougeot, L. J. A. & Sawio, C. J. (1994). Cities feeding people: an examination of urban agriculture in East Africa. Ottawa, ON, IDRC,. xiv + 146 p.: ill. Ehrenfeld, G. J. (2000). Evaluating wetlands within an urban context. Urban Ecosystems, 4: 69–85 Eppinka, F. V., van den Bergh, C. J. M. J. & Rietveld, P. (2004). Modelling biodiversity and land use: urban growth, agriculture and nature in a wetland area. Ecological Economics, 51: 201– 216

Erenstein, O., Oswald, A. & Mahaman, M. (2006). Determinants of lowland use close to urban markets along an agro-ecological gradient in West Africa. Agriculture, Ecosystems & Environment, 117(2-3): 205-217 Esty, D. C., Levy, M. A., Srebotnjak, T. & de Sherbinin A. (2005). 2005 Environmental Sustainability Index: Benchmarking National Environmental Stewardship. New Haven, Conn.: Yale Center for Environmental Law & Policy Faithful, J. & Finlayson, W. (2005). Water quality assessment for sustainable agriculture in the Wet Tropics - a community-assisted approach. Marine Pollution Bulletin, 51(1-4): 99-112 FAO (1985). Review of the state of aquatic pollution of West and Central African inland waters Series. CIFA Occasional Paper-CIFA/OP12, 32pp. FAO (1998). Wetland characterization and classification for sustainable agricultural development. Food and Agriculture Organization of the United Nations (FAO), Sub-Regional Office for East and Southern Africa (SAFR)

FAO (2007). Profitability and sustainability of urban and peri- urban agriculture. Agricultural Management, Marketing and Finance Occasional Paper 19. ISBN 978-92-5-105881-7. FAO, Rome Federal Interagency Committee for Wetlands Delineation (1989). Federal Manual for identifying and delineating jurisdictional wetlands. US Government Printing Office, Washington, DC, 76pp.

Federal Register (1980). Guidelines for Specification of Disposal Sites for Dredged or Fill Material". U.S. Government Printing Office, Washington, DC

Fedeski, M. & Gwilliam, J. (2007). Urban sustainability in the presence of flood and geological hazards: The development of a GIS - based vulnerability and risk assessment methodology. Landscape and Urban Planning, 83: 50–61 Finlayson, C. M., Davidson, N. C., Spiers, A. G. & Stevenson, N. J. (1999). Global wetland inventory–current status and future priorities. Marine & Freshwater Research, 50(8): 717–727 Fobil, J. N. & Atuguba, R. A. (2004). Ghana: changing urban environmental ills in slum communities. International Journal of Environmental Policy and Law, 34(4-5): 206-215

166

Foeken, D. (2006). “To Subsidise My Income”: Urban farming in an East African Town. Koninklijke Brill NV, Leiden, The Netherlands. xvi + 224 pp. Foeken, D. W. J. & Owuor, S O. (2008). Farming as a livelihood source for the urban poor of Nakuru, Kenya. Geoforum, 39(6): 1978-1990 Fortney, R. H. Benedict, M., Gottgens, J. F, Walters, T. L., Leady B. S. & Rentch, J. (2004). Aquatic plant community composition and distribution along an inundation gradient at two ecologically-distinct sites in the Pantanal region of Brazil. Wetlands Ecology and Management, 12: 575–585 Fritioff, Å. (2005). Metal accumulation by plants - evaluation of the use of plants in stormwater treatment. Dissertation for the Degree of Doctor of Philosophy in Plant physiology presented at Stockholm University September 16, 2005. Stockholm University, ISBN 91-7155-111-5 pp 1-56 Ghana Environmental Protection Agency (2001). Initial national communication to the United Nations Framework Convention on Climate Change (UNFCCC). EPA, Accra Ghana News Agency (2010). Ghana to get radical on preservation of wetlands. Ghanaweb Article 176195. Thursday, 4 February 2010

Ghana Statistical Service (2002a). 2000 Population and Housing Census: Summary Report of Final Results. Accra, Ghana Ghana Statistical Service (2002b). 2000 Population and Housing Census: Special Report on 20 Largest Localities. Accra, Ghana

Ghana Statistical Service (2005). 2000 Population and Housing Census: Ashanti Region Analysis of District Data and implications for Planning. Accra, Ghana Ghrefat, H. & Yusuf, N. (2006). Assessing Mn, Fe, Cu, Zn, and Cd pollution in bottom sediments of Wadi Al-Arab Dam, Jordan. Chemosphere, 65: 2114–2121 Gibbs, R. J. (1977). Transport phases of transition metals in the Amazon and Yukon Rivers. Geological Society of America Bulletin, 88: 829–843 Gilbert, R.O. (1987). Statistical methods for environmental pollution monitoring. Van Nostrand Reinhold, New York Government of Ghana (1992). The Constitution of the Republic of Ghana. Asempa Publisher s, Accra Government of Ghana (1993). Republic of Ghana Act 462 - Local Government Act of 1993. Asempa Publishers, Accra

Government of Ghana (1996). Republic of Ghana Act 522 - Water Resources Commission Act, 1996. Asempa Publishers, Accra

Green, M. B. & Upton, J. (1994). Constructed reed beds: Appropriate technology for small communities. Wetland Systems for Water Pollution Control, 32(3): 339-348

Griebenow, G. & Kishore, S. (2009). Mainstreaming Environment and Climate Change in the Implementation of Poverty Reduction Strategies. The International Bank for Reconstruction and Development/The World Bank, Washington, D.C. 20433, USA

167

Griggs, G. B. & Paris, L. (1982). Flood control failure: San Lorenzo River, California. Environmental Management, 6(5): 407-419 Grimm, N. B., Faeth, S. H., Golubiewski, N. E., Redman, C. L., Wu, J., Bai, X. & Briggs J. M. (2008). Global change and the ecology of cities. Science, 319 (5864): 756-760 Gunter, B. G., Cohen M. J. & Lofgren, H. (2005). Analysing Macro-Poverty Linkages: An Overview. Development Policy Review, 23 (3): 243-265

Haflinger, E. & Scholz, H., (1980a). Grassweeds 1. CIBA-GEIGY Ltd, Basle, Switzerland. Haflinger, E. & Scholz, H., (1980b). Grassweeds 2. CIBA-GEIGY Ltd, Basle, Switzerland

Haflinger, E., Kuhn, U., Hamet-Ahti, L., Cook, C. D. K., Faden, R. & Speta, F. (1982). Monocot Weeds 3. CIBA-GEIGY Ltd, Basle, Switzerland Halls, A. J. (ed.) (1997). Wetlands, Biodiversity and the Ramsar Convention: The Role of the Convention on Wetlands in the Conservation and Wise Use of Biodiversity. Ramsar Convention Bureau, Gland, Switzerland

Harikumar, P. S., Prajitha, K. & Silpa, S. (2010). Assessment of heavy metal contamination in the sediments of a river draining into a Ramsar site in the Indian sub-continent. Journal of Advanced Laboratory Research in Biology, 1(2): 157-169 Harper, D. M., Adams, C. & Mavuti, K. (1995). The aquatic plant communities of the Lake Naivasha wetland, Kenya: pattern, dynamics and conservation. Wetlands Ecology and Management, 3(2): 111-123

Harvey, D. (1985a). Consciousness and the urban experience: studies in the history and theory of capitalist urbanization. Johns Hopkins University Press, Baltimore, Md. xix, 293pp.

Harvey, D. (1985b). The urban experience. Baltimore, MD: The John Hopkins University Press Harvey, K. R. & Hill, G. J. E. (2001). Vegetation mapping of a tropical freshwater swamp in the Northern Territory, Australia: a comparison of aerial photography, Landsat TM and SPOT satellite imagery. International Journal of Remote Sensing, 22(15): 2911–2925 Haslam, S. M. (2003). Understanding Wetlands: Fen, Bog and Marsh. London, Taylor and Francis Group

Hasse, J. E., Lathrop, R. G. (2003). Land resource impact indicators of urban sprawl. Applied Geography, 23: 159–175

Hejcmanovā-Nežerková, P. & Hejcman, M. (2006). A canonical correspondence analysis (CCA) of the vegetation–environment relationships in Sudanese savannah, Senegal. South African Journal of Botany, 72: 256 – 262 Hidalgo, J., Masson, V., Baklanov, A., Pigeon, G. & Gimeno, L. (2008). Advances in urban climate modelling. Annals of the New York Academy of Sciences, 1146: 354-74 Holland, M. D., Kasanga, R. K., Lewcock, C. P. & Warburton, H. J. (1996). Peri-urban baseline studies, Kumasi, Ashanti region, Ghana. Final Technical Report for project R6448. Chatham: NRI

168

Huang, S., Taniguchi, M. Yamano, M. & Wang, C. (2009). Detecting urbanization effects on surface and subsurface thermal environment — a case study of Osaka. Science of the Total Environment, 407(9): 3142-3152

Intergovernmental Panel on Climate Change (IPCC) (2007). Climate Change 2007. Fourth Assessment Report. Cambridge University Press, Cambridge, UK

International Recovery Platform (2009). Environment Issues in Recovery. http://www.gdrc.org/uem/disasters/disenvi/environment-recovery.pdf (accessed 22/06/2010) Islam, Md. A., Thenkabail, P. S., Kulawardana, R. W., Alankara, R., Gunasinghe, S., Edussriya, C., & Gunawardana, A. (2008). Semi-automated Methods for mapping wetlands using Landsat ETM+ and SRTM data. International Journal of Remote Sensing, 29 (24): 7077 - 7106

Jacobs, G., Macfarlane, R. & Asokan, N. (1997). Comprehensive Theory of Social Development. International. Center for Peace and Development, 2352 Stonehouse Drive, Napa, CA 94558, USA Janssen, R., Goosena, H., Verhoevenb, M. L., Verhoevenb, J. T. A., Omtzigta, A. Q. A. & Maltby, E. (2005). Decision support for integrated wetland management. Environmental Modelling and Software, 30: 215-229 Jenerettea, G. D., Marussicha, W. A. & Newell, J. P. (2006). Linking ecological footprints with ecosystem valuation in the provisioning of urban freshwater. Ecological Economics, 59: 38 – 47 Joosten, H. & Clarke, D. (2002). Wise Use of Mires and Peatlands- Background and Principles Including a Framework for Decision-making. International Mire Conservation group and International Peat Society. 304pp. Kaplan, D., Wheeler, J. & Holloway, S. (2009). Urban Geography (2nd Edition). John Wiley & Sons, Inc. NJ, USA Keddy, P. A. (2000). Wetland Ecology: Principles and Conservation. Cambridge University Press Keraita, B., Drechsel, P. & Amoah, P. (2003). Influence of urban wastewater on stream water quality and agriculture in and around Kumasi, Ghana. Environment and Urbanization, 15(2): 171-178

Keraita, B., Drechsel, P. & Konradsen, F. (2008a). Using on-farm sedimentation ponds to improve microbial quality of irrigation water in urban vegetable farming in Ghana. Water Science and Technology, 57(4): 519-525. Keraita, B., Drechsel, P., Huibers, F. & Raschid-Sally, L. (2002). Wastewater use in informal irrigation in Urban and Peri-urban areas of Kumasi, Ghana. Urban Agriculture Magazine, 8: 11-13

Keraita, B., Drechsel, P., Huibers, F. & Raschid-Sally, L. (2002). Wastewater use in informal irrigation in Urban and Peri-urban areas of Kumasi, Ghana. Urban Agriculture Magazine, 8: 11-13

169

Keraita, B., Drechsel, P., Konradsen, F. & Vreugdenhil, R. C. (2008b). Potential of simple filters to improve microbial quality of irrigation water used in urban vegetable farming in Ghana. Journal of Environmental Science and Health, 43: 749–755.

Kidane-Mariam, T. (2003). Environmental and Habitat Management: The Case of Ethiopia and Ghana. Environmental Management, 31(3): 313–327

Killeen, D., Khan, S. R (2002). Poverty and Environment. A contribution to the World Summit on Sustainable Development Process. International Institute for Environment and Development, London. Kim, J. (2006). A growth model for root systems of virtual plants with soil and moisture control. IEICE Transactions on Information and Systems, E89-D(5): 1743-1750 Klinkenberg, E, McCall, P. J., Wilson, M. D., Amerasinghe F. P., & Donnelly M. J. (2008). Impact of urban agriculture on malaria vectors in Accra, Ghana. Malaria Journal, 7: 1-9

Kumasi Metropolitan Assembly (2006). A Profile of Ashanti Region of Ghana. KMA, Kumasi. 43pp.

Lant, C. L. (1994). The role of property rights in economic research on U.S. wetlands policy. Ecological Economics, 11(1): 27-33 Levy, P. E., Cannell, M. G. R. & Friend A. D. (2004). Modelling the impact of future changes in climate, CO 2 concentration and landuse on natural ecosystems and the terrestrial carbon sink. Global Environmental Change, 14: 21–30 Lunetta, R. S., Balogh, M. E. & Merchant, J. W. (1999). Application of multi-temporal Landsat 5 TM imagery for wetland identification. Photogrammetric Engineering and Remote Sensing, 65, 1303-1310 Lydecker, M. & Drechsel, P. (2010). Urban agriculture and sanitation services in Accra, Ghana: the overlooked contribution. International Journal of Agricultural Sustainability, 8(1&2): 94-103 Lyon, J. G. (1993). Practical handbook for wetland identification and delineation. CRC Press, Inc., Boca Raton, Florida

Mabogunje, A. L. (1990). Urban planning and the post-colonial state in Africa: a research overview. African Studies Review, 33(2): 121–203

Magigi, W. (2008). Improving urban land governance with emphasis on Integrating agriculture based livelihoods in Spatial land use planning practise in Tanzania. PhD thesis at Albert-Ludwigs-Universität Freiburg im Breisgau, Germany Magnani, E. (2000). The Environmental Kuznets Curve, environmental protection policy and income distribution. Ecological Economics, 32: 431–443 Magnani, E. (2001). The Environmental Kuznets Curve: development path or policy result? Environmental Modelling & Software, 16:157–165 Maltby, E. & Barker, T. (eds) (2009). The Wetlands Handbook. Wiley and Blackwell, Oxford

170

Marcotullio, P. J. & Boyle, G (eds) (2003). Defining an Ecosystem Approach to Urban Management and Policy Development. United Nations University Institute of Advanced Studies (UNU/IAS)

Mathias, M. E. & Moyle, P. (1992). Wetland and aquatic habitats. Agriculture, Ecosystems & Environment, 42: 165-176

Matthys, B., Vounatsou, P., Raso, G., Tschannen, A. B., Becket, E.G., Gosoniu, L., Cissé, G., Tanner, M., N’goran, E. K., & Utzinger, J. (2006). Urban farming and malaria risk factors in a medium-sized town in Côte D’Ivoire. The American Journal of Tropical Medicine and Hygiene, 75(6): 1223–1231 Mazzotti, F. J. & Morgenstern, C. S. (1997) A scientific framework for managing urban natural areas. Landscape and Urban Planning, 38: 171-181

McClean, D. (ed.) (2010). World Disasters Report 2010 - focus on urban risk. International Federation of Red Cross and Red Crescent Societies, Geneva, Switzerland. ISBN 978-92- 9139-156-1 Mcdonnell, M. J. & Pickett, S. T. A. (1990). Ecosystem structure and function along urban- rural gradients: an unexploited opportunity for ecology. Ecology, 71(4): 1232-1237

McGregor, D., Simon, D. & Thompson, D. (eds) (2006). The Peri-Urban Interface: Approaches to sustainable natural and human resource use. Earthscan/ James & James, xxiv + 336pp. McPherson, M. A. & Nieswiadomy, L. M., (2005). Environmental Kuznets curve: threatened species and spatial effects. Ecological Economics, 55: 395– 407 Millennium Ecosystem Assessment (2000). Ecosystems and human well-being: current state and trends. Island Press Millennium Ecosystem Assessment (2005). Ecosystems and human well-being: current state and trends. Island Press Miller, R. L. & Fujii, R (2010). Plant community, primary productivity, and environmental conditions following wetland re-establishment in the Sacramento-San Joaquin Delta, California. Wetlands Ecology & Management, 18: 1–16 Mills, J. H. & Waite T. A. (2009). Economic prosperity, biodiversity conservation, and the environmental Kuznets curve. Ecological Economics, 68(7): 2087-2095 Ministry of Lands and Forestry (1999a). Managing Ghana's Wetlands: a National Wetlands Conservation Strategy. Accra, Ghana Ministry of Lands and Forestry (1999b). National Land Policy. Accra, Ghana.

Ministry of Water Resources, Works and Housing (2007). National Water Policy. Accra, Ghana Mireri, C., Atekyerezab, P., Kyessi, A. & Mushi, N. (2007). Environmental risks of urban agriculture in the Lake Victoria drainage basin: A case of Kisumu municipality, Kenya. Habitat International, 31: 375–386

171

Mitsch, W. J. & Gosselink, J. G. (1993). Wetlands. 2nd ed. Van Nostrand. Reinhold, New York Mitsch, W. J. Gosselink J.G. (2000). The value of wetlands: importance of scale and landscape setting. Ecological Economics, 35: 25–33 Modarres, R. & da Silva, V. R. (2007). Rainfall trends in arid and semi-arid regions of Iran. Journal of Arid Environments, 70: 344–355

Mougeot, L. J. A. (1999). Urban agriculture: definition, presence, potentials and risks, main policy challenges. In Proceedings of International Workshop on Growing Cities Growing Food: Urban Agriculture on the Policy Agenda, La Habana, Cuba Müller, G. (1969). Index of geoaccumulation in the sediments of the Rhine River. Geojournal, 2: 108–118 Mumford, L. (1961). The City in History: Its Origins, Its Transformations, and Its Prospects. New York: Harcourt, Brace Mumphrey, A. J. & Whalen, T. F. (1977). Economic growth, lifestyle preferences, and urban size. Urban Ecology, 2: 259-278 Munasinghe, M. (1999). Is environmental degradation an inevitable consequence of economic growth: tunnelling through the environmental Kuznets curve. Ecological Economics, 29: 89 – 109 Mundia, C. N. & Aniya, M. (2006). Dynamics of landuse/cover changes and degradation of Nairobi City, Kenya. Land Degradation & Development, 17: 97–108 National Development Planning Commission & UNDP (2010). 2008 Ghana Millennium Developme nt Goals Report. National Developme nt Planning Commiss io n, Accra, Ghana National Development Planning Commission (2001). Poverty Reduction Strategy (PRS) (2001– 2005). National Development Planning Commiss io n, Accra, Ghana National Development Planning Commission (2005). Growth and Poverty Reduction Strategy (GPRS II) (2006 – 2009). National Development Planning Commiss io n, Accra, GhanaGhana Nelson, K. C., Palmer, M. A., Pizzuto, J. E., Moglen, G. E., Angermeier, P. L., Hilderbrand, R. H., Dettinger, M. & Hayhoe, K. (2009). Forecasting the combined effects of urbanization and climate change on stream ecosystems: from impacts to management options. Journal of Applied Ecology, 46(1): 154-163 Niasse, M., Afouda, A. & Amani, A. (eds) (2004). Reducing West Africa's vulnerability to climate on water resources, wetlands and desertification: Elements for a regional strategy for preparedness and adaptation. IUCN: Gland, Switzerland and Cambridge, UK. xviii + 66pp. Niemelä, J. (1999). Ecology and urban planning. Biodiversity and Conservation, 8(1): 119- 131 Obeng, H. B. (1971). Soil Map of Ghana. Soil Research Institute, CSIR, Ghana

Obeng-Odoom, F. (2007). Enhancing urban productivity in Africa. Opticon 1826, 2(1): 1–9

172

Obuobie, E., Keraita, B., Danso, G., Amoah, P., Cofie, O. O., Raschid-Sally L. & Drechsel, P. (2006). Irrigated urban vegetable production in Ghana: Characteristics, benefits and risks. IWMI-RUAF-CPWF, Accra, Ghana: IWMI. 150pp.

Odemerho, F. 0. (1993). flood control failures in a third world city: Benin City, Nigeria -some environmental factors and policy issues. GeoJournal, 29(4): 371-376

Okley, B. L. (2004). Legislation and implementation of international environmental law by African countries: a case study of Ghana. LLM Theses and Essays, University of Georgia. Paper 38 Orloci, L. (1978). Multivariate Analysis in Vegetation Research. Dr W. Junk b.v., Publishers. The Hague. 451pp. Ot’ahel’ová, H., Valachovič, M. & Hrivnák, R. (2007). The impact of environmental factors on the distribution pattern of aquatic plants along the Danube River corridor (Slovakia). Limnologica, 37: 290–302 Owen, C. R. (1999). Hydrology and history: land use changes and ecological responses in an urban wetland. Wetlands Ecology and Management, 6: 209–219 Ozesmi, S. L. & Bauer, M. E. (2002). Satellite remote sensing of wetlands. Wetlands Ecology and Management, 10: 381 – 402 Pacione, M. (2009). Urban Geography: A Global Perspective (3rd edition). Routledge. 736pp.

Panayotou, T. (1993). Empirical tests and policy analysis of environmental degradation at different stages of economic development. Working Paper WP238, Technology and Employment Programme, Geneva: International Labour Office

Panayotou, T. (1997). Demystifying the Environmental Kuznets Curve: turning a black box into a policy tool. Environment and Development Economics, 2: 465-484

Panayotou, T. (2003). Economic growth and the environment. Paper prepared for and presented at the Spring Seminar of the United Nations Economic Commission for Europe, Geneva, March 3, 2003 Pauleit, S. & Duhme, F. (2000). Assessing the environmental performance of land cover types for urban planning. Landscape and Urban Planning, 52: 1-20 Payne, G. (1997). Urban land tenure and property rights in developing countries: a review. Intermediate Technology, London, UK. 80pp. Petrov, L. O., Barredo J. I., Lavalle, C., Sagris, V., De Roo, A., Gierk, M. & Genovese, E. (2005). Adaptation strategies for flood mitigation in the urban environment: the case study, Friuli-Venezia Giulia (FVG), Italy. International conference on innovation advances and implementation of flood forecasting technology, 17 to 19 October 2005, Tromsø, Norway. ACTIF/FloodMan/FloodRelief

Phillips, A. (ed) (1998). Economic Values of Protected Areas: Guidelines for Protected Area Managers. Best Practice Protected Area Guidelines Series No. 2. Task Force on Economic Benefits of Protected Areas of the World Commission on Protected Areas (WCPA) of IUCN, in collaboratio n with the Economics Service Unit of IUCN

173

Pickett, S. T. A., Burch, W. R., Dalton, S. E., Foresman, T. W., Grove, J. M. & Rowntree, R. (1997). A conceptual framework for the study of human ecosystems in urban areas. Urban Ecosystems, 1: 185–199

Pickett, S. T. A., Kolasa, J., Armesto, J. J. &. Collins, S. L. (1989). The ecological concept of disturbance and its expression at various hierarchical levels. Oikos, 54: 129-136

Pongrácz, R., Bartholy, J. & Dezső, Z. (2006). Remotely sensed thermal information applied to urban climate analysis. Advances in Space Research, 37: 2191–2196 Pongrácz, R., Bartholy, J. & Dezső, Z. (2010). Application of remotely sensed thermal information to urban climatology of Central European cities. Physics and Chemistry of the Earth, Parts A/B/C, 35(1-2): 95-99 Qadeer, M. A. (2004). Urbanization by implosion. Habitat International, 28: 1–12

Quansah, C., Drechsel, P. & Bakang, J. A. (2005). Participatory communication strategies to improve natural resources management in peri-urban Kumasi, Ghana. In Mear, A. (Ed.). Food and Water for Millions. Participatory communication for land and water management. IDRC, Ottawa, p.81-88 Rahman, A. (2006). Assessing income-wise household environmental conditions and disease profile in urban areas: Study of an Indian city. GeoJournal, 65: 211–227 Ramsar Convention Secretariat (2005). Wetlands and water: supporting life, sustaining Livelihoods. 9th Meeting of the Conference of the Parties to the Convention on Wetlands (Ramsar, Iran, 1971). Kampala, Uganda, 8-15 November 2005

Ramsar Convention Secretariat (2006). The Ramsar Convention Manual: a guide to the Convention on Wetlands (Ramsar, Iran, 1971), 4th ed. Ramsar Convention Secretariat, Gland, Switzerland Ramsar Convention Secretariat (2007). Ramsar handbooks for the wise use of wetlands, 3rd edition. Ramsar Convention Secretariat, Gland, Switzerland Ramsar Convention Secretariat (2008). Healthy Wetlands, Healthy People. 10th Meeting of the Conference of the Contracting Parties to the Convention on Wetlands (Ramsar, Iran, 1971). Changwon, Republic of Korea, 28 October - 4 November, 2008 Ravallion, M, Shaohua, C. & Sangraula, P. (2007a). The Urbanisation of Global Poverty. Background paper to the 2008 World Development Report Ravallion, M., Chen, S. & Sangraula, P. (2007b). New Evidence on the Urbanization of Global Poverty. World Bank Policy Research Working Paper 4199. Development Research Group, World Bank

Rees, W. E. (2003). Understanding urban ecosystems: an ecological economics perspective. In: Understanding Urban Ecosystems: A New Frontier for Science and Education. A.R. Berkowitz, C.H. Nilon, and K.S. Hollweg (eds.), Springer-Verlag, New York, 115–136 Reinelt, L., Horner, R. & Azous, A. (1998). Impacts of urbanization on palustrine (depressional freshwater) wetlands—research and management in the Puget Sound region. Urban Ecosystems, 2: 219–236

174

Rietbergen, S., Bishop, J. & Mainka S. (2002). Ecosystem Conservation - A neglected tool for poverty reduction. A contribution to the World Summit on Sustainable Development Process. International Institute for Environment and Development, London

Risbey, J. S., Hamza, K. & Marsden, J. S. (2007). Use of climate scenarios to aid in decision analysis for interannual water supply planning. Water Resources Management, 21: 919–932

Rodrigo, F. S., Pozo-Vazquez, D., Esteban-Parra, M. J. & Castro-Diez, Y. (2001). A reconstruction of the winter North Atlantic Oscillation index back to A.D. 1501 using documentary data in southern Spain. Journal of Geophysical Research, 106: 805–814 Roy, M. (2009). Planning for sustainable urbanisation in fast growing cities: mitigation and adaptation issues addressed in Dhaka, Bangladesh. Habitat International, 33: 276–286 Sachs, J. McArthur, J., Schmidt-Traub, G., Bahadur C., Faye M. & Kruk, M. (2004) Millennium Development Goals needs assessments. Country case studies of Bangladesh, Cambodia, Ghana, Tanzania and Uganda. UN Mille nnium Project Sánchez-Rodríguez, R., Seto, K. C., Simon, D., Solecki, W. D., Kraas, F. & Laumann, G. (2005). Science Plan: Urbanization and Global Environmental Change. IHDP Report No. 15. International Human Dimensions Programme on Global Environmental Change (IHDP). Bonn, Germany Sassa, L. L., Bozek, M. A., Hauxwell, J. A., Wagner, K. & Knight, S. (2010). Response of aquatic macrophytes to human land use perturbations in the watersheds of Wisconsin lakes, U.S.A. Aquatic Botany, 93: 1-8 Satterthwaite, D. (2003). The links between poverty and the environment in urban areas of Africa, Asia, and Latin America. Annals of the American Academy of Political and Social Science, 590: 73–92 Satterthwaite, D. (2008). Climate change and urbanization: effects and implications for urban governance. United Nations Expert Group Meeting on Population Distribution, Urbanization, Internal Migration and Development. UN/POP/EGM-URB/2008/16 Schiff, K. C. & Weisberg, S. B. (1999). Iron as a reference element for determining trace metal enrichment in Southern California coast shelf sediments. Marine Environmental Research, 48: 161–176 Schmidt, S. (2005). Cultural influences and the built environment: an examination of Kumasi, Ghana. Journal of Urban Design, 10(3): 353-370 Schropp, S. J. & Windom, H. L. eds. (1988). A guide to the interpretation of metal concentrations in estuarine sediments. Office of Water Policy, Florida Department of Enviro nme nta l Protection, Tallahasse e

Secretariat of the Convention on Biological Diversity (2005). Handbook of the Convention on Biological Diversity Including its Cartagena Protocol on Biosafety, 3rd edition. Montreal, Canada Secretariat of the Convention on Biological Diversity (2007). Cities and Biodiversity: Achieving the 2010 Biodiversity Target. Mayors’ Meeting on the Contribution of Cities to the Achievement of the 2010 Biodiversity Target. Curitiba, Brazil, 26-28 March 2007. UNEP/CBD/Cities/1/3

175

Sekabira, K., Oryem Origa, H., Basamba, T. A., Mutumba, G. & Kakudidi, E. (2010). Assessment of heavy metal pollution in the urban stream sediments and its tributaries. International Journal of Environmental Science and Technology, 7(3): 435-446

Shutes, R. B. E. (2001). Artificial wetlands and water quality improvement. Environment International, 26: 441 - 447

Shutes, R. B. E., Revitt, D. M., Lagerberg, I. M. & Barraud, V. C. E. (1999). The design of vegetative constructed wetlands for the treatment of highway runoff. Science of the Total Environment, 235: 189-197 Silvius, M. J., Oneka, M. & Verhagen, A. (2000). Wetlands: lifeline for people at the edge. Physics and Chemistry of the Earth (B), 25(7-8): 645-652 Simex, S. A. & Helz, G. R. (1981). Regional geochemistry of trace elements in Checapeake Bay. Environmental Geology, 3: 315-323

Sims, K. R. E. & Schuetz, J. (2009). Environmental regulation and land use change: do local wetlands bylaws slow the conversion of open space to residential uses? Regional Science and Urban Economics, 39(4): 409-421 Small, C. & Lu, J. W. T (2006). Estimation and vicarious validation of urban vegetation abundance by spectral mixture analysis. Remote Sensing of Environment, 100: 441 – 456 Small, C. (2001). Estimation of urban vegetation abundance by spectral mixture analysis. International Journal of Remote Sensing, 22(7): 1305-1334 Smith, M. B. (1993). A GIS-based distributed parameter hydrologic model for urban areas. Hydrological Processes, 7: 45-61

Smith, M. E. (2005). City size in late post-classic Mesoamerica. Journal of Urban History, 31(4): 403-434

Sorrell, B. K., Partridge, T. R., Clarkson, B. R., Jackson, R. J., Chague´-Goff, C., Ekanayake, J., Payne, J., Gerbeaux, P. & Grainger, N. P. J. (2007). Soil and vegetation responses to hydrological manipulation in a partially drained polje fen in New Zealand. Wetlands Ecology and Management, 15: 361–383

Stern, D. I. & Common, M. S. (2001). Is there an environmental Kuznets curve for sulfur? Journal of Environmental Economics and Management, 41: 162–178

Stern, D. I. (2004). The rise and fall of the environmental Kuznets curve. World Development, 32(8): 1419–1439

Takebayashi, H. & Moriyama, M. (2009). Study on the urban heat island mitigation effect achieved by converting to grass-covered parking. Solar Energy, 83(8):1211-1223 Taylor, B. L. (1993). The Influence of Wetland and Watershed Morphological Characteristics on Wetland Hydrology and Relationships to Wetland Vegetation Communities. MSCE thesis, Department of Civil Engineering, University of Washington Terry, P. J. (1983). Some common crop weeds of West Africa and their control. USAID Regional Food Crop Protection Project

176

The World Bank (1991). Urban Policy and Economic Development; an agenda for the 1990s. A World Bank Policy Paper, World Bank The World Bank (2000). Entering the 21st Century: World Development Report, 1999-2000, pp. 126-138 The World Bank (2007). Ghana: Country Environmental Analysis. Environment and Natural Resources. Agence Française de Développement (AFTEN), Royal Netherlands Embassy and Department for International Development, Africa Region. Report No. 36985-GH Thenkabail, P. S. & Nolte, C. (1996). Capabilities of Landsat- 5 Thematic Mapper (TM) data in regional mapping and characterization of inland valley agroecosystems in West Africa. International Journal of Remote Sensing, 17(8): 1505-1538 Thenkabail, P. S., Nolte, C. & Lyon, J. G. (2000). Remote sensing and GIS modeling for selection of benchmark research area in the inland valley agroecosystems of West and Central Africa. Photogrammetric Engineering and Remote Sensing, 66(6): 755-768 Thibault, P. A. & Zippererb, W. C. (1994). Temporal changes of wetlands within an urbanizing agricultural landscape. Landscape and Urban Planning, 28: 245-251 Tomlinson, D. L., Wilson, J. G., Harris, C. R., & Jeffrey, D.W. (198 0). Problems in the assessment of heavy- metal levels in estuaries and the formation of a pollution index. Helgolander Wissenschaftliche Meeresuntersuchungen, 33: 566–575 Töyrä, J., Pietroniro, A., Martz, L. W. & Prowse, T. D. (1999). A multi-sensor approach to wetland flood monitoring. Hydrological Processes, 16 (8): 1569-1581

Tschakert, P., Huber-Sannwald, E, Ojima, S. D, Raupach, M. R, Schienke, E. (2008). Holistic, adaptive management of the terrestrial carbon cycle at local and regional scales. Global Environmental Change, 18: 128–141 Turekian, K. K. & Wedepohl, K. H. (1961). Distribution of the elements in some major units of the earth's crust. Geological Society of America Bulletin, 72: 175-192 Turner, R. K., van den Bergh, J. C. J. M., Söderqvist, T., Barendregt, A., van der Straaten, J., Maltby, E. & van Ierland, E. C. (2000). Ecological-economic analysis of wetlands: scientific integration for management and policy. Ecological Economics, 35: 7–23 U. S. Environmental Protection Agency (1985). Sediment sampling quality assurance user’s guide. EPA/600/4-85/048. PB85-233542. Environmental Monitoring Systems Laboratory. Las Vegas, NV UN-HABITAT (1996). An Urbanizing World-Global Report on Human Settlements 1996. Oxford. Oxford University Press. ISBN: 0-19-823346-9

UN-HABITAT (2008a). Country Programme Document 2008 – 2009: Ghana. Regional Office for Africa and the Arab States, Nairobi, Kenya

UN-HABITAT (2008b). State of the Word’s Cities 2008/2009: Harmonious Cities. Earthscan, London

UN-HABITAT (2009). Ghana urban profile. UN-Habitat Regional and Information Office, Nairobi, Kenya

177

United Nations (2008). World Urbanization Prospects: The 2007 Revision (Highlights). United Nations, New York United Nations Department of Economic and Social Affairs (2005). Changing unsustainable patterns of consumption and production: human settlements and water. Background Paper No.4 DESA/DSD/2005/4

United Nations Inter-Agency Secretariat of the International Strategy for Disaster Reduction (UN/ISDR) (2005). Hyogo Framework for Action 2005-2015: Building the Resilience of Nations and Communities to Disasters. World Conference on Disaster Reduction 18-22 January 2005, Kobe, Hyogo, Japan (A/CONF.206/6) United Nations Population Fund (UNFPA) (2007). State of World Population 2007. UNFPA. UNU-IAS Urban Ecosystems Management Group (2004). Cities as Drivers of Sustainable Development. World Urban Forum 2004 Discussion Paper. United Nations University - Institute of Advanced Studies Vincent, M. A. & Saatchi, S. S. (1999). Comparison of Remote Sensing Techniques for Measuring Carbon Sequestration. Jet Propulsion Laboratory, California Institute of Technology

Wagner, M. (2008). The carbon Kuznets curve: A cloudy picture emitted by bad econometrics? Resource and Energy Economics, 30: 388–408 Water Resources Commission (2008). Final draft buffer zone policy for managing river basins in Ghana. Accra, Ghana

White, H., Killick, T., Kayizzi- Mugerwa, S. & Savane, M. (2001). African Poverty at the Millennium: Causes, Complexities, and Challenges. The International Bank for Reconstruction and Development/The World Bank. Washington, DC Wikum, W. A. & Shanholtzer, G. F. (1978). Application of the Braun-Blanquet cover- abundance scale for vegetation analysis in land development studies. Environmental Management, 2(4): 323-329 Wilby, R. L. (2007). A Review of Climate Change Impacts on the Built Environment. Built Environment, 33: 31-45 Williams, M. (ed). (1990). Wetlands: a threatened landscape. Basil Blackwell Ltd., Oxford, U.K. 419pp. World Meteorological Organisation (2006). Environmental aspects of integrated flood management. The Associated Programme on Flood Management (APFM) Technical Document No. 3, Flood Management Policy Series. Geneva, Switzerland

Xu, H., Wang, X. & Xiao, G. (2000). A Remote Sensing and GIS integrated study on urbanization with its impact on arable lands: Fuqing City, Fujian Province, China. Land Degradation & Development, 11: 301-314 Yeboah, M. A. & Appiah-Yeboah, K. (2009). An examination of the cultural and socio- economic profiles of porters in Accra, Ghana. Nordic Journal of African Studies, 18(1): 1 –21 Zedler, J. B. & Leach, M. K. (1998). Managing urban wetlands for multiple use: research, restoration, and recreation. Urban Ecosystems, 2: 189–204 178

9 APPENDICES Appendix 1. Survey instrument used for the wetland associated communities in Kumasi

Questionnaire number…………………………Area………………………………...... Type of house...... Date…………………………………

INTRODUCTION The researcher is a student of the University of Bremen, Germany working on urban wetland ecology and flooding in Kumasi. The information you provide will be treated as confidential and will be used for academic purposes only. Thank you for your time and cooperation.

1.1 Do you own the house you live in yes [ ] no [ ] 1.2 How did you get access to this land? Family inheritance [ ] Chief [ ] Bought from another person [ ] Squatter [ ] Not applicable [ ] 1.3 For how long have you been living in the area? Below 5years [ ] 5-10 years [ ] 10-15 years [ ] 15-20 years [ ] more than 20 years [ ] Please rank the following options according the criterion below: 0= no change/no influence/no contribution/nothing, 1=… 2= … 3=… 4=… in increasing order to 5=highest increase/most influential/most contribution. 1.4 What do you do for a living? Job types 0 1 2 3 4 5 1. Trading 2. Farming 3. Artisan (e.g. car fitting/Saw milling/carpentry etc.) 4. Tie-dye manufacturing 5. Oil extracting (palm oil/palm kernel oil etc.) 6. Formal sector job 7. Nothing/unemployed/dependant Other, please specify 1.5 What are the predominant economic activities in this area? Activity/use of land at time of moving in 0 1 2 3 4 5 1. Agriculture (vegetables, animals, etc.) 2. Tie-dye industries 3. Hunting 4. Fishing 5. Car washing bays

179

6. Car fitting/spraying shops 7. Carpentry/sawmill 8. Trading Other (specify) 1.6 What was the previous use of this wetland to the people of this area? ……………...... 1.7 What is/are the current uses of this stream/wetland to the people of this area? Uses 0 1 2 3 4 5 1. Agricultural purposes 2. Tie-dye industries 3.Waste disposal 4. Car washing bays 5. Car fitting shops 6. Carpentry 7. Water supply 8. Fishing 9. Rangeland/forage area for cows sheep and goats 10. Recreation/tourism

Other, please state 1.8 What are the major landuse activities around the wetland (100 m from water channel)? Activities 0 1 2 3 4 5 1. Agriculture (vegetables, animals, etc.) 2. Tie-dye industries 3. Waste disposal 4. Car washing bays 5. Car fitting shops 6. Carpentry 7. Rangeland/forage area 8. Recreation/tourism 9. Housing/construction Other, please specify 1.9 What have been the effects of these activities (1.8 above) on the environment/wetland? Changes 0 1 2 3 4 5 1. Channel width (size of the river) 2. Water depth 3. Number and type of living organisms (e.g. fish, plants animals) 4. Storm flow (amount of water in the river/channel immediately after rainfall) 5. Suitability of water for domestic/industrial purposes

180

Other, please specify 2.0 What changes have you observed in the wetland vegetation since you first settled? Changes 0 1 2 3 4 5 1. Plant species diversity 2. Number of woody plants (e.g. trees) 3. Number of herbaceous plants (e.g. grasses) 4. Distribution of plants 5. Density of plants 6. Plant cover 7. Loss of some species 8. Presence of new species Other, please specify 2.1 Buildings in wetland areas are major problems in Kumasi. Strongly agree [ ] Agree [ ] Disagree [ ] Not sure [ ] 2.2 What are some of the effects building on wetland areas have on our environment? Effects 0 1 2 3 4 5 1. Nothing 2. Causes flooding 3. Makes the environment dirty 4. No place to dump wastes anymore 5. Does not make the city beautiful Other, please specify 2.3 Why do you think people build in wetland areas? Reasons 0 1 2 3 4 5 1. Wetland areas are cheap 2. Close to amenities 3. Direct use of the water for domestic purposes 4. Close to family 5. Do their economic activities 6. It is “no man’s land” Other, please specify 2.4 Have you experienced flood in this area? Yes [ ] No [ ] 2.5 If yes, how many times in a year do you experience floods? Once [ ] Twice [ ] Thrice [ ] more than thrice [ ] 2.6. Which month(s) of the year do you experience the floods?………………………………………

181

2.7. How long do the floodwaters take to recede? Less than 5 days in a year [ ] less than 10 days in a year [ ] less than 15 days in a year [ ] less than 20 days in a year [ ] less than 25 days in a year [ ] About a month in a year [ ] 2.8. How do you assess the frequency of the floods compared to 10 years ago? Less frequent [ ] More frequent [ ] 2.9 How do you deal with floods as an individ ua l and community? (mark I or C or both) Activities 0 1 2 3 4 5 1. Do nothing 2. Fetch/pump water out of flooded areas 3. Move out till water subsides 4. Create channels for easy flow of water 5. Dredge stream 6. Build embankments along stream 7. Complain to authorities Others, please specify 3.0 Have you ever been educated on the usefulness of a wetland? Yes [ ] no [ ] If yes, give the name(s) of the organisation(s) ………………………………………………………………………………………………… 3.1 Has there been any attempt by an organisation to move you from this site? yes [ ] no [ ] If yes, name of organisation…………………………………………………………………… 3.2 What will make you move out of this place? Motivation/reason 0 1 2 3 4 5 1. Nothing, I love it here 2. Given money to move 3. New land/plot somewhere 4. When I make some money 5. When the water gets too much 6. When we are forced 7. I am already preparing to move 8. If I am provided alternative shelter Other, please specify 4.0 Do you think the wetland should be protected? yes [ ] no [ ] 4.1 How are you and your community protecting this wetland? Activities 0 1 2 3 4 5 1. We are doing nothing 2. Stop waste disposal on wetland 3. Stop building too close to wetland 4. Create channels for easy flow of water 5. Dredge stream 182

6. Build embankments along stream 7. Stop all economic activities on wetland 8. Complain to authorities Others, please specify 4.2 Have houses in this neighbourhood ever been marked for demolition because of the wetland? Yes [ ] No [ ] If yes, when less than 2 yrs age [ ] 2-5 yrs ago [ ] 5-10 yrs ago [ ] more than 10 yrs [ ] 4.3 Has any demolition been carried out in this area? Yes [ ] No [ ] 4.4 Do you support the demolition along water ways? Yes [ ] No [ ] 4.5 Who are responsible for the management of the wetlands in Kumasi? Agencies/institutions 0 1 2 3 4 5 1. KMA 2. Hydrological Services Department 3. Town and Country Planning Department 4. The community and Assemblyman 5. NGOs (e.g. IWMI, Friends of Rivers and

Waterbodies) 6. NADMO 7. The chief/land owners Others, please specify 4.6 Are there any on-going management practices to protect the wetland? Yes [ ] No [ ] 4.7. What factors affect the implementation of wetland policies and management plans? Factors 0 1 2 3 4 5 1. Politica l influe nce 2. Changes in government 3. Corruption 4. Poverty levels 5. Lack of politica l will 6. Cost of demolition 7. Cultural values Other, please specify 4.8 Do you know any of the laws regarding building or sitting of infrastructure in wetland areas? Yes [ ] No [ ] If yes what is it Background Information 5.1 Age: below 20yrs [ ] 20-30yrs [ ] 30-40yrs [ ] 40-50yrs [ ] 50-60yrs [ ] above 60yrs [ ] 5.2 Sex: Male [ ] Female [ ]

183

5.3 Education: Primary [ ] Middle School/JHS [ ] Secondary/SHS [ ] Post-secondary/Tertiary [ ] No formal education [ ] Other [ ] 5.4 Occupation: …………………Household size:......

184

Appendix 2. Survey instrument used for wetland associated NGOs and government agencies in Kumasi

Date……………………………….

INTRODUCTION The researcher is a student of the Kwame Nkrumah University of Science and Technology working on urban wetlands in Kumasi. The information you provide will be treated as confidential and will be used only for academic purposes. Thank you for your time and cooperation.

1. In your opinion; what has been the overall change in the state of the wetland?

2. What are the tasks of inspectors/ officers-in- charge of developments on wetland areas?

3. How often do developers who flout the regulations come to your notice?

4. What punitive measures are there for such offenders?

5. Are these punitive measures effective?

6. How can the effectiveness be improved?

7. Are prospective (building) developers aware of these regulations/rules?

8. How often does your outfit organise educative seminars for people living on wetlands?

9. What medium do you use for such educative forum?

10. Are there any risks/dangers your inspectors/ officers face in visiting development sites in wetland (e.g. from land guards, mob attack, etc)?

11. Under what circumstances can a building on wetlands be pulled down?

12. Are there maps/demarcations showing wetland areas?

13. What major constraints affect the implementation of wetlands management plan?

14. How do non-economic variables other than per capita income, such as the features of the political system, and some cultural values, affect the implementation of wetland policies and management plan?

185

Appendix 3. Field data sheet used for wetland vegetation and ecological studies in Kumasi Distance from water channel wate r species 10m 20m 30m 40m ≥50m Re mark/comme nt channel # cover # cover # cover # cover # cover # cover

186

Appendix 4. Study sites and anthropogenic activities within 100 m from the water channel [minimum buffer distance set by Ministry of Lands and Natural Resources (Water Resources Commission, 2008)] KNUS T Atonsu High School FRNR Family Golden Kaase- Bekwai Police Spetinpom Dakwodjom Sawmill Junction farms Chapel Tulip Hotel Guinness roundabout Station F K A B C E G H J D % % % % % % % % % land % land Activity/ # land # land # land # # land # land # land # land # land # cover cover landuse cover cover cover cover cover cover cover cover Hous ing 74 34 52 35 1 0 19 6 65 50 45 55 45 20 37 30 152 55 30 20 Farming 2 0 3 5 9 2 15 2 7 10 11 5 12 12 3 15 41 20 3 5 Natural vegetation 50 45 98 91 30 30 65 50 0 20 70 Waste disposal 3 10 4 5 0 0 3 0 7 5 2 3 2 0 1 0 2 0 2 0 Schools 0 0 1 0 0 0 1 1 1 2 1 1 0 0 2 1 3 0 0 0 Church 0 0 5 0 0 0 1 0 7 1 0 1 2 0 4 1 3 0 2 0 Mos que 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 Carpentry/saw mills 2 5 15 1 0 0 0 0 9 0 0 0 0 0 0 0 0 0 6 0 Car washing bay 2 0 5 0 0 0 0 0 1 0 1 0 2 0 2 0 12 1 0 0 Car repairs 2 0 22 5 0 0 0 0 15 2 0 0 0 0 22 2 12 0 6 2 Road 2 0 5 0 1 0 6 0 6 0 2 0 6 2 3 0 6 2 2 0 Gas/fuel filling station 0 0 1 0 0 0 0 0 2 0 3 1 2 1 3 1 3 0 0 0 Market 1 1 2 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Kios k s 4 0 6 2 0 0 2 0 0 0 0 2 3 0 12 0 17 0 9 0 Oil extraction 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 6 0 Tie-Dye industry 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Playground 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 2 1 3 Fish processing 1 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 22 0 Total 94 100 125 100 11 100 47 100 135 100 65 100 74 100 89 100 251 100 89 100

Appendix 5. A Wilcoxon Signed Ranks Test comparison of the seasonal differences in physicochemical parameters of the wetlands in Kumasi

187

Mean Sum of Test Statistics N Rank Ranks Z As ymp. S ig . (2 -tailed) rainy_ pH - dry_ pH Negative Ranks 10a 5.50 55.00 -2.803 0.005 Positive Ranks 0b .00 .00 Ties 0c Total 10 rainy_Temp - dry_Temp Negative Ranks 7d 5.86 41.00 -1.376 0.169 Positive Ranks 3e 4.67 14.00 Based on positive Ties 0f ranks Total 10 Rainy_Sal - dry_Sal Negative Ranks 5g 3.30 16.50 -.211 0.833 Positive Ranks 3h 6.50 19.50 Based on negative ranks Ties 2i Total 10 Rainy_DO - dry_DO Negative Ranks 1j 2.00 2.00 -2.599 0.009 Positive Ranks 9k 5.89 53.00 Based on negative Ties 0l ranks Total 10 rainy_TDS - dry_TDS Negative Ranks 5m 5.40 27.00 -.051 0.959 Positive Ranks 5n 5.60 28.00 Based on negative Ties 0o ranks Total 10

188

Appendix 6. Metal concentration (mg/kg) in sediments of the different study sites in Kumasi study sites High KNUST Golden Kaase Bekwai Atonsu School FRNR Family Spetinpo Dakwadwo Police Tulip Guinnes roundabou Sawmill Junctio farms Chapel m m Station Hotel s t n Arsenic 138.6 117.5 22.5 175.8 44.0 37.2 54.7 25.2 45.0 24.5 Copper 652.5 513.2 152.1 72.6 621.3 184.2 584.5 245.2 445.8 125.2 Lead 425.3 189.2 35.8 45.3 452.1 49.5 65.2 75.3 122.0 87.5 rainy Zinc 842.6 752.5 154.2 152.2 825.2 152.0 242.2 239.0 234.2 541.2 season Chromium 542.2 452.2 142.3 145.3 1078.3 142.8 221.5 236.5 189.5 425.3 17/06/2010 Cadmium 0.9 0.6 1.0 0.6 1.0 0.6 0.5 0.6 0.6 0.8 Nickel 443.8 479.2 145.2 109.2 542.2 105.2 179.2 102.7 107.2 102.2 Mercury 2.1 2.5 0.9 0.7 0.6 0.8 0.8 0.8 0.8 0.8 Phosphorus 7785.2 3022.2 2230.5 18897.5 15202.5 1745.2 4421.9 5232.2 4523.2 2202.0

Arsenic 48.6 185.7 25.1 85.2 200.1 47.2 295.4 277.2 189.1 31.3 Copper 652.5 624.6 214.2 127.4 1052.1 984.2 497.2 589.2 389.5 323.2 Lead 425.3 489.2 38.5 47.6 612.9 84.9 86.2 405.4 342.4 332.1 dry Zinc 1542.6 1252.5 177.2 242.2 1041.2 427.5 327.6 557.2 534.2 582.1 season Chromium 542.2 752.2 152.8 324.2 532.9 1542.9 324.9 723.1 287.5 325.1 14/01/2010 Cadmium 0.8 1.0 1.3 0.6 0.8 1.0 1.0 0.8 0.9 1.3 Nickel 843.8 779.2 107.6 104.9 642.2 112.7 124.2 148.2 185.2 134.2 Mercury 2.5 3.7 0.8 1.0 2.1 0.8 0.8 1.0 0.8 0.8 Phosphorus 17785.2 19022.2 11852.3 22521.2 16125.2 12521.3 16895.0 11214.4 12523.2 17542.4

189

Appendix 7. Pearson correlation matrix of heavy metals and other measured environmental variables (significant relationships shown in bold)

Arseni Copper Le ad Zinc Chromiu Cadmiu Nickel Mercur Phosphor pH Te m pe r at u Salinity Dissolve Dissolve sand silt clay c m m y us re d O2 d solids Arsenic Copper 0.287 Le ad 0.358 0.538 Zinc 0.127 0.592 0.852 Chromium 0.053 0.743 0.188 0.330 Cadmium -0.398 -0.003 -0.116 -0.127 -0.142 Nickel 0.121 0.510 0.736 0.939 0.196 -0.155 Mercury 0.141 0.426 0.654 0.867 0.257 -0.164 0.952 Phosphorus -0.066 -0.254 0.027 0.255 -0.165 -0.332 0.335 0.416 pH 0.167 0.243 0.567 0.691 0.330 0.032 0.647 0.775 0.165 Te m pe r at u 0.438 0.068 0.406 0.204 -0.041 -0.305 0.043 -0.088 -0.305 0.090 re Salinity 0.441 0.203 0.533 0.292 0.015 -0.309 0.057 -0.101 -0.201 0.042 0.632 Dissolved -0.222 0.352 -0.092 -0.067 0.148 0.381 0.108 0.113 -0.259 -0.207 -0.401 -0.666 O2 Dissolved 0.434 0.202 0.517 0.279 0.032 -0.325 0.042 -0.115 -0.203 0.032 0.628 0.999 -0.672 solids sand 0.733 0.155 0.382 0.099 -0.147 0.179 0.156 0.238 -0.223 0.362 0.187 0.078 0.136 0.058 silt -0.401 -0.427 -0.465 -0.460 -0.536 -0.129 -0.349 -0.502 -0.082 -0.774 -0.136 -0.009 0.046 -0.001 -0.528 clay -0.415 0.198 0.017 0.270 0.607 -0.131 0.064 0.075 0.304 0.186 0.015 0.073 -0.289 0.090 -0.677 -0.246

190

Appendix 8. Principal Component Analysis of environmental variables and heavy metals assessed in the wetlands of Kumasi

Eigenvalues PC Eigenvalues % Variation Cum. % Variation 1 5.52 32.4 32.4 2 3.48 20.5 52.9 3 2.52 14.8 67.7 4 1.93 11.3 79.1 5 1.3 7.7 86.8

Eigenvectors (Coefficients in the linear combinations of variables making up PC's) Variable PC1 PC2 PC3 PC4 PC5 Arsenic 0.197 -0.257 0.313 0.059 -0.214 Copper 0.273 0.098 0.048 -0.442 0.217 Lead 0.378 -0.065 0.062 0.028 0.239 Zinc 0.387 0.105 -0.113 0.066 0.238 Chromium 0.188 0.143 -0.178 -0.510 -0.209 Cadmium -0.100 0.197 0.236 -0.191 -0.009 Nickel 0.344 0.194 -0.031 0.191 0.309 Mercury 0.337 0.261 -0.007 0.231 0.099 Phosphorus 0.056 0.153 -0.289 0.498 -0.146 pH 0.326 0.151 0.017 0.116 -0.318 Temperature 0.152 -0.373 0.026 -0.066 0.007 Salinity 0.182 -0.450 -0.093 -0.073 0.122 Dissolved O2 -0.076 0.355 0.282 -0.231 0.324 Dissolved solids 0.177 -0.452 -0.105 -0.082 0.115 sand 0.156 -0.054 0.553 0.094 -0.210 silt -0.290 -0.124 -0.120 0.122 0.543 clay 0.078 0.083 -0.539 -0.233 -0.224

191

Appendix 9. Comparison of means of heavy metal enrichment factors for rainy and dry season in Kumasi

Normality Test: Failed (P < 0.050)

Test execution ended by user request, Rank Sum Test begun

Mann-Whitney Rank Sum Test

Group N Missing Median 25 % 75 % EF ra iny season 90 0 3.241 1.968 6.742 EF dry season 90 0 5.994 2.842 16.063

Mann-Whitney U Statistic= 2694.000 T = 6789.000 n(small)= 90 n(big)= 90 (P = <0.001)

The difference in the median values between the two groups is greater than would be expected by chance; there is a statistically significant difference (P = <0.001)

Appendix 10. Comparison of means of heavy metal geoaccumulation indices for rainy and dry season in Kumasi

Normality Test: Failed (P < 0.050)

Test execution ended by user request, Rank Sum Test begun

Mann-Whitney Rank Sum Test

Gr ou p N Mis s ing Me di an 2 5 % 7 5 % geoaccumulation rainy season 90 0 1.050 0.330 2.107 geoaccumulation dry season 90 0 1.937 0.861 3.359

Mann-Whitney U Statistic= 2694.000

T = 6789.000 n(small)= 90 n(big)= 90 (P = <0.001)

The difference in the median values between the two groups is greater than would be expected by chance; there is a statistically significant difference (P = <0.001)

193

Appendix 11. Comparison of means of Pollution Load Indices for the different sites in the rainy and dry season in Kumasi

Normality Test: Passed (P = 0.051)

Equal Variance Test: Passed (P = 0.374)

Group Name N Missing Mean Std Dev SEM PLI rainy s 10 0 4.082 2.199 0.695 PLI dry sea 10 0 6.841 3.062 0.968

Difference -2.759 t = -2.314 with 18 degrees of freedom. (P = 0.033)

95 percent confidence interval for difference of means: -5.264 to -0.254

The difference in the mean values of the two groups is greater than would be expected by chance; there is a statistically significant difference between the input groups (P = 0.033).

Power of performed test with alpha = 0.050: 0.498

194