LAND USE/LAND COVER CHANGES AND ASSOCIATED DRIVING FORCES IN

BALE ECO-REGION, ETHIOPIA

MSc. THESIS

BY: ADANE MEZGEBU NIGUSSIE

HAWASSA UNIVERSITY WONDO GENET COLLEGE OF FORESTRY AND NATURAL

RESOURCES, WONDO GENET, ETHIOPIA

DECEMBER, 2016

LAND USE/LAND COVER CHANGES AND ASSOCIATED DRIVING FORCES IN

BALE ECO-REGION, ETHIOPIA

BY: ADANE MEZGEBU NIGUSSIE

A THESIS SUBMITTED TO SCHOOL OF NATURAL RESOURCE AND

ENVIRONMENTAL STUDIES, HAWASSA UNIVERSITY WONDO GENET COLLEGE

OF FORESTRY AND NATURAL RESOURCES, HAWASSA UNIVERSITY, WONDO

GENET, ETHIOPIA

IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF

MASTER OF SCIENCE IN CLIMATE CHANGE AND DEVELOPMENT

DECEMBER, 2016

Approval sheet I

This is to certify that the thesis entitled “Land Use/Land Cover Changes and Associated

Driving Forces in Bale Eco-Region, Ethiopia” submitted in partial fulfillment of the

requirement for the degree of Master’s with specialization in Climate change and

Development, the graduate program of the school of Natural Resources and Environmental

Studies, and has been carried out by Adane Mezgebu Id. No MSc/CcDe/R002/07, under my

supervision. Therefore, I recommend that the student has fulfilled the requirements and hence

hereby can submit the thesis to the department.

Motuma Tolera (PhD) _______________ ____________

Name of Major Advisor Signature Date

Menfese Tadesse (PhD) _________________ ____________

Name of co-advisor Signature Date

Approval sheet II

We the undersigned members of the Board of Examiners of the final open defense by Adane

Mezgebu have read and evaluated his thesis entitled “Land Use/Land Cover Changes and

Associated Driving Forces in Bale Eco-Region, Ethiopia” and examined the candidate.

Accordingly this is to certify that the thesis has been accepted in partial fulfillment of the

requirement for the degree of Master of Science.

__________________ _______________ ____________

Name of the Chairperson Signature Date

_____________________ _________________ ____________

Name of Major Advisor Signature Date

________________________ _______________ ____________

Name of Internal Examiner Signature Date

________________________ _________________ ____________

Name of External Examiner Signature Date

i

Acknowledgement

First of all my everlasting thanks go for the Almighty God to his eternal love and kindness.

This thesis has been possible with the help of many peoples. My special thanks go to my two

advisors Dr. Motuma Tolera and Dr. Menfese Tadesse for their indispensable support in all

stage of my thesis. I deeply thanks them for devoting their precious time in reading,

commenting and correcting my thesis.

I am also thankful to Bale Zone agriculture bureau, Wereda offices in Goba, Harena Buluk

and Delo Mena, farmers and field guides for their immeasurable cooperation and participation

during the course of data collection.

Many thanks also go to the SHARE Bale project in Bale Eco-Region for facilitating and

financing this study and to Hawassa University for allowing me to stud my MSc. in Wondo

Genet

I thank my beloved friend Mr. Getachew Werkineh who took his precious time to review and

comment my thesis. And I am thankful to my family Ato Mezgebu Nigussie and W/ro Abebu

Asefa and my aunt Abonesh Nigussie for their endless support and encouragement.

At last but not the least I thank all my classmates in Wondo Genet College of Forestry and

Natural resource, Climate change and Development department for their support and

suggestions.

.

ii

Table of Contents

Acknowledgement ....................................................................................................................... i

Table of Contents ....................................................................................................................... ii

Dedication ................................................................................................................................... v

Acronyms and Abbreviations .................................................................................................... vi

List of Tables .......................................................................................................................... viii

List of Figures ............................................................................................................................ ix

List of Appendixes ...................................................................................................................... x

1 INTRODUCTION ............................................................................................................... 1

1.1 Background .................................................................................................................. 1

1.2 Statement of the Problem ............................................................................................. 3

1.3 Objectives of the Study ................................................................................................ 5

1.3.1 General Objective ................................................................................................. 5

1.3.2 Specific Objectives ............................................................................................... 5

1.4 Research Questions ...................................................................................................... 6

1.5 Significance of the Study ............................................................................................. 6

1.6 Scope of the Study ....................................................................................................... 7

2 LITERATURE REVIEW .................................................................................................... 8

2.1 Land Use/ Land Cover Change: Definition and Concepts ........................................... 8

2.2 Drivers of Land Use/Land Cover Change .................................................................... 9

2.2.1 Proximate Causes ................................................................................................ 10

2.2.2 Underlying Causes .............................................................................................. 10

2.3 Application of Remote Sensing and GIS Techniques for LU/LCC Study................. 13

2.4 Modeling Land Use/Land Cover Change .................................................................. 14

iii

2.4.1 Markov Chain Modeling ..................................................................................... 15

2.5 LU/LCC at Global Perspective .................................................................................. 17

2.6 State of LU/LCC in Ethiopia...................................................................................... 18

3 METHODS AND MATERIALS ...................................................................................... 21

3.1 Description of Study Area .......................................................................................... 21

3.1.1 Location .............................................................................................................. 21

3.1.2 Demographic and Socio-Economic Characteristics ........................................... 22

3.1.3 Biophysical Characteristics................................................................................. 23

3.2 Study Site Selection Procedures................................................................................. 25

3.3 Sources and Types of Data ......................................................................................... 27

3.4 Data Acquisition......................................................................................................... 27

3.4.1 Satellite Image and GIS Data Collection ............................................................ 27

3.4.2 Field Survey and Data Collection ....................................................................... 29

3.5 Data Analysis ............................................................................................................. 32

3.5.1 Satellite Image Analysis ..................................................................................... 32

3.5.2 Data Analysis for Driving Forces of LU/LCC ................................................... 35

3.6 Land Use/Land Cover Change Modeling .................................................................. 35

3.7 Data presentation ........................................................................................................ 36

4 RESULTS AND DISCUSSION........................................................................................ 38

4.1 Characteristics of LU/LC units .................................................................................. 38

4.2 Land use/Land Covers of the Study Area in 1986, 1996, 2006 and 2016 ................. 40

4.3 Land Use/Land Cover Change Detection .................................................................. 42

4.3.1 Trend of LU/LCC in Bale Eco-Region ............................................................... 42

4.3.2 Land Use/Land Cover Change Matrix ................................................................ 48

4.4 LU/LCC under different institutional set-up in Bale Eco-Region ............................. 49

iv

4.5 Predicting LU/LCC Based on the Markov Model ..................................................... 53

4.6 Causes of LU/LCCs in Bale Eco-Region ................................................................... 54

4.6.1 Proximate (Direct) Causes .................................................................................. 54

4.6.2 Underlying Causes .............................................................................................. 58

5 CONCLUSION AND RECOMMENDATION ................................................................ 67

5.1 Conclusion ................................................................................................................. 67

5.2 Recommendation........................................................................................................ 69

6 References ......................................................................................................................... 71

v

Dedication

To my beloved friend Getachew Werkineh and my family Ato Mezgebu Nigussie and W/ro

Abebu Asefa

vi

Acronyms and Abbreviations

BERSMP Bale Eco-Region Sustainable Management Plan

BER Bale Eco-Region

BMNP Bale Mountain National Park

CSA Central Statistical Agency

DAs Development Agents

ETM+ Enhanced Thematic Mapper Plus

FDRE Federal Democratic Republic of Ethiopia

FGD Focus Group Discussion

FZS Frankfurt Zoological Society

GIS Geographic Information System

GPS Global Positioning System

GTP Ground Truthing Points

IWMI International Water Management Institute

KII Key Informant Interview

LU/LCC Land Use/Land Cover Change

LU/LC Land Use/Land Cover

MCM Markov Chain Model

vii

MoME Ministry of Mines and Energy

MSS Multispectral Scanner

NTFPs Non-Timber Forest Products

OFWE Oromia Forest and Wildlife Enterprise

OLI/TIRS Operational Land Imager/Thermal Infrared Sensor

PA Peasant Association

PFM Participatory Forest Management

PHEEC People Health and Environment Ethiopia Consortium

PRM Participatory Rangeland Management

RS Remote Sensing

SHARE European Union’s Support for Horn of Africa Resilience

TM Thematic Mapper

USGS-EROS United State Geological- Earth Resource Observation and Science

UTM Universal Transverse Mercator

WGS84 World Geodetic System 84

viii

List of Tables

Table 1: Summary of spatial data sets used in this study ......................................................... 28

Table 2: Software used in the course of the study .................................................................... 29

Table 3: Description of major LU/LC types identified in Bale Eco-Region ............................ 39

Table 4: Rate and percentage change of LU/LCs in Bale Eco-Region .................................... 44

Table 5: LU/LCC under different institutional arrangements in BER, Note: NP = National

park, W.admin = Wereda administration, PFM = participatory forest management

and PRM = participatory rangeland management. .................................................. 50

Table 6: Underlying Causes of LU/LCC in BER ..................................................................... 59

Table 7: LU/LCC matrix between 1986 and 1996 ................................................................... 86

Table 8: LU/LCC matrix between 1996 and 2006 ................................................................... 86

Table 9: LU/LCC matrix between 2006 and 2016 ................................................................... 87

Table 10: Transitional probability matrix derived from LU/LC map of 2006 and 2016 .......... 87

Table 11: LU/LCC matrix between 1986 and 2016 ................................................................. 88

Table 12: Error matrix for the LU/LC map of 1986 ................................................................. 89

Table 13: Error matrix for the LU/LC map of 1996 ................................................................. 90

Table 14: Error matrix for LU/LC map of 2006 ....................................................................... 91

Table 15: Error matrix for LU/LC map of 2016 ....................................................................... 92

ix

List of Figures

Figure 1: Study area map. Note: the intervention Weredas with olive color in figure (d)

indicates Weredas used for LU/LCC analysis, whereas sample Weredas and

Kebeles describe in figure (a) indicates Weredas and Kebeles selected for filed data

collection. BER in figure (d) represents Bale Eco-Region. ..................................... 22

Figure 2: Flow chart that shows the general methodology of this research. Adopted from Sang

(2010) and Shiferaw (2011) with some modification .............................................. 37

Figure 3: Area of LU/LC units at different periods in Bale Eco-Region ................................. 40

Figure 4: Map of LU/LC types of Bale Eco-Region produced based on unprocessed satellite

images obtained from USGS.................................................................................... 41

Figure 5: Trend of LU/LCC in Bale Eco-region....................................................................... 43

Figure 6: Map that shows LU/LCC across different intuitions in BER.................................... 53

Figure 7: Population growth in seven Weredas of BER (1994-2016) drived from CSA (Central

Statistical Agency). Note: Due to the data gap from the CSA total population for the

years between 1994 and 2004 and 2009 were not available for use. ....................... 61

x

List of Appendixes

Appendix 1: Land Use/Land Cover Change Matrixes .............................................................. 86

Appendix 2: Error Matrixes ...................................................................................................... 89

Appendix 3: Field Observation Sheet Format .......................................................................... 93

Appendix 4: Checklist for Focus Group Discussion and Key Informant Interview ................. 93

xi

Land Use/Land Cover Changes and Associated Driving Forces in Bale Eco-Region,

Ethiopia

Adane Mezgebu Nigussie

Mobile phone: +251-9-18470641 E-mail: adanelove25@gmail.com

Abstract

Land Use/ Land Cover Change (LU/LCC) is one of the major human induced global changes.

Information on LU/LCC and the forces and processes behind such changes are essential for

proper understanding of how land was being used in the past, what type of changes have

occurred and are expected in the future. This study was carried out to examine land use/land

cover changes and driving forces behind the changes in the Bale Eco-Region, Ethiopia. It was

conducted using satellite image of Landsat5 TM 1986 and 1996, Landsat7 ETM+ 2006 and

Landsat8 OIL/TIROS 2016. In addition, field observations, Key informant interview (KII) and

Focus Group Discussion (FGD) were also conducted. ERDAS Imagine 9.2, ArcGIS 9.3 and

IDRSI Selva 17.00, softwares were used for satellite image processing and map preparation

and LU/LCC prediction respectively. The main finding of this study revealed an expansion of

agriculture/settlement and reduction of woodland and forest over the last 30 years between

1986 and 2016. Agriculture/settlement increased by 173369 ha, with a corresponding 296692

ha and 85184 ha decline in the area of woodland and forest respectively. If the current rate of

LU/LCC continues, agriculture/settlement is predicted to increase by 28% in 2026. In contrast

woodland and forest are predicted to shrink by 24% and 17% respectively. Analysis of

LU/LCC under different institutional set-ups between 2006 and 2016 showed highest

expansion (1285 ha/year) of agriculture/settlement in the lowland Kebele under Wereda land

administration followed by 290 ha/year in Kebele under the national park. LU/LCC in the

BER is a result of several proximate and underlying drivers. The major proximate driving

forces of LU/LCC in the BER are agricultural expansion, fire, illegal logging and fuel wood

extraction, overgrazing and expansion of illegal and unplanned settlements. Demographic,

economic, technological, institution and policy, socio-cultural and biophysical factors

constitute the major underlying drivers of LU/LCC in the BER. Hence, the right policy

packages are required to control the expansion of agriculture at the expance of woodland and

forest resources in the study area.

Key words: Bale Eco-Region, Change prediction, Drivers, Institutions, Land use/Land cover

1

1 INTRODUCTION

1.1 Background

Land is an essential natural resource which has numerous social, economic, and biophysical

uses. It used to create wealth and employment, grow economies and also use as a source of

water, food and energy. It provides services such as conserving biodiversity, storing carbon,

purifying and storing water and regulating the Earth’s climate by absorbing the heat from the

sun as well (Molla, 2014; Sall, 2014). These services will continue if only the land is not

destroyed or degraded by human induced actions.

According to Agarwal et al. (2002) increase in atmospheric carbon- dioxide concentrations;

alterations in the biochemistry of the global nitrogen cycle; and on-going Land Use/ Land

Cover Change (LU/LCC) are the three major human induced global changes. LU/LCC is an

endless process taking place on the earth surface starting from ancient time (Shiferaw, 2011;

Worku et al., 2014). Expansion of agriculture to meet the demand of growing population such

as food and fiber at the expense of vegetated lands is the most significant historical change in

all parts of the world (Lambin et al., 2003). During the last 3 centuries around 1.2 million km2

of forest and woodland and 5.6 million km2 of grassland and pasture have been converted to

other uses globally, while cropland has increased by 12 million km2 (Agarwal et al., 2002).

Both natural and human activities are responsible for LU/LCC (Burka, 2008) while the latter

is increasingly recognized as a dominant force in LU/LCC (Lamichhane, 2008). Human

activities are responsible for the conversion and transformation of plentiful of the world’s

natural land covers (Hamza and Iyela, 2012). For instance, over the last 10,000 years, about

50% of the ice-free land surface has been changed by human activities (Lambin et al., 2003).

2

Since 1850 around 6 million km2 and 4.7 million km2 of forest/woodland and grassland areas

have been converted to agricultural land worldwide in that order, to meet the demand for food

and fiber (Lambin et al., 2003; Hamza and Iyela, 2012).

LU/LCC has negative consequences on both the quality of environment and life (Molla,

2014). Gashaw and Dinkayoh (2015) noted that the environmental consequences of LU/LCC

are as large as that of climate change. LU/LCC can affect food security, biodiversity,

biogeochemical cycles, soil fertility, hydrological cycles, energy balance, land productivity

and the sustainability of environmental service provision (Burka, 2008; Molla, 2014). Apart

from these, it also contributes to global warming (Molla, 2014).

Assessment of spatial and temporal distribution of LU/LC is essential pre-requisite for land

resources planning, management and monitoring programs (Mani and Krishnan, 2013). Ebro

et al. (2011) and Tefera (2011) have also noted that, precise information on LU/LCC and its

driving forces are essential to understand what type of changes have occurred and are

expected in the future. Moreover, analysis of LU/LCC and its drivers provides important

information for monitoring biodiversity loss and natural disasters (e.g. drought, floods,

wildfires), and for identifying areas threatened with severe land degradation (e.g.

deforestation, overgrazing, diversion of water resources, etc.) (Ebro et al., 2011).

GIS and Remote sensing technologies has made possible to assess and analyze LU/LCC in

less time, at low cost and with better accuracy (Abdullah et al., 2013; Mani and Krishnan,

2013). Availability of remote sensed data in various spatial and temporal resolutions made

mapping and assessing LU/LCC possible (Mani and Krishnan, 2013). On the other hand GIS

3

has tools for collecting, storing, analyzing and visualization of the outcome of analysis (Reis,

2008).

Being the second populous country in Africa, Ethiopia is experiencing enormous LU/LCC

(Kindu et al., 2013; Gashaw and Dinkayoh, 2015). These changes are mainly from natural

vegetation land to agricultural land and settlement. The LU/LCC problem is more severe in

the highlands of Ethiopia (Eshetu and Hogberg, 2000). It is because these areas were

characterized by high population pressure and cultivated for long period of time (Kindu et al.,

2013).The highland areas in Ethiopia cover nearly 45% of the country’s landmass (Tefera,

2011). Different studies have been conducted to quantify LU/LCC in both highland and

lowland parts of Ethiopia (Belay, 2002; Woldeamlak, 2002; Tefera, 2011; Kindu et al., 2013;

Molla, 2014; Alemu et al., 2015). According to these studies, the country is characterized by

reduction of forest, woodlands, grasslands and shrub lands, but a remarkable expansion of

agricultural land and bare lands in space and time. These studies also identified population

growth owing to natural increase, in-migration and resettlement; overgrazing by livestock;

climate change; land tenure arrangements; livelihood strategies and commercial agricultural

investments as the main driving forces for LU/LCC in Ethiopia.

1.2 Statement of the Problem

Bale Eco-Region (BER) which is characterized by wealth of biodiversity and ecosystem

services is one of the most important eco-region in Ethiopia and Sub-Saharan Africa (FARM

Africa, 2008). It is home of several fauna and flora including the endangered and endemic

species (FARM Africa, 2008). According to Watson (2013) the BER is one of 34 global

biodiversity hotspots which contain more than 1,500 species of vascular plants as endemics.

4

The Bale Mountains National Park (BMNP) which lies at the heart of the BER is also one of

the most important conservation areas in Ethiopia (Watson, 2013). The livelihood of millions

of people living in the eco-region depends on the ecosystem services provided from the eco-

region. The region is also well known in its water resources. These water resources are critical

for the livelihoods and wellbeing of hundreds of thousands of people in the highlands of

Southeast Ethiopia and an estimated 12 million people in the lowlands of Southeast Ethiopia,

Northern Kenya and Somalia (FARM Africa, 2008).

However, this globally important eco-region is under increasing threat from a growing human

population, fire and rapid immigration with unplanned and unrestricted settlement (SOS Sahel

Ethiopia, 2010; Teshoma, 2010). Disturbance of the water systems and deforestation and

forest degradation are occurring due to poor management of natural resources, the conversion

of natural habitat to farmland, overgrazing by livestock and unsustainable fuel wood and

timber extraction (Teshoma, 2010; Hailemariam et al., 2015). These, coupled with impacts

from climate change are influencing its unique flora and fauna and both lowland and highland

communities that depend on the BER’s ecosystem services (FARM Africa, 2008).

Gebrehiwet (2004), Oumer (2009), Molla et al. (2010), Muluneh (2010), Shiferaw (2011),

Tefera (2011), Gebreslassie (2014), Molla (2014), Tsegaye (2014), Alemu et al. (2015),

Gashaw and Dinkayoh (2015) have studied LU/LCC in different parts of Ethiopia using GIS

and Remote Sensing techniques. Few studies like Morie (2007), Walellegn (2007), and

Teshome et al. (2008) have been conducted to understand LU/LCC in BER. Even though,

these studies are found in the study area they dealt with quantifying LU/LCC using remote

sensing tools, which give quantitative descriptions, but fail to assess the drivers of LU/LCC.

5

Then again studies those links LU/LCC under different institutional arrangements of natural

resource management and predict how LU/LCC will unfold in the future are lacking in study

area. Therefore in view of the literature gaps indicated above this research analyzed the

LU/LCC and driving forces of change in BER from 1986 to 2016. It also analyzed LU/LCC

under different institutional arrangements (Federal/Bale Mountains National Park, Oromia

regional government, Participatory Forest Management (PFM) and Participatory Rangeland

Management (PRM)) from 2006 to 2016. The research predicted future LU/LCC in study area

as well.

1.3 Objectives of the Study

1.3.1 General Objective

This study generally aimed at assessing land use/land cover changes and driving forces behind

the changes in the Bale Eco-Region, Ethiopia.

1.3.2 Specific Objectives

The specific objectives of this study were to:

1. Assess historical land use/land cover change in the Bale Eco- Region between

1986 and 2016

2. Assess land use/land cover change under different institutional arrangements of

natural resource management in Bale Eco-Region between 2006 and 2016

3. Predict future land use/land cover changes in the Bale Eco- Region (from 2016 to

2026)

6

4. Identify major driving forces of land use/land cover changes in the Bale Eco-

Region

1.4 Research Questions

The Study has tried to answer the following research questions

1. What is the historical trend of land use/land cover change in the Bale Eco-Region

between 1986 and 2016?

2. What is the land use/land cover change look like under different institutional

arrangements of natural resource management in Bale Eco-Region?

3. What will the future land use/ land cover change look like in the Bale Eco-Region?

4. What are the major driving forces to land use/land cover change in the Bale Eco-

Region?

1.5 Significance of the Study

The study will have its own rationalities both for study site in one way and for

LU/LCC literatures. By analyzing the LU/LCC trend and driving forces behind the

changes, it helps to understand how land was being used in the past, what type of

changes have occurred and are expected in the future. Moreover it can provide data to

policy and decision makers to design appropriate policies and strategies for monitoring

resource degradiation and promote sustainable management of natural resources. More

sustainable management of natural resources in turn can enhance agricultural

productivity and builds the resilience of rural communities to shocks. A large number

7

of government or non-government development agencies, researchers and local

communities can benefit from the outputs of this research.

1.6 Scope of the Study

Conceptually, the study assessed LU/LCC and drivers of the change, and then it predicted

future LU/LCC. In its geographic scope it was undertaken in Bale Eco- Region covering an

area of 1,577,067 ha. Methodologically the study used different methods and approaches

(remote sensing, field observation, key informant interview, focus group discussion, and use

of secondary data). For predicting future LU/LCC the study used Markov chain Model.

8

2 LITERATURE REVIEW

2.1 Land Use/ Land Cover Change: Definition and Concepts

Land cover and land use are the two interrelated ways of observing earth’s surface (Duhamel,

2011). The former represents the biophysical state of the earth’s surface and immediate

subsurface, while the later indicates the manner human population manipulate the biophysical

attributes of the land and the purpose for which land is used (Meyer, 1995; Turner et al., 1995;

Lambin et al., 2003; Pellikka, 2008; Duhamel, 2011). Some examples of land use are grazing,

recreation, agriculture, urban development, logging and mining (Opeyemi, 2006).

The relationship between land use and land cover can be described as: change in land use can

affect and be affected by land cover, however the change in either of them is not necessarily

the product of the other. Single land use system may correspond to a single land cover or it

may involve several distinct covers (Briassoulis, 2011). For instance a farming system may

involve several distinct covers such as cultivated land, woodlots, improved pasture, and

settlements. On the other hand single class of cover may support multiple uses. For example

the area covered by natural forest can be used for hunting and gathering, fuel wood collection,

recreation and wild life preservation (Briassoulis, 2011; Verheye, 2011).

A change in land cover refers to conversion of one land cover type to a new cover type or

modification within one land cover category (Meyer and Turner, 1992; Lambin and Geist,

2003). On the other hand land use change refers to a conversion of land use due to the

interference of human being for different purposes such as for settlement, infrastructural

development, agriculture and recreational uses (Meyer and Turner, 1992; Turner et al., 1995).

9

LU/LCC refers to the human modification and conversion of the earth terrestrial surface

(Lambin et al., 2003; Hamza and Iyela, 2012; Shrestha, 2012). Modification occurs when the

change affect only the characteristics of the land cover without causing a complete shift from

one LU/LC type to the other. On the other hand conversion of LU/LC occurs when one

LU/LC type completely replaced by another (Turner et al., 1995; Lambin et al., 2003; Alemu

et al., 2015).

2.2 Drivers of Land Use/Land Cover Change

The world’s land surface estimated to cover about 13,340 million ha (Verheye, 2011). Of

which 54% of this land surface disturbed by both human activities and natural factors

(Briassoulis, 2011). The change in LU/LC at all level is associated with several natural and

human induced factors (Rahdary et al., 2008). The natural or biophysical causes of LU/LCC

include: slop, climate change, soil type, wildfire, pest infestation, flood and drought (Garedew,

2010; Shiferaw, 2011). Human induced or anthropogenic driving forces of LU/LCC grouped

as the direct effects of human activity (proximate causes) and indirect effects of human

activity (underlying driving forces) (EPA, 1999 report cited in Morie, 2007). The former

comprises agricultural expansion, wood extraction and infrastructure expansion while the later

includes demographic, economic, technological, policy and institutional and cultural factors

(Geist and Lambin, 2002). The human induced causative factors increasingly recognized as a

dominant force in LU/LCC (Lamichhane, 2008; Chang-Martínez et al., 2015). According to

Briassoulis (2011) one-third to one-half of the global land surface change by human activities

such as logging, agricultural expansion, over grazing, fire management, forest harvesting and

urban and suburban construction and development.

10

2.2.1 Proximate Causes

Proximate causes of LU/LCC are immediate actions of local communities and directly exerted

on land resources due to different underlying causes such as economic, social, political, etc

(Geist and Lambin, 2002; Shiferaw, 2011). They operate at the local level (individual farms,

households or communities) and explain how and why local land covers and ecosystem

processes are modified and converted directly by humans (Lambin et al., 2003; Lambin and

Geist, 2007).

According to Geist and Lambin (2002) agricultural expansion, wood extraction and

infrastructure expansion are major proximate causes of LU/LCC. De Sherbinin (2002)

explained that agricultural expansion is the dominant proximate cause for LU/LCC.

Agricultural expansion comprises permanent cultivation (large scale, smallholder subsistence

and commercial), shifting cultivation (slash & burn) and cattle ranching (large-scale and

smallholder) (Geist and Lambin, 2002). Crop land and pastures are now among the dominant

ecosystems on the planet, occupying more than 35% of the world’s ice-free land surface (Paul

and Lisa, 2011). Over 50% of the global agricultural lands increased in the past 100 years. In

the developing world, half of the land cover conversion occurred in just the past 50 years

(Houghton, 1994). In Ethiopia large areas, which were once under vegetation cover are now

changed to cultivated land and expose to soil erosion resulting into environmental degradation

and serious threat to the land (Amare, 2007).

2.2.2 Underlying Causes

Underlying causes of LU/LCC involves the structural (or systemic) factors that trigger the

proximate causes (Geist and Lambin, 2002; Lambin et al., 2003). They operate at the regional

11

(districts, provinces, or country) or even global levels by changing one or more proximate

causes (Lambin et al., 2003; Lambin and Geist, 2007). They are external to the local

communities and not controlled at the local level. According to De Sherbinin (2002) and

Geist and Lambin (2001, 2002) underlying causes of LU/LCC originate from a complex

interaction of social, policy and institutional, economic, demographic, technological, cultural

and biophysical factors.

Economic factor is one of the major underlying causes of LU/LCC particular for tropical

deforestation (Geist and Lambin, 2002). Economic variables such as low domestic costs (for

land, labor, fuel or timber), increase in product price (mostly for cash crops) influence land

use decision making, thereby impacting the land cover (Geist and Lambin, 2002). Besides

these, change in prices, taxes, and subsidies on land use inputs and products, change in the

costs of production and transportation and access to credit, market, and technology also plays

vital role in LU/LCC ( Lambin and Geist, 2007).

Political, legal, economic and traditional institutions and their interaction with individual

decision making also influence LU/LCC (Lambin and Geist, 2003; Lambin and Geist, 2007).

Institutional causes of LU/LCC must be considered both at large scale (international or

national level) and local level (Lambin et al., 2003). This is because the implementation of

large scale policies is practiced at local level and local people’s access to land, capital,

technology and information influenced by the structure of both local and large scale policies

(Lambin and Geist, 2003). On the other hand LU/LCC influenced significantly when local

institutions are undermined by large scale institutions (Lambin et al., 2003). Policy and

institutional cause of LU/LCC include; land tenure system, shift in land holding system from

12

communal (traditional) to formal (state), government policies on land use and economic

development, property-rights, environmental policies, decision-making systems for resource

management (e.g., decentralized, democratized, state-controlled, local and communal) and

social networks concerning distribution and access to resources (Geist and Lambin, 2002;

Lambin et al., 2003; Lambin and Geist, 2007).

Agro-technological changes such as the adoption of mechanized large scale agriculture,

modification of farming systems through intensification and extensification and poor

technological applications in the wood sector (leading to wasteful logging practices) are

amidst technological factors causing LU/LCC, particularly tropical deforestation (Geist and

Lambin, 2002).

Demographic changes are the dominant causes of LU/LCC in most of Africa, Asia and

L/America countries (Turner and Meyer, 1994). Demographic change include shifts in fertility

and mortality, changes in household structure, the breakdown of extended families into

multiple nuclear families and dynamics including; labor availability, migration, urbanization

(Geist and Lambin, 2002).

Cultural factors encompass motivations, collective memories, personal histories, attitudes,

values, beliefs, and perceptions of individuals, communities and land managers. These factors

influence land use decisions and land covers, sometimes profoundly (Lambin and Geist,

2007).

13

2.3 Application of Remote Sensing and GIS Techniques for LU/LCC Study

Geographical Information Systems (GIS) in conjunction with Remote Sensing (RS) has been

recognized as a powerful and effective tool in LU/LCC analysis (Weng, 2002; Rimal, 2011;

Abdullah et al., 2013). They provide accurate, cost effective and timely information and

methods for monitoring, modeling and mapping of LU/LCC across a range of spatial and

temporal scales. The information from GIS and RS also helps to assess the extent, direction,

causes, and effects of the LU/LCC (Reis, 2008; Oumer, 2009; Rimal, 2011). In LU/LCC

assessment some studies have utilized RS techniques; others have integrated remote sensing

techniques with GIS. GIS is the technology which has been used to view and analyze data

from a geographic perspective (Rimal, 2011).

It is a useful tool to measure the LU/LCC trends between two or more time by using statistical

and analytical functions (Abdullah et al., 2013). It provides a flexible environment for

collecting, storing, displaying and analyzing digital data necessary for LU/LCC detection and

tools for land use planning and modeling (Reis, 2008; Rimal, 2011). In the context of

LU/LCC, RS means the ability to detect change on the earth’s surface through space-borne

sensors (Abdullah et al., 2013). RS becomes useful tool for understanding landscape

dynamics over time and space, irrespective of the causal factors. This is because of the fact

that it provides multi-temporal and multi- spectral remotely sensed data (Oumer, 2009; Rimal,

2011). Application of RS for LU/LCC analysis depends on: (i) sensor capability, (ii) wealth of

information captured, (iii) objective of the intended study and (iv) spatial and spectral

properties of satellite images acquired by different versions of a particular sensor instrument

(Oumer, 2009). Landsat imagery provides a better understanding of land resources. The most

14

important reason for this is a continuous improvement in radiometric and spectral property of

images over time (Oumer, 2009). Since the starting of Landsat program in 1972 Landsat

Multispectral Scanner (MSS), Thematic Mapper (TM) and Enhanced Thematic Mapper Plus

(ETM+) data have been broadly employed in LU/LCC studies, mainly in forest and

agricultural areas (Reis, 2008).

2.4 Modeling Land Use/Land Cover Change

Models are used in a variety of fields including in LU/LCC studies (Brown et al., 2004). In

LU/LCC studies models have been developed to address the questions when, where and why

LU/LCC occurs (Brown et al., 2000). They usually involve empirically fitting the models to

some historical pattern of change, then extending those patterns into the future for prediction.

Modeling LU/LCC plays a significant role for understanding the factors that cause LU/LCC

and the impacts of the changes (Araya, 2009; Adedeji et al., 2015). Models are useful for

sorting out the complex groups of socio-economic and biophysical forces that influence the

rate and spatial pattern of LU/LCC and for estimating the impacts of changes in LU/LCC

(Verburg, 2004).

Furthermore, models can support prediction of future LU/LCCs under different scenario

conditions, based on past evidence (recent past) (Verburg, 2004; Chang-Martínez et al., 2015).

Models also enable the projection into the future of the expected effects of governmental

programs aiming at the conservation and utilization of resources. Assessing and predicting

LU/LCC would help for effective environmental management and sustainable resources use.

Additionally it would help to better understand the functioning of the LU/LC system and to

support land use planning and policy as well as development plans and decision making

15

(Araya, 2009). Numbers of LU/LCC modeling approaches have been described in different

literatures (Veldkamp and Lambin, 2001). Depending on the purpose they are designed, their

spatial and temporal scale, data availability, expertise knowledge and etc. there exist different

LU/LCC modeling approaches and they may differ from each other (Araya, 2009). There are

technically complex models which require advanced programming and expertise knowledge,

while some others are simple and provide user-friendly tools that someone can apply with

limited experiences, e.g. markove chain model (Veldkamp and Lambin, 2001; Araya, 2009).

2.4.1 Markov Chain Modeling

Markov chain model (MCM) has been used extensively for urban and rural LU/LCC modeling

(Brown et al., 2000; Arsanjani, 2012). It is a discrete-time stochastic model, describing the

probabilistic movement of one state (LU/LC type) to another state (LU/LC type) (Arsanjani,

2012; Sayemuzzaman and Jha, 2014; Iacono et al., 2015). MCM consider LU/LCC as a

stochastic process and different LU/LC categories are the states of a chain (Weng, 2002). The

model specifies both time and a finite set of states as discrete values (Iacono et al., 2015). The

applicability of MCM in LU/LCC modeling is promising because of its ability: (i) to represent

all of the possible directions of LU/LCC among all of the land use categories, (ii) to quantify

the states of conversion between LU/LC types and (iii) to quantify the rate of conversion

among the LU/LC types (Sang, 2011; Han et al., 2015).

For LU/LCC prediction MCM utilizes two historic LU/LC images as input and produces a

transition probability matrix, a transition areas matrix and a set of conditional probability

images (Eastman, 2006; Sayemuzzaman and Jha, 2014). The former represents the probability

that each land use/land cover category will change to every other category over the specified

16

number of time units, while the later represents the amount of pixels that are anticipated to

change from each land use/land cover type to each other land use/land cover type over the

specified number of time units (IDRISI Seva help system; Sang, 2011; Arsanjani, 2012). The

conditional probability images report the probability that each land cover type would be found

at each pixel after the specified number of time units (IDRISI Selva help system)

According to different authors (Weng, 2001; Sayemuzzaman and Jha, 2014; Adedeji et al.,

2015; Iacono et al., 2015; Mirkatouli et al., 2015) MCM have several assumptions. One basic

assumption is that a future state of LU/LC at a time (t+1) can be determined only as a function

of its current state (t). Mathematically this can be expressed as . Path of past

states Xt-1, Xt-2, Xt-3,…X0 that the process passed through in arriving at does not determine

the future state at It also assume that the observed trends of LU/LCC will remain the

same (stationary process), thus allowing their projection to the future.

The current state of LU/LC distributions (Xt) and the future state of LU/LC distributions at

(Xt + 1) time period, as well as a transition probability matrix (Pij) representing, m × m matrix

which expresses the probability that a site in state i at time t will transfer to state j at time t+1,

are used to construct the Markov model (Brown et al., 2000; Sang, 2011; Adedeji et al., 2015;

Han et al., 2015) which is expressed as follows:

Where Xt + 1 and Xt Represent the states of land use at given point t + 1 and t, respectively. The

matrix P is row-standardized, such that the sum of transition probabilities from a given state is

always equal to one.

17

2.5 LU/LCC at Global Perspective

LU/LCC occurs at local, regional and global scales and changes at local scales can have

cumulative impacts at broader scales (Burka, 2008). LU/LCC is as old as the age of human

kinds (Gebreslassie, 2014). Human land use activities spread over about 50% of the ice free

land surface starting from the control over fire and domestication of animals and plants

(Lambin and Geist, 2006). The spread of human land use activities were mainly at the expance

of forest lands, resulting reduction of global forest cover from 50% to less than 30% (Lambin

et al., 2003). Between 1700 and 1980 for instance, at the global scale total cultivated land was

estimated to have increased by 46% (Turner et al., 1992 Cited in Muluneh, 2003).

Furthermore, between the years 1700 and 1990, the area under cropland and pasture has

increased from an estimated 300-400 million ha to 1500-1800 million ha, and 500 million ha

to 3100 million ha respectively. This resulted in reduction of forests from around 5000-6200

million ha in 1700 to 4300-5300 million ha in 1990 and natural grasslands, steppes and

savannas from around 3200 million ha in 1700 to 1800- 2700 million ha in 1990 (Lambin et

al., 2003).

However LU/LCC is not homogeneous across all parts of the world (WIREs Climate Change,

2014). There are regions still with relatively undisturbed land cover such as parts of tropics

and Polar Regions. Some other regions experienced huge LU/LCC mainly expansion of

agricultural lands at the expance of vegetated lands. This is true in regions with a shorter

history of human development such as Africa, south and Southeast Asia and Latin America

countries. Since 1850, these regions have experienced dramatic increases in cropland,

especially during the second half of the twentieth century (Lambin et al., 2003; WIREs

18

Climate Change, 2014). On the other hand expansion of agricultural lands reduced in

European countries (WIREs Climate Change, 2014).

2.6 State of LU/LCC in Ethiopia

As per Ministry of Mines and Energy (MoME, 2003) the total area of Ethiopia covers above

1.12 million km2. About 55% of this area is below 1500m a.m.s.l. which is lowland, whereas

the remaining 45% of the area, with an altitude of greater than 1500m is highland (Tefera,

2011). In Ethiopia the land is dominantly used for mixed farming system, by smallholders

who farm for subsistence (Tefera, 2011; Geremew, 2013).

The country also kwon by several environmental, climatic, and socio-economic problems such

as: environmental degradation, erratic rainfall, recurrent droughts and drought-related

distressing famines, prevalence of malaria and HIV/AIDS, widespread poverty and poor

governance (Tefera, 2011). The aforementioned problems are directly or indirectly linked

with Climate change and LU/LCC.

LU/LCC including forest cover change is one of the major environmental problems in

Ethiopia (Alemu et al., 2015). Albeit, most of the researches were conducted in the northern

highland, there are numbers of LU/LCC studies carried out in Ethiopia, at catchment, zone,

watershed and village levels. For instance Zeleke and Hurni (2001) in Dembecha area of

Gojjam; Woldeamlak (2002) in Chemago watershed, Gojjam; Garedew (2010) in the Semi-

Arid Areas of Central Rift Valley of Ethiopia; Gebrehiwot et al. (2010) in Koga watershed at

the headwaters of the Blue Nile Basin; Tsegaye et al. (2010) in North eastern Afar range

lands; Ebro et al. (2011) in Adami Tulu and Fantale Weredas, in the rift valley of Ethiopia;

19

Tefera (2011) in Nonno Wereda, Central Ethiopia; Fentahun and Gashaw (2014) in Bantneka

Watershed, Ethiopia; Molla (2014) in Arsi Negele District, Central Rift Valley Region of

Ethiopia; Worku et al. (2014) in Ameleke Watershed, South Ethiopia; Gashaw and Dinkayoh

(2015) in Hulet Wogedamea Kebele, Northern Ethiopia. Most of these researches reported the

decline of grassland and natural vegetation including forests, shrub lands and woodlands due

to conversion to croplands, grazing lands, open areas and settlements areas. This idea is in line

with (Munchen, 2012) who stated that almost all LU/LCC studies conducted in Ethiopia have

common characteristics, such as expansion of agriculture land and the loss of natural

vegetation, combined with a loss of biodiversity.

In the highland parts of Ethiopia there was expansion of agriculture at the expance of

vegetated lands mainly shrub land, woodland and forest land since 1860s (Girma, 2014).

However according to the author expansion of agriculture at the expance of vegetated lands

worsened since 1980s.

In Ethiopia expansion of agricultural land and loss of natural vegetation are associated with

population growth, poor economic condition, unclear land tenure right and several other

biophysical and socio-political factors (Melaku, 2003). According to Sege (1994) and Turner

and Meyer (1994) in most developing countries including Africa, Asia and L/America

countries population growth and LU/LCC have a strong statistical correlation. In agreement

to these different studies undertaken in different parts of Ethiopia also reported population

growth as a major cause for LU/LCC. Population growth was the major cause for the

expansion of agriculture and reduction of vegetation covers in Ethiopian highlands (Muluneh,

2010), Borena Wereda South Wello Highland (Shiferaw, 2011); Nono Wereda, Central

20

Ethiopia (Tefera, 2011), West Guna Mountain South Gondar (Tsegaye, 2014) and Northwest

lowland of Ethiopia (Alemu et al., 2015). The total population of Ethiopia during the first

population and housing census (1984) was 39,868,572. However, during the census of 1994

and 2007 it increased to 53,477,265 and 73,918,505 respectively (Minale, 2012). This implies

that between 1984 and 2007 the total population of the country increased by more than 34

million persons. This population growth has led to expansion of agriculture and settlement by

clearing forest, grass and woodlands (Minale, 2012).

21

3 METHODS AND MATERIALS

3.1 Description of Study Area

3.1.1 Location

Bale Eco-Region (BER) situated within the Oromia Regional State, forming part of the Bale

and West Arsi Zonal administration in the south eastern Ethiopian Highlands (Watson, 2013).

The astronomical location of BER ranges between 5°16ꞌ54ꞌꞌ and 7°52ꞌ55ꞌꞌN latitude and

38°37ꞌ52ꞌꞌ and 41°13ꞌ0ꞌꞌE longitude. BER covers 2,217,600 ha over 16 Woredas: Adaba,

Agarfa, Berbere, Dinsho, Dodola, Gasera, Goba, Gololcha, Goro, Harena Bulluk, Kokosa,

Delo Mena, Nensebo, Mede Welabu, Gora Damole and Sinana (Hailemariam et al., 2015).

However, the intervention Weredas for LU/LCC analysis of this study were only seven

namely Adaba, Dinsho, Goba, Harena Bulluk, Delo Mena, Mede Welabu and Berbere. Hence

hereafter BER in this study refers only these seven Weredas. These Woredas composed of 120

Kebeles, which are the smallest local government unit.

22

Figure 1: Study area map. Note: the intervention Weredas with olive color in figure (d) indicates

Weredas used for LU/LCC analysis, whereas sample Weredas and Kebeles describe in figure (a)

indicates Weredas and Kebeles selected for filed data collection. BER in figure (d) represents Bale

Eco-Region.

3.1.2 Demographic and Socio-Economic Characteristics

As per the Central Statistical Agency (CSA, 2013) projected figure for 2016, Bale and West

Arsi Zonal administrations in which BER is situated together have a total population of

4,319,447 of which 50% are male. BER on the other hand has a total population of 2,071,862

23

of which 1,036,206 are male. The total population for the seven intervention Weredas is

878,493. About 797,123 (91%) and 81,370 (9%) of this population live in rural and urban

areas respectively. Small-scale subsistence agriculture using traditional technologies is the

major sector that supports the livelihood of households and communities in the area (World

Bank, 2007; Rosell, 2011). It contributes about 85% to household’s economy. Agriculture in

the BER involves two major activities: Farming and Livestock husbandry (BMNP GMP,

2007). Farmers in the region grow cereal crops (e.g. Corn, Barley, Teff and Wheat), cash

crops (chat, coffee and rice) and horticultural crops particularly vegetables (e.g. onion, potato

and cabbage). Cattle, Goats, Sheep, Horses, Mule and Donkeys are important livestock species

reared by farmers in the BER for destructive (skins, selling and meat), and non-destructive

(transport, ploughing, reproduction and milk) purposes (BMNP GMP, 2007). Rural

households also generate significant portions of their income from forest products including

firewood and charcoal and non-timber forest products (NTFPs) such as honey, Arabica Coffee

and medicinal plants (Hailemariam et al., 2015). According to the estimation by Tesfaye et al.

(2011) about 34% of per capita income in the BER obtained from forest and forest products.

This implies that farming, livestock husbandry and forest are the three main livelihood sources

of local communities in BER.

3.1.3 Biophysical Characteristics

BER is a high land area with a vertical expansion ranging from 1500 to 3500 meter above sea

level (FARM/SOS, 2007). However the central afro-alpine plateau of the eco-region reaches

more than 4000 meters above sea level. The climate of BER is influenced by low level

easterly winds from the Indian Ocean (Walelegn, 2007) and characterized by eight months

24

rainy season (March to October) and four months dry season (November to February)

(Hailemariam et al., 2015). The annual rainfall of the region ranges from 1000 to 1400mm

(Walelegn, 2007). Temperature records from the BER indicate that the wet seasons are

comparatively warm and the dry seasons are characterized by extremely cold night and warm

day time. The lowest recorded temperature at highest plateau of Bale (Sanette) was -15 ºc and

the maximum record was 26 ºc in Delo Mena Wereda. Similarly the lowest recorded

temperature in Dinsho area is -6 ºc (Hillman, 1986; Hailemariam et al., 2015).

BER, with the BMNP at its heart is the largest Afro-alpine area left in Africa and

characterized by forest areas, afro-alpine plateau, mountains and valleys, grasslands and

agricultural land (FARM Africa, 2008). The Afro-alpine plateau and forests in BER are home

of globally unique and diverse fauna and flora, including a significant number of rare and

endemic species (FARM Africa, 2008). BER is named as a ‘water tower’ of south-eastern

Ethiopia, Somalia and Northern Kenya. This is because over 40 streams and springs and five

major rivers namely; the Web, Wabi Shebele, Welmel, Dumal and Ganale arise from this area

supplying water for around 12 million people in the lowlands of southeast Ethiopia, Northern

Kenya and Somalia (FARM Africa, 2008; OFWE et al., 2014). The ecosystem also provides

several goods and services for millions of peoples living in the highland and lowland part of

the region. The Harena forest including its large genetic pool of wild Arabica Coffee and vast

carbon store is the second largest stand of moist tropical forest in Ethiopia (Watson, 2013).

Main soil types in the BER are Cambisols, Vertisols, Luvisols, Lithosols and Nitosols (OFWE

et al., 2014).

25

3.2 Study Site Selection Procedures

There are about 16 Weredas in the BER. Currently the SHARE Bale Eco-Region project

“sustainable biodiversity conservation and improved local livelihoods at Bale Eco-Region” is

running by a consortium of organizations, which are Farm Africa, SOS Sahel Ethiopia,

Frankfurt Zoological Society (FZS), Population Health Environment Ethiopia Consortium

(PHEEC) and International Water Management Institute (IWMI) in the Bale Eco-Region. This

project is implementing its activities in seven Weredas namely; Adaba, Dinsho, Goba

(Highland Woredas), Harena Buluk (Mid altitude Woreda) and Delo Mena, Meda Welabu,

Berbere (Low altitude Woredas). Accordingly this research focused on these intervention

Weredas with a financial support from SHARE Bale Eco-Region project.

For focus group discussion and key informant interview three Woredas were selected by using

multistage sampling technique. First, the existing seven Weredas were stratified agro-

ecologically into highland, midland and lowland Weredas. Then three Weredes, (one from

each agro-ecology) were selected purposively to represent the three agro-ecologies. Basically

these sample Weredas were selected based on four purposive criteria. The first was an area

where rapid population growth and associated natural resource degradation was observed,

secondly, livelihood strategies of the local communities (high dependency on forest resource

and livestock production), thirdly Weredas under PFM (participatory forest management) and

PRM (Participatory rangeland management) projects and national park interactions. This is

important for comparative analysis of LU/LCC under different institutional arrangements in

BER. The fourth criterion was their accessibility for the researcher. Information for the

aforementioned criteria was gathered during reconnaissance filed survey and from

26

unsupervised image classification information undertaken preliminary to major field survey.

Accordingly Goba (from highland Woredas), Harena Buluk (from mid altitude Woredas) and

Delo Mena (from low altitude Woredas) were sample geographic units selected for this study.

Each Weredas composed of Kebeles which are the smallest local administrative units. Out of

all Kebeles existing in the sampled Weredas, six Kebeles (two from each representative

Weredas) were selected purposely. Representative Kebeles were selected purposively based

on their membership in target project area of SHARE Bale Eco-region project and their

accessibility for the researcher.

For the analysis of LU/LCC under different institutional set-up in the BER seven sample

Kebeles were selected from the three agro-ecologies purposively. The main criteria used to

select sample Kebeles were representativeness of Kebeles to major LU/LC types namely;

woodland, forest and grazing land, existing natural resource management institutions such as

Federal government /BMNP, Oromia regional government land administration, PFM and

PRM and existing proximate drivers of LU/LCC such as agricultural expansion. Accordingly

Wagitu Shabe, Tosha, Rira from Goba Wereda (highland agro-ecology), Hawo and Shawe

from Harena Buluk Wereda (midland agro-ecology) and Berak and Naniga Dera from Delo

Mena Wereda (lowland agro-ecology) are sample Kebeles selected for LU/LCC analysis

under different institutional set-up. Wagitu Shabe and Shawe are Kebeles under PFM and

Tosha and Berak are under Wereda land administration. Rira and Naniga Dera are Kebeles

under the Federal government /BMNP and PRM respectively. Hawo is the only Kebele

situated under three institutional set-ups namely; Federal government /BMNP, PFM and

Wereda land administration.

27

3.3 Sources and Types of Data

To meet the objectives of this research, different kinds of data were collected from both

primary and secondary data sources. Data from primary sources include satellite imagery and

field data. Secondary data such as census records and unpublished official documents and

reports were gathered from Central Statistical Authority (CSA) of Ethiopia and offices of

Agricultural and Rural Development, and Land and Environmental Protection respectively.

Additionally past research works were used as supportive secondary data sources.

3.4 Data Acquisition

3.4.1 Satellite Image and GIS Data Collection

Time series Landsat images of 1986, 1996, 2006 and 2016 were used to analyze LU/LCC of

entire BER. The reason why images of these years were selected was in order to match the RS

data with major events undertaken during these years. These are the 1985/86 resettlement

program by the Derge government, implementation of participatory forest management (PFM)

by GIZ and Oromia region agricultural bureau starting from 1995 and finally implementation

of PFM by Farm Africa and SOS Sahel starting from 2006. On the other hand Landsat images

of 2006 and 2016 were used for analysis of LU/LCC under different institutional arrangements

(Federal government /BMNP, Oromia regional government land administration, PFM and

PRM). All data were collected from U.S. Geological Survey Center for Earth Resources

Observation and Science (USGS-EROS) (http://geography.usgs.gov) which comprised of the,

Landsat Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+) and Operational

Land Imager/Thermal Infrared Sensor (OLI/TIRS). All images used in this study had 30 m

spatial resolution and below 10% cloud cover. Detailed characteristics of the sources of data

28

used for the study are shown in Table 1. To reduce the effect of cloud cover and seasonal

variation on the classification result, the researcher tried to consider Landsat images of the

same season (December to February). As it was described in sub section 3.1.3 four months

from November to February are dry season in BER and have relatively cloud free sky.

Therefore images acquired during this season have relatively low cloud cover. Landsat image

data were preferred in LU/LCC analysis over other multispectral data for example; spot

images (Bakker et al., 2001). This is because of its free availability and inclusion of middle

infrared bands, a legal sharing of data among government department and donor agencies and

for having the longest record of global scale data for earth observation (Bakker et al., 2001;

Gilani et al., 2014).

Table 1: Summary of spatial data sets used in this study

Dataset type Acquisition

Date

Pixel Resolution

(m)/Scale

Path/Row Producer

Satellite data

Landsat TM 1986-01-14/ 21

&1996-01-23/

30

30m 167/ 055 & 056

& 168/055

&056

USGS

Landsat ETM+ 2006-12-12/ 19 30m 167/ 055 & 056

& 168/055

&056

USGS

Landsat

OLI/TIRS

2016-02-02 &

2016-01-24

30m 167/ 055 & 056

& 168/055

&056

USGS

Ancillary data

Field data March, 2016-

May, 2016

29

Table 2: Software used in the course of the study

Software Application

ARCGIS 9.3 Image processing and map preparation

ERDAS EMAGINE 9.2 Image preparation and processing

IDRSI Selva 17.00 LU/LCC prediction

MS EXCEL Statistical analysis and Chart and graphs

preparation

MS WORD Word processing

3.4.2 Field Survey and Data Collection

The required field data concerning the existing LU/LC types, historical trends in prevailing

LU/LCC and possible drivers of LU/LCC in the study area were collected by using different

data collection tools such as; Focus Group Discussion (FGD), Field Observation and Key

Informant Interview (KII). The field survey was conducted in two phases. The first phase was

a reconnaissance survey and it was conducted between March 29, 2016 and April 3, 2016 on 5

Weredas of the BER. The overall aim of this survey was to get a broad picture of the study

area and pretest FGD and KII checklists. The information obtained from the reconnaissance

survey was used to select sample Weredas and Kebeles and redesign FGD and KII checklists.

The second phase of the field survey was conducted starting from April 10, 2016 until end of

May, 2016.

30

I). Reconnaissance Survey

Before starting the fieldwork a reconnaissance survey was carried out in the month of March.

During the reconnaissance survey informal interviews and meetings were conducted with Bale

Zone agricultural bureau, Wereda pastoral office staff, Wereda natural resource management

experts, SHARE project site coordinators, Kebele administration staff, and elder peoples.

Information such as the current land use practices, natural resource management institutional

set-ups, status of forest and land resources and population and livelihood strategies in terms of

their pressure on land resources has been obtained during the reconnaissance survey.

II). Focus Group Discussion (FGD)

As suggested by many authors like Single (1996) cited in Bhawana (2015), Robinson (1999)

and Jayasekara (2012), based on theme of study and researchers interest the number of

participants in FGD can range from 4 to 10. Accordingly, groups containing 8-10 elderly

people were used in this study. One female group and two male groups from each sample

Kebele were formed. Totally 18 groups were formed from six Kebeles. The reason for

forming separate male and female groups was traditionally in the study area females are not

allowed to sit and speak out full in front of males. The participants of FGDs were selected

purposively from both sexes. Two purposive criteria were used to select participants in FGD.

The first was age of participants i.e. elder people who have lived long time in the study area

and had detail information about the past and present situations of the study sites. A second

criterion is capability to understand the topics and express their feelings and opinions. The

selection was performed with the help of Development Agents (DAs) and peasant associations

(PA) councils. The FGDs was guided by a list of questions as a checklist (Appendix IV). The

aim of FGD was to assess and analyze the extent and trend of changes that discussants

31

perceived to have occurred on their lands and their surroundings in the past 30 year period

between 1986 and 2016 and the driving forces behind such change. This can help to compare

discussants perception with GIS and remote sensing analysis.

III). Key Informant Interview (KII)

With the intention of obtaining in-depth information and cross-checking the data collected

from focus group discussions, few key informant interviews were conducted. In this, one elder

person from each sampled Kebeles, one Wereda Natural Resource Conservation and

Management Expert from each sampled Weredas, one Wereda Environmental and Land

Administration Office Coordinator from each sample Weredas, one PFM /PRM Coordinator

from each sample Weredas and one Kebele Administrator from each sample Kebeles were

involved. Totally, 15 key informants (5 from each Weredas) were selected. The selection of

elder key informants was executed using snowball sampling method with the help of FGD

participants.

IV). Field Observation

All studied Kebeles were visited with one local assistant from each Kebeles to gain insight of

knowledge. Field observation was carried out continuously throughout the data collection

period in the field. During field observation Ground Truthing Points (GTPs) and photos of

existing LU/LC types were collected by using handheld GPS (Global Positioning System) and

digital camera to aid different steps of image processing, classification and accuracy

assessment. This field observation also helped to validate information obtained from FGD and

KII.

32

3.5 Data Analysis

3.5.1 Satellite Image Analysis

Image pre-processing

The study area covers four scenes (Table 1). Therefore firstly the four scenes were

downloaded from USGS-EROS using the path and rows indicated in (Table 1). Secondly the

panchromatic bands in each scene were stacked together so as to produce a multispectral

image for each scene. Thirdly a mosaic of four scenes covering the study area was created by

merging the stacked image via mosaic operation. Subsequently geometric and radiometric

corrections and image enhancement were conducted. All the above mentioned pre-process

were performed on ERDAS IMAGINE 9.2 software prior to the image classification.

Geometric correction involves conversion of data to ground coordinates e.g. UTM by removal

of distortions from sensor geometry. Radiometric correction on the other hand involves

correcting unwanted sensor or atmospheric noise and correcting the data for sensor

irregularities (Seyoum, 2012). Image enhancement is to improve the appearance of the

imagery to assists in image analysis, classification and visual interpretation (Bakker et al.,

2001). Accordingly in order to have all the data in the same coordinate system and ensure

consistency between datasets during analysis all maps and satellite images used in this study

were projected to Universal Transverse Mercator (UTM) projection system Zone 37N and

datum of World Geodetic System 84 (WGS84). Radiometric corrections such as haze

reduction and cutting dark areas were executed as well. To increase the tonal distinction

between various features in the image and to enhance specific spatial patterns in an image two

enhancement function namely contrast stretching and spatial filtering were executed. Finally

33

after passing the aforementioned processes each image are clipped using the boundary of BER

and selected Kebeles for institutional analysis.

Image classification

Image classification involves categorizing raw remotely sensed satellite images into a fewer

number of individual LU/LC classes, based on the reflectance values (Adedeji et al., 2015).

This study used a hybrid classification method involving both unsupervised and supervised

image classification techniques. First unsupervised classification was carried out before field

work to understand the general LU/LC classes of the study area and to select sample training

sites for data collection during field work. This is because unsupervised classification is

automatic and requires little knowledge of the study area. And then after the field work

maximum likelihood supervised classification was carried out to categorize the images using

training sites. Training sites were defined by using original images, the results of unsupervised

classification, field study knowledge and ancillary data (Google Earth). Image classification

was performed by using ERDAS IMAGINE 9.2 software. Ground Truthing Points and

ancillary data (Google Earth) were used to define training sites for the recent image

(OLI/TIRS) classification. Training sites for classification of older images (TM and ETM+)

were defined based on the result of unsupervised classification, spectral values of recent image

and information obtained from elder peoples. In this study a total of 250 Ground Truthing

Points(GTPs) (50 from agriculture/settlement, 50 from forest, 50 from woodland, 50 from

grassland/rangeland and 50 from scrub/bush land) were collected from the field using hand

held global positioning system (GPS). Of the total GTPs collected during the field work 40%

or 100 GTPs (20 from each LU/LC types) were used to assist classification of recent year

34

image (OLI/TIRS), while the remaining 60% (150 GTPs) were used for classification

accuracy assessment of the 2016 image. For classification accuracy assessment of older

images (TM and ETM+) a total test samples of 150 were randomly selected from the original

mosaic image of the respective years.

Accuracy Assessment

To verify to what extent the produced classification is compatible with what actually exists on

the ground it is important to evaluate the accuracy of classification results. Accordingly error

matrix was produced for all images in this study. An error matrix is a square array of rows and

columns and presents the relationship between the classes in the classified and reference data.

The reference data used for accuracy assessment were obtained from GPS points during field

work and original mosaic image. The GPS points used in classification accuracy assessment

were independent of the ground truths used in the classification. Based on the error matrix

overall accuracy and kappa statistics were used to illustrate the classification accuracy.

Therefore an overall accuracy of 86%, 87%, 89% and 87% was achieved for the Landsat TM

of 1986 and 1996; Landsat ETM+ of 2006 and Landsat OLI/TIRS of 2016 respectively

(Appendix II). These imply excellent classifications of Landsat images.

LU/LCC Detection analysis

LU/LCC statistics were computed in three different ways:

1) Total LU/LCC in hectare calculated by

35

Where, Area is extent of each LU/LC type. Positive values suggest an increase whereas

negative values imply a decrease in extent.

2) Percentage LU/LCC calculated using the following equation:

Where, Area is extent of each LU/LC type. Positive values suggest an increase whereas

negative values imply a decrease in extent.

3) Rate of LU/LCC: computed using the following simple formula

Where, r, Q2, Q1 and t indicates rate of change, recent year LU/LC in ha, initial year LU/LC

in ha and interval year between initial and recent year in that order

3.5.2 Data Analysis for Driving Forces of LU/LCC

Data concerning the driving force of LU/LCC collected via FGD and KII were analyzed

qualitatively. Speech transcription and comprehension of speeches techniques were applied.

LU/LCC drivers category provided by Geist & Lambin (2001, 2002) were used to structure

the assessment and analysis of LU/LCC drivers.

3.6 Land Use/Land Cover Change Modeling

Markov chain model implemented in IDIRIS Selva were used to predict LU/LCC for the year

2026. MCM provides transition probability matrix, a transition areas matrix and a set of

conditional probability images. The former represents the probability that each land cover

36

category will change to every other category and the later represents the number of pixels that

are expected to change from each land cover type to each other land cover type over the

specified number of time units. The conditional probability images report the probability that

each land cover type would be found at each pixel after the specified number of time units

(Sang, 2011).

Prior to predicting future LU/LC in 2026 the predictive power of the model was first validated

by predicting the LU/LC for the year 2016. Araya (2009) stated that comparing the result of

the model prediction for time t2 (in this case 2016) to the real map of time t2 (2016) is the only

way to quantify the predictive power of the model. Accordingly the LU/LC for the year 2016

was predicted considering the LU/LC map of 1996 and 2006. This helped to compare the

result of the prediction with the actual LU/LC in 2016. After validating the performance of the

model, a real “prediction” for the year 2026 was carried out. LU/LCC maps for the year 2006

and 2016 were used to predict the land requirement in 2026. This is because the model

determines the prediction result based on the observed pattern between the initial year (2006)

and base year (2016). The year 2026 is selected for prediction, because Markov chain model

requires the time interval between base year (2016) and predicted year (2026) to be analogous

with the time interval between the initial year (2006) and base year (2016).

3.7 Data presentation

The result from both field work and image analysis, were presented in the form of figures, (i.e.

maps and graphs) and tables. Graphs were prepared using Microsoft excel 2007 and Microsoft

word 2007.

37

Figure 2: Flow chart that shows the general methodology of this research. Adopted from Sang (2010)

and Shiferaw (2011) with some modification

38

4 RESULTS AND DISCUSSION

4.1 Characteristics of LU/LC units

Five, major LU/LC types were identified by using the field data and satellite images of

Landsat TM, 1986. These were forest, woodland, scrub/bush land, grassland/rangeland and

agriculture/settlement (Table 3). Water was added as a new LU/LC type in the Landsat images

of TM 1996, ETM+ 2006 and OIL/TIROS 2016 owing to the presence of dam in these

images. Rivers, streams and springs were not included in the classification. This is due to

resolution problem of the image (30m), the very low likelihood of identifying springs and

rivers from riverine vegetation. In the classification forest was made to include human made

plantation forest, riverine forests, dry ever green forest and moist mountain forest. This is

because as they had the same spectral nature on the images, it was difficult to differentiate one

from the other. This classification is in line with Tesfaye et al. (2014) who grouped both

natural forest and plantation forests under forest category. On the other hand it was difficult to

identify settlements especially rural settlements from agricultural land on 30m spatial

resolution image and in most cases the two are spatially integrated. Therefore settlements were

grouped under agricultural land covers. This classification is in agreement with Desalegn et al.

(2014) who grouped agriculture and settlement under one class for the same reason listed

above.

39

Table 3: Description of major LU/LC types identified in Bale Eco-Region

LU/LC types Their description

Forest Areas that are covered with dense growth of trees with closed

canopies. It was made to include human made plantation forest,

riverine forests, dry ever green forest and moist mountain forest.

Woodland The land covered with both open and closed (high) woodland with

dominant species of Acacia-Commiphora vegetation. It also

includes the scattered rural settlements found within the Woodland

(Molla et al., 2010)

Scrub/Bush land Land area covered by Asta scrubland, Erica bushes, alpine

vegetation (vegetation with small white leaves found at top of

Sanette Platue and habitats of Ethiopian Wolf. It includes Lobelia

rhynchopetalum and Helichrysum species) and ground covered by

Artemesia afra, Alchemilla johnstoni and Knifofia

Grassland/Range land Both communal and\or private grazing lands that are used for

livestock grazing. The land is basically covered by small grasses,

grass like plants and herbaceous species. It also includes land

covered with mixture of small grasses, grass like plants and shrubs

less than 2m and it is used for grazing

Agriculture/settlement Made to include areas allotted to rain fed cereal crops (e.g. Corn,

Barley, Teff, and Wheat), cash crops (chat) and horticultural crops

particularly vegetables (e.g. onion, potato and cabbage). Crop

cultivation both annuals and perennials, mostly in subsistence

farming and the land covered by rural villages and scattered rural

settlements

Water Land area covered by dam and small ponds.

40

4.2 Land use/Land Covers of the Study Area in 1986, 1996, 2006 and 2016

Starting from 1986 to 2006 woodland and forest were the dominant LU/LC types in the BER.

However by 2016 these LU/LC types were overtaken by agriculture/settlement.

Agriculture/settlement also predicted to dominate the land cover of BER by 2026 (Figure 3).

Figure 3: Area of LU/LC units at different periods in Bale Eco-Region

LU/LC analysis from the Landsat imagery of TM and ETM+ showed that starting from mid-

1980s to mid-2000s woodland and forest were the dominant LU/LC types in the study area.

These two LU/LC types together accounted for 1029134 ha (65%), 878905 ha (56%) and

827275 ha (53%) of the total area of BER in the years 1986, 1996 and 2006 respectively

(Figure 3). However LU/LC analysis from the Landsat OLI/TIRS imagery of 2016 indicated

that the area coverage of these two LU/LC types was overtaken by agriculture/settlement.

Agriculture/settlement was covered about 444345 ha (28%) of the study area (Figure 3). In

contrast during the same period woodland and forest covered only 268455 ha (17%) and

378803 ha (24%) of the study area respectively. On the other hand during the entire study

periods starting from 1986 to 2016 the smallest portion of the land in the study area was

covered by scrub/bush land (Figure 3). Scrub/bush land accounted for 110523 ha (7%),

41

140366 ha (9%), 166845 ha (11%) and 163247 ha (10%) of the total area of BER in the years

1986, 1996, 2006 and 2016 respectively

Figure 4: Map of LU/LC types of Bale Eco-Region produced based on unprocessed satellite images

obtained from USGS

As it is seen on the maps in 1986, 1996 and 2006 the greatest share of the land was covered by

woodland and forest. These LU/LC categories occupied the southern part of the study area

along the middle and lower parts of the eco-region. However, in 2016 and 2026 the lion share

of the land was covered by agriculture/settlement. In 1986 most of the agriculture/settlement

land cover was found on the northern parts of the study area along Adaba, Dinsho and Goba

42

Weredas. However, starting from 1996 to 2016 this LU/LC type was greatly expanded to the

southern part of the Harena forest and throughout the low land areas. It also expected to

expand for the year 2026 as well. In all maps the majority of scrub/bush land were occupied

the northern part of the study area above 3800 meters above sea level at Sanette Platue. In

contrast the majority of grassland/rangeland was located at southern part of the Harena forest

and throughout the lowland areas. There is a great expansion of agriculture/settlement and

grassland/rangelands along the southern part of the study area throughout the whole years.

4.3 Land Use/Land Cover Change Detection

4.3.1 Trend of LU/LCC in Bale Eco-Region

Bale Eco-region experienced different LU/LCC between 1986 and 2016. The land under

woodland decreased continuously between the indicated years. In contrast the area of forest,

scrub/bush land and grassland/rangeland showed a fluctuating trend between the study periods

(Figure 5).

43

Figure 5: Trend of LU/LCC in Bale Eco-region

Woodland showed the largest decline with a rate of decline of about 9890 ha/year. This was

followed by forest which was losing an estimated 2839 ha/year. Agriculture/settlement

showed the highest increase inclining by an estimated 5779 ha/year in the period from 1986 to

2016 (Table 4).

44

Table 4: Rate and percentage change of LU/LCs in Bale Eco-Region

LU/LC

category

1986-

1996

1996-

2006

2006-

2016

1986-

2016

2016-

2026

Rate

(ha/year)

%

change

(ha/year)

%

change

(ha/year)

%

change

(ha/year)

%

change

(ha/year)

%

change

Water -5 -2 -128 -53 0 0

Agriculture/ 7100 26 3350 10 6887 18 5779 64 12270 28

settlement

Woodland -9711 -17 -7680 -16 -12277 -31 -9890 -53 -6344 -24

Forest -5312 -11 2517 6 -5724 -13 -2839 -18 -6339 -17

Scrub/bush 2984 27 2648 19 -360 -2 1757 48 -589 -4

land

Grassland/range

land

4693 28 -830 -4 11603 57 5155 93 1001 3

In the period between 1986 and 1996 the land under woodland and forest decreased by 97114

ha (17%) and 53115 ha (11%) respectively, while agriculture/settlement increased by 71000

ha (26 %) (Figure 5 and Table 4). This implies that woodland and forest were declining at the

rate of 9711 ha and 5312 ha per annual respectively, while land under agriculture/settlement

increased at the rate of 7100 ha per year over the ten year period (Table 4). As reported from

discussion and interview with focus groups and key informants the rise of

agriculture/settlement between 1986 and 1996 linked with village establishment during the

Derge regime which was made effective around 1987, the 1985/86 resettlement program from

Harrerge and influx of illegal migrants. The efforts to improve agricultural systems by the

Derge government also played a great role for the expansion of agriculture. According to

FGD participants the massive reduction of vegetation in between 1986 and 1996 was during

the transitional period (i.e. 1990/1991). It is because during this transitional period the new

government in power was not capable to manage the country and no one was in charge of

45

protecting the natural resources of the country. Following the end of the war local peoples

participating in the war were returned to their environment and subsequently cleared the forest

to fulfill their livelihood requirements. Beside the local communities peoples coming from

other regions were also participated in the deforestation. Therefore lack of administration

coupled with lack of awareness among the local communities about the consequences of forest

conversion was the reasons behind this historic deforestation.

The result for the second period (1996-2006) indicated that the land under forest cover

increased by 25174 ha (6%) as compared to the first period (1986 - 1996) (Figure 5 and Table

4). On the other hand woodland and agriculture/settlement continued to decrease and increase

in the second period respectively. Woodland decreased by 76804 ha (16%), while

agriculture/settlement expanded by 33500 ha (10%). Increase in forest resource during this

period linked with different factors such as the integrated and participatory forest management

project which was implemented in the BER between 1995 and 2006 by GIZ and the Oromia

Bureau of Agriculture and Rural Development, extensive plantations carried out by the project

in Adaba Wereda and plantations by smallholder farmers in Goba Wereda. In line to the

finding of this study Tesfaye et al. (2014) reported increment in forest cover between 1986

and 2008 in Gilgel Tekeze catchment, Northern Ethiopia. The researcher has claimed that

increment in forest cover was due to tree plantation campaign and construction of terraces in

the hill slopes. Desalegn et al. (2014) also reported the rise in forest cover between 1975 and

1986 due to implementation of huge afforestation campaign by the Derge government in the

central highlands of Ethiopia.

46

The third period (2006-2016) result shows that grassland/rangeland increased greatly (i.e.

11603 ha/year (57%)) in the southern part of the Harena forest and throughout the low land

areas, with a corresponding decrease in woodland (12277 ha/year (31%)) (Table 4).

Grassland/rangeland increased at the expense of other LU/LC categories mainly woodland and

agriculture/settlement (Table 9). Shifting cultivation practices contributed for conversion of

woodlands to rangelands in the lowland parts of the eco-region creating huge openings in

woodland after cultivation is abandoned. On the other hand fallowing is one way of farmland

management practices in the study area, implying that the fallow farmlands used as a grazing

land for cattle. In some cases cultivated lands also permanently left for grazing. In agreement

to the finding of this study Alemayehu (2015) also reported expansion of grassland at the

expance of agricultural land in Fagita Lekoma Woreda, Awi Zone, North Western Ethiopia

between 1973 and 2015. Shiferaw (2011) also reported expansion of grassland at the expense

of forest and shrub land in Borena Woreda of South Wollo Highlands, between 1985 and

2003. Reduction of scrub/bush lands between 2006 and 2016 (Table 4) linked with deliberate

fire around scrub lands in order to expand grazing land. According to discussants of Goba

Wereda conversion of grazing lands in to agriculture and Eucalyptus tree plantation forced

inhabitants to send their livestock inside the afro-alpine scrub land to graze.

During the 30 year period between 1986 and 2016 the proportion of area covered by woodland

was continually decreasing as it was 565147 ha (36%) in 1986 and 268455 ha (17%) in 2016

(Figure 3). In contrast agriculture/settlement was continuously increasing as it was 270976 ha

(17%) in 1986 and 444345 ha (28%) in 2016 (Figure 3).

47

Generally the major finding from the analysis of Landsat images revealed a great reduction in

the area of woodland and forest and a corresponding increase in the area of

agriculture/settlement over the 30 year period. Focus group discussions and interviews

conducted in BER also support this trend showing increase in land under

agriculture/settlement over time with a corresponding reduction in land under forest and

woodland cover. In 1986 woodland and forest were the dominant LU/LC types in the study

area. This is because, during this time the area were characterized by relatively low population

pressure, small agricultural activities and to some extent undisturbed environmental condition.

However the largest part of lands that were covered by woodland and forest before 30 years

now replaced by agriculture and settlement. In agreement to the finding of this research,

studies conducted in dry and semi-dry land parts of Ethiopia such as Garedew (2010), Tefera

(2011), Alemu et al. (2015) documented reduction of area under woodland and increase in

area under agricultural land. Rapid reduction in woodland and forest and increase in

agriculture and settlement were also reported by Zeleke and Hurni (2001) in Dembecha area

of Gojjam, Molla et al. (2010) in the mountain landscape of Tera Gedam and adjacent agro-

ecosystem, Northwest Ethiopia and Kindu et al. (2013) in Munessa Shashemene landscape of

the Ethiopian highlands. However it is contrary to the work of Alemayehu (2015) who

reported expansion of forest land between 1973 and 2015 with corresponding reduction of

cultivated land in Fagita Lekoma Woreda, Awi Zone, Northwestern Ethiopia.

The information obtained from FGD participants and key informants, confirmed that the major

reasons for the continual expansion of agriculture/settlement between 1986 and 2016 in the

eco-region are rapid population growth, gradual change in the economic activities of

48

communities in the area from pure pastoralist to agro-pastoralist, loss of soil fertility,

vilagization and resettlement policies and poverty and food insecurity.

4.3.2 Land Use/Land Cover Change Matrix

Conversion matrixes were analyzed for each period to clearly show the source and destination

of the major LU/LCCs. Analysis of conversion matrix were computed by overlaying classified

images of two study year on ArcGIS 9.3. Results of the analysis are presented under Appendix

I. In all change matrixes the row of the table stand for the initial year and the column of the

table symbolize the final year of the change. More over except Table 11 all change matrixes

shows gross gain and loss of each land cover category during the study periods. The diagonal

numbers in bold show the unchanged pixels.

During the study period between 1986 and 2016 about 837446 ha (53%) of the study area

landscape remained unchanged. This implies around 47% of the total landscape of the study

area was converted from one LU/LC type to the other (Appendix I, Table 11). The level of

conversion varies amidst the LU/LC types. The woodland in the landscape was mainly

converted to grassland/rangeland and agriculture/settlement (Appendix I, Table 11). Forest

land was mainly converted to agriculture/settlement and woodland. Agriculture/settlement

replaced about 173117 ha of the land that used to be covered by other LU/LC types. The

major conversion were from woodland (150050 ha), forest (75989 ha) and grassland/range

land (37453 ha) (Appendix I, Table 11). Of all LU/LC types woodland experienced the lowest

persistence, whereas forest land was the most persistent cover type (Appendix I, Table 11).

Out of 567916 ha of woodland in 1986 about 390263 ha (69 %) were converted to other

LU/LC in 2016, while it is only 121,035 ha (26%) of forest land converted to other LU/LC

49

types between the indicated period. The net persistence for agriculture/settlement, woodland

and grassland/rangeland was large (relatively far from zero in both direction), whereas it is

closer to zero for the remaining LU/LC types (Appendix I, Table 11). The net persistence

closer to zero indicates the higher tendency of LU/LC types to persist rather than decline or

increase.

4.4 LU/LCC under different institutional set-up in Bale Eco-Region

Analysis of LU/LCC under different institutional set ups (Federal government /BMNP,

Oromia regional government land administration, PFM and PRM) showed considerable

difference in LU/LCC between 2006 and 2016. The most important change was the expansion

of agriculture/settlement and reduction of woodland and forest in all institutions, but with

differing rates (Table 5).

50

Table 5: LU/LCC under different institutional arrangements in BER, Note: NP = National park,

W.admin = Wereda administration, PFM = participatory forest management and PRM = participatory

rangeland management.

As it is seen in Table 5 agriculture/settlement in the lowland Kebele under Wereda institution

was expanding more widely than in the rest of institutions. Under this institution the rate of

expansion for agriculture/settlement in the last ten years was 1285 ha/year. This is an

indication of woodland and rangeland loss. Results from survey interviews showed that

expansion of both commercial and small scale farming activities and extraction of wood trees

for charcoal and fuel wood are the major causes of woodland and rangeland fragmentation.

This result well agree with the information obtained from the Delo Mena Wereda investment

51

office i.e. between 2011 and 2015/16 in Berak, Haya oda, Kale Golba and Naniga Dera

Kebeles about 6163 ha of rangeland and woodlands were given for agricultural investment. Of

this about 4952 ha (80%) is found in Berak Kebele. The result of this research under Table 11

is also consistence with this finding which documented the conversion of woodland and range

lands to agriculture/settlement.

Rira Kebele which is in highland agro-ecology and under national park is the second

institution where the rate of agriculture/settlement expansion was high followed by Kebeles

under PFM in midland agro-ecology. In the Kebele under national park agriculture/settlement

was expanding at the rate of 290 ha/year and it was at the expance of forestland. According to

the result of the LU/LCC analysis the rate of agriculture/settlement expansion is much smaller

in Tosha Kebele which is in the highland agro-ecology and under Wereda institution with 27

ha/year.

Comparing the loss of woodland and forest across different institutions, woodland in the

lowland Kebele under Wereda land administration showed the highest rate of reduction (564

ha/year) followed by forest in Kebele under federal institution (the national park) (394

ha/year). In contrast woodland in lowland Kebele under PRM and forest in highland and

midland Kebeles under PFM demonstrated the lowest rate of reduction. However, grassland in

Kebeles under PFM in highland and mid lands showed the highest rate of reduction (133

ha/year) over the last ten years as compared to other institutions.

From this one can conclude that between 2006 and 2016 much of the expansion in

agriculture/settlement was undertaken in the lowland and highland Weredas of BER especially

on the land under the Wereda administration and national park respectively. This expansion of

52

agriculture/settlement is at the expance of woodland and forest. This implies lack of cross-

institutional collaboration among natural resource management institutional arrangements

working in BER

53

Figure 6: Map that shows LU/LCC across different intuitions in BER

4.5 Predicting LU/LCC Based on the Markov Model

The predicted LU/LC type of 2026 is dominated by agriculture/settlement, which covers an

area of 567044 ha (36 %) of the total area. Forest and grassland/rangeland will cover an area

of 315414 ha (20%) and 331110 ha (21%) respectively, whereas the area coverage of

woodland, scrub/bush land and water will be 205019 ha (13%), 157361 ha (10%) and 1120 ha

(0.07%) in that order. This shows that in 2026 more than half (54%) of the study area is

expected to be covered by forest, grassland/rangeland and woodland. This is lowered by

about 22 % as compared to the initial year (1986) coverage.

As it is stated by Araya (2009) trend of the LU/LCC in the future time period can be detected

when predicted LU/LC at time t2 compared with LU/LC of the base year at time t with

54

reference to the class area metrics. Therefore as compared to the base year 2016 in 2026

agriculture/settlement is predicted to increase by 28%, while woodland and forest are

predicted to decrease by 24% and 17% respectively (Table 4).

The growth of agriculture/settlement will come largely at the expense of scrub/bush land,

forest and woodland respectively. This is because as it is seen in the probability matrix

(Appendix I Table 10) the probability of these LU/LC categories to change to

agriculture/settlement is high i.e. 46, 40 and 33 percent in that order. As it is indicated in the

probability matrix (Appendix I Table 10) in 2026 we expect 12% of woodland and 24% of

forest to persist and 88% of woodland and 76% of forest to change to other LU/LC. However

these two LU/LC categories are expected to gain 54% and 74% from other LU/LC categories.

This implies that woodland and forest will have a net loss of 34% and 2% respectively.

4.6 Causes of LU/LCCs in Bale Eco-Region

LU/LCC in the BER is a result of several proximate and underlying causes.

4.6.1 Proximate (Direct) Causes

The series of discussions and interviews conducted with the FGD participants and key

informants in the study area indicated that five major proximate (direct) driving forces appear

to explain a large part of LU/LCC in the BER. These are; (i.) expansion of agriculture (ii.) fire

(iii.) illegal logging and fuel wood extraction (iv.) overgrazing and (v.) expansion of illegal

and unplanned settlements. Urban expansion and construction of infrastructures such as school

and road also take part in changing the LU/LC of the BER. However they have a minor role

on the LU/LCC of the eco-region.

55

As reported from discussion and interview with focus groups and key informants expansion of

agriculture including crop farming (both subsistence and commercial farming), forest-coffee

farming (by small-holder farmer) and other cash crop like chat farming are the major drivers

of LU/LCC in BER. Agriculture is expanding in all parts of the eco-region at the expanse of

grassland/rangeland, forest and woodlands. Subsistence crop farming is the major driving

forces for forest, grassland and scrub/bush land cover change in highland Weredas like Goba

and Dinsho. In Dinsho Wereda a number of hills which were previously covered by small

grass, bush and forest have been converted to small scale crop farms. Moreover, as per the

Goba Wereda investment office in 2015 alone about 100 ha of grassland was given to

agricultural investment in Ashuta Kebele. Chat and forest coffee farming are the major

drivers of forest and grassland cover change in Harena Buluk Wereda and in the high land

Kebeles of Delo Mena Wereda. Rangelands and woodlands in lowland Weredas like Delo

Mena, Mede Welabu and Berbere were converted to subsistence and commercial crop farms.

This is in agreement with Teshome (2010) who reported that between 1986 and 2006 about

65% and 10% of forest in Goba and Delo Mena respectively were converted to agriculture.

Participants in the FGD also identified fire as one of the major proximate causes of LU/LCC

next to agricultural expansion in the study area. They perceived two major fire incidents over

30 year period that caused destruction of woodland, forest and scrub/bush lands. These were

the 2000 and 2008 fires. The discussants stated that most of the forest and woodland areas

destroyed by these major fires are now converted to agriculture and settlement areas. There are

studies that quantified the destruction caused by these two major fire incidents and by the

1984 fire. Accordingly the fire occurred in 1984 destroyed about 195 km2 (Belayneh et al.,

2013) of vegetated lands in different parts of the eco-region, while in year 2000 fire destroyed

56

approximately 20,000 ha of moist evergreen forest (Wakjira, 2015). On the other hand in the

year 2008 fire destroyed 12, 825 ha of vegetated land in the eco-region (Belayneh et al.,

2013). Out of the 12, 825 ha destroyed by 2008 fire about 11,972 ha (93%) was destroyed

within the intervention Weredas. Abera and Kinahan (2011) also reported about 142 fire

incidents that occurred within the national park boundary between 1999 and 2008. According

to researcher the fires occurred within this range of period caused destruction of 38, 150 ha of

woodland, forest and Erica shrub. As it was stated by key informants and focus group

discussants the source of nearly all fire occurred in the eco-region were human during illegal

hunting, honey harvesting (in which smoke is used to protect the beekeepers) and farm land

clearing. Farmers also burn the afro-alpine scrub lands to initiate fresh grass for their cattle.

Illegal logging and fuel wood extraction in the form of charcoal and fire wood is also a major

driver for the diversion of forest and wood lands in BER. Especially in Goba, Delo Mena and

Berbere Weredas forests and woodlands are highly exploited for the purpose of charcoal

preparation and fire wood. According to key informants forests and woodlands provides the

major towns (i.e. Robe, Goba town, Delo Mena town and Haro Dumal town) with charcoal

and firewood as the energy source of majority of population living in these towns depend on

fuel wood. On the other hand discussants of this study also confirmed that rural population

especially those vulnerable to climatic change and those economically poor use charcoal and

firewood as a source of income to cope with the effects of climatic hardships and fulfill the

livelihood requirements of their family. In addition to this most of the migrants coming from

different parts of the country are highly involved in charcoal production and fire wood sale

besides to converting forest and woodlands in to farmland and settlement.

57

Overgrazing is another major proximate cause of LU/LCC identified by FGD participants and

key informants. They stated that increase in the number of livestock from time to time and

conversion of grasslands and rangelands to agriculture created livestock pressure on currently

existing grasslands and rangelands. This further forced the local communities to send their

livestock inside the forest and woodland resources for grazing, thereby exerting severe

pressure on the forest and woodland through browsing and trampling. This problem is most

common in woodland and forested Weredas of BER.

Results from survey interview and group discussion also indicated that expansion of Illegal

and unplanned settlements inside the dense forests and woodland are the other major

proximate driver of forest and woodland cover fragmentation in BER. They stated that such

settlement is the major problem in Harena forest. This is because a number of illegal migrants

from different parts of the country settled inside the Harena forest and cleared the forest for

the purpose of settlement and coffee farm. According to the discussants most of illegal settlers

of Harena forest came from South region especially from Sidama Zone, Amhara region

(Godjam and Gondar) and Oromia region from Shewa (north and west shewa), East and West

Harrage, and from the arid Arsi Zone. On the other hand illegal and unplanned settlements by

the local peoples to expand crop land and settlement also contributed for fragmentation of

forest and woodlands, especially for woodlands of lowland Weredas like Delo Mena.

Expansion of urban areas like Delo Mena, Angetu, Haro Dumal and Bidire towns and

construction of infrastructures like school and roads also contributed for the conversion of

grassland, forest and woodlands. As it is stated by discussants of this study the current schools

in their locality replaced the land that was covered by grassland and woodland before two to

58

three decades. On the other hand dense forests and woodland areas were opened through road

construction.

4.6.2 Underlying Causes

The above mentioned proximate causes were triggered by different underlying causes of

LU/LCC. From a range of demographic, economic, technological, institution and policy,

socio-cultural and biophysical factors more than 20 underlying drivers of LU/LCC were

identified by the FGD participants and key informants in the study area (Table 6). However

population growth, poverty and food insecurity, gradual change in the economic activities of

communities in the area from pure pastoralist to agro-pastoralist, weak law enforcement and

drought appear to explain a large part of underlying causes.

59

Table 6: Underlying Causes of LU/LCC in BER

Drivers Their Category

Population growth Demographic

Poverty and food insecurity, unemployment,

change in rural economic activity and

opportunity to drive high economic benefits

Economic

Improvement in road networks, access to

markets, change in farming technologies

and access to agricultural inputs such as

inorganic fertilizers, improved seed and

herbicides

Technological

Villagezation and resettlement policies,

national and regional policies on land use

and economic development, lack of proper

land use plans, weak law enforcement, low

investments in management and protection

of natural resources and change in land

tenure

Institution and policy

Resource use interdependency, resource use

competition, lack of awareness, distribution

of land and other resource between

generations and traditional land use systems

Socio-cultural

Drought Biophysical (natural)

Demographic Factors

Focus group discussants and key informants of this study confirmed that population growth is

the first of all underlying drivers of LU/LCC identified in the BER. In line to this, studies like

Oumer (2009), Tefera (2011) and Alemu et al. (2015) also identified population pressure as

one of the major underlying drivers of LU/LCC in different parts of Ethiopia. Key informants

60

and FGD participant perceived that population of BER has been increasing from time to time.

They identified three major reasons for BER’s population growth namely; (i) resettlement

programs carried out around 1985/86, 1999 and 2003 from Harrerge Zone (ii) influx of

migrants from different parts of the country and (iii) natural increase i.e. high fertility rates

owing to early marriages and polygamy. According to the knowledge of FGD participant

population growth in the area increased the demand for agriculture, settlement and fuel wood

and construction materials. This in turn resulted in forest and woodland encroachment for

settlement, new agricultural land and fuel wood extraction. The evidence obtained from local

informants confirmed that the current resettlement sites and areas where illegal settler found

were covered by forests, grassland and woody plants in the previous three decades. The

perception of discussants and key informants is in agreement with CSA reports that shows rise

in total population of the study area. In 1994 the total population in the study area was 475,515

(CSA, 1996). In 2016 it increased to 878,493 with population density of 52 person/ km2 (CSA,

2013). This implies that between these two years the number of population in study area

increased by about 402,978 with annual rate of about 18318 persons / year. As shown in the

Figure 7 between 1994 and 2016 total population of the study area was increased by 85%.

61

Figure 7: Population growth in seven Weredas of BER (1994-2016) drived from CSA (Central

Statistical Agency). Note: Due to the data gap from the CSA total population for the years between

1994 and 2004 and 2009 were not available for use.

Economic Factors

The economic causes of LU/LCC in BER includes: poverty and food insecurity,

unemployment owing to lack of off-farm jobs (especially for landless and educated youth),

change in rural economic activity and opportunity to derive high economic benefits from the

sale of cash crops such as coffee and chat. Key informants and focus group discussants stated

that these were the major underlying economic factors behind the expansion of agriculture

especially coffee and chat farming inside the forest and illegal logging and fuel wood

extraction in the form of charcoal and fire wood. Due to lack of off-farm employment

opportunities adults in BER remain in their area as unemployed. This resulted in land

62

fragmentation due to sharing of lands from their families and encroachment of forest,

grassland and woodland in search of new agricultural land and fuel wood. To show the role of

poverty on environmental change the World Commission on Environment and Development

pointed out that people who are poor and hungry always destroy their immediate environment

in order to survive (Belay, 1995). Accordingly in the study area those economically poor and

landless households were engaged in fuel wood extraction in the form of charcoal and

firewood to fulfill the livelihood requirement of their family. Change in the rural economy

from pure pastoralism to agro-pastoralism and establishment of economic linkage between

rural and urban centers also played a greatest role for the expansion of agriculture in the BER.

Economic linkage between rural and urban areas was facilitated by infrastructural

developments such as road and markets. Traditionally the local peoples in almost all parts of

the BER were pure pastoralist. According to FGD participant agriculture started in the area

during Derge regime around 1975/76 following the “land to tiller” policy of the Derg

government. However, they agreed that more expansion was observed in the past two –three

decades. One of the reasons for this expansion was influx of legal and illegal migrants

especially in Harena Buluk Wereda. Influx of new migrants to the study area introduced new

livelihoods centered on crop farming. Subsequently non migrant local communities adopted

migrant’s livelihood and changed their livelihood from pure pastoralist to agro –pastoralist.

Technological Factors

According to discussants and key informants the technological factors responsible for

LU/LCC in BER were improvement in road networks, access to markets and change in

farming technologies (from hand tool (locally known as Haarkoo) to animal power and to

63

tractors) and access to agricultural inputs such as inorganic fertilizers, improved seed and

herbicides. Distance from the market and low road facilities decreases the share of agricultural

land. On the other hand it increases the share of forest, woodland, shrub and grasslands

(Girma and Hassan, 2014). This implies that access to market and improvement of road

network promotes the conversion of forest, woodland and grassland in to agricultural land.

FGD participants stated that during the Derge regime there were very few markets, with

average travel distance of 4-5 hours to get market. On the other hand they used to travel by

using traditional mode of transportation (animal power). But currently there are many rural

and urban market places and peoples are shifted from tradition mode of transportation to

modern i.e. using motor cycles and car subsequently reduced their travel hours to 1hour-

30minuets. All the above mentioned factors created a big motivation among the local

communities to increase their agricultural products through increasing their land holding size

and this can be at the expanse of vegetated lands mainly forest, woodland and grassland.

Policy & Institutional Factors

As per the information obtained from key informant villagezation policy during the Derge

regime where by people were clustered in villages called “Sefera” and the resettlement policy

contributed to expansion of settlements and agriculture. The other most important policy

contributed for agricultural expansion in the study area during the Derge regime was “ Land to

Tiller” where by privatization of communal lands were carried out. National and regional

policies on land use and economic development such as infrastructural expansion (e.g. roads,

schools, markets etc.), attaining food self-sufficiency through investment on agriculture are

the other factors contributing to LU/LCC. These lead to expansion of small and large scale

64

agriculture, construction of several infrastructures at the expanse of forest, woodland and

grassland

Lack of proper land use plans is another policy related driver of forest and woodland cover

change. It is characterized by encroachment of vegetated lands especially forest, woodlands

and national park for settlement, pasture and agriculture, cultivation of steep slope and

opening of very dense forest areas through road construction.

Weak law enforcement is also significant driver of LU/LCC. As per the information obtained

from FGD and KII, manifestations of weak law enforcements in the study area were

corruption, lack of benefit sharing and delay in decision making by the courts. These all are

the major reasons for inability of local institution and community based organizations to

discourage illegal settlements in vegetated areas and to control fire, livestock grazing inside

the forest and expansion of both small holder and commercial agriculture including forest

coffee farming at the expense of forests, woodland and grasslands. Lack of benefit sharing

between local communities and government and non-government organizations was raised as

problem especially for those Kebeles found under the jurisdiction of Oromia Forest and

Wildlife Enterprise (OFWE) and Bale Mountain National Park (BMNP). The other policy and

institutional related drivers of LU/LCC was change in institutional power on land use from

shared and customary systems to privatization and formal institutional system.

Change in land tenure system was another policy related driver of forest and woodland cover

change. Participants of the FGD informed that during the Derge regime huge size of forest ,

grassland and woodlands in BER had been converted to other land use types due to change in

the land policy of the previous government. Distribution of land and resource among small

65

scale farmers following the 1975 land to tiller policy and launch of state farms resulted in

conversion of vegetated lands to agricultural land. However in FDRE government land is a

public property and administered by government (Tefera, 2011). Rural peoples have the right

to use land indefinitely and to lease/rent, and transfer the land; correspondingly the land policy

of the FDRE government looks better as compared to the Derge regime (Tefera, 2011).

However in the study area participants of the FGD agreed that farmers are still lacking

confidence and feel as they have no right over their land. This coupled with a very low land

holding per household motivated local farmers to encroach in to vegetated lands for cropping,

grazing and settlement.

Socio-cultural Factor

Resource use interdependency between lowland and highland communities and resource use

competition among non-migrant local communities and between migrant and non-migrant

communities are the major socio-cultural causes of LU/LCC identified by discussants and key

informants of this study. They stated that resource use competition between the migrant and

non-migrant local communities is the most sever and it was responsible for the conversion of

forest and woodlands to agriculture. This problem is common in the areas where there are high

migrants. Local peoples in these areas are not happy with migrants assuming that these people

are overriding future resources of their children. On the other hand during the dry season and

drought years, transhumance peoples migrate from the lowland Weredas and Kebeles to the

highland part of the eco-region in search of pasture to their cattle. These people use forests and

grasslands in the mid land as pasture and shade to their livestock. On the other hand highland

communities expand agriculture in lowland parts of the eco-region at the expance of woodland

66

and rangelands. This shows resource use interdependency between lowland and highland

communities.

Lack of awareness about the negative impacts of forest conversion, distribution of land and

other resource between generations and traditional land use systems brought by incoming

migrants such as Chat and other crop farming like maize and rice are also the other socio-

cultural causes for the expansion of agricultural lands at the expance of other LU/LC types

mainly woodland and forest.

Biophysical (Natural) Factors

The biophysical factor considered in this study was drought. It was mentioned by participants

of the FGD as one of the factors contributing for LU/LCC in the study area especially in low

land Weredas like Delo Mena. Focus group discussants in Delo Mena Wereda perceived two

drought years over 30 year period i.e. 1997 and 2011. They stated that during the drought

years people in their locality highly involve in extraction of fuel wood in the form of charcoal

and fire wood to fulfill the livelihood requirement of their family. On the other hand people

also migrate to midland and highland parts of the eco-region in search of pasture and water to

their livestock. This demonstrates the potential threat of such activities to woodlands and

rangelands in the lowland and forests and grasslands in high and mid altitude.

67

5 CONCLUSION AND RECOMMENDATION

5.1 Conclusion

Bale Eco-region has been experiencing different LU/LC changes. The main finding of this

study revealed a continual increase in agriculture/settlement at the expense of woodland and

forest between 1986 and 2016. During 30 years period agriculture/settlement increased by

173369 ha (64%), with a corresponding 296692 ha (52%) and 85184 ha (18%) decline in the

area of woodland and forest. If the current rate of LU/LCC continues agriculture/settlement is

predicted to increase by 122699 ha (28%) in 2026. On the other hand woodland and forest are

predicted to shrink by 63436 ha (24%) and 63389 ha (17%) respectively.

Findings of the LU/LCC analysis under different institutional arrangements between 2006 and

2016 showed expansion of agriculture/settlement and reduction of woodland and forests under

all institutions, but with differing rates. The rate of agriculture/settlement expansion was high

in lowland Kebele under Wereda land administration with (1285 ha/year) followed by Kebele

under the national park with (290 ha/year) and Kebele under PFM in mid lands with (245

ha/year). In contrast rate of agriculture/settlement expansion was much smaller in the highland

Kebele under the Wereda institution with 27 ha/year. On the other hand rate of woodland and

forest loss was high in parts of BER that are managed by Wereda and national park with 564

ha/year and 394 ha/year respectively. On the contrary the rate of deforestation is low in forests

and woodlands managed under PFM and PRM institutional arrangements.

LU/LCC in the BER is a result of different interactions between proximate and underlying

causes. The major proximate driving forces of LU/LCC in the BER are expansion of

agriculture, fire, illegal logging and fuel wood extraction, overgrazing and illegal and

68

unplanned settlement. On the other hand from a range of demographic, economic,

technological, policy and institution, socio-cultural and biophysical factors more than 20

underlying drivers of LU/LCC are identified by key informant and focus group discussants of

this study. However population growth, poverty and food insecurity, gradual change in the

economic activities of communities in the area from pure pastoralist to agro-pastoralist, weak

law enforcement and drought appear to explain a large part of underlying causes.

69

5.2 Recommendation

Expansion of agriculture especially small scale agriculture by small holder farmers is

the major proximate/direct causes of LU/LCC in BER causing loss of several hectares

of forest and woodlands. Therefore, controlling the expansion of agriculture at the

expense of forest and woodlands requires the right policy packages by national and

regional governments such as livelihood diversification and improving the productivity

of existing farm lands through the provision of improved production inputs.

Population growth is the major root cause for LU/LCC in the study area. Traditional

practices such as early marriages and polygamy and illegal migrations are the reasons

for population growth. Therefore, controlling the population growth and its associated

impacts on the natural environment requires the right policy packages by national and

regional governments such as awareness creation, provision of family planning

services, increasing productivity, working on the pushing factors of migration and

controlling illegal settlements. Working on the pushing factors of migration may

require a joint action between the senders and recipients of migrants. Currently there

are undergoing efforts in the study area but these efforts should be strengthened more.

Poverty and food insecurity were the other most important root cause for land use/land

cover change in the study area. Combating this problem therefore requires designing of

good polices and strategies. Thus both national and regional governments should

design policies and strategies like creating and strengthening environmental friendly

non-farm/off-farm income generating activities and provision of safety net programs

PFM project which is implemented by Farm Africa and OFWE have brought several

positive benefits especially on forest resources. However implementation problems

70

such as lack of benefit sharing were reported by local communities. Hence

performance assessment is required for all institutions working in the study area to

better understand existing problems and make immediate corrections.

Even though there are some positive benefits especially on forest resources under the

jurisdiction of PFM/PRM still there is resource degradation. The finding of this study

also revealed high loss of resources such as forest, woodland and range lands under the

jurisdiction of Wereda administrations and Bale Mountain National Park as compared

to areas under PFM/PRM. In this regard, two major recommendations are forwarded

by the researcher. First currently ongoing efforts under PFM/PRM should be

strengthened more. Secondly the successes achieved in PFM/PRM should be extended

to other institutional arrangements in BER.

This study addressed only the change in LU/LC and driving forces behind the change.

Therefore, further study is required to assess impacts brought by LU/LCC.

71

6 References

Abdullah, F., Alqurashi and Kumar, L. 2013. Investigating the Use of Remote Sensing and

GIS Techniques to Detect Land Use and Land Cover Change: A Review. Advances in

Remote Sensing (2): 193-204.

Abera, K. and Kinahan, A. A. 2011. Factors Affecting Fire Extent and Frequency in the Bale

Mountains National Park. Walia Journal of the Ethiopian Wildlife and Natural History

Society pp.146-157.

Adedeji, O. H., Tope-Ajayi, O. O. and Abegunde, O. L. 2015.Assessing and Predicting

Changes in the Status of Gambari Forest Reserve, Nigeria Using Remote Sensing and

GIS Techniques. Journal of Geographic Information System (7): 301-318.

Agarwal, C., Green, G. M., Grove, J. M., Evans, T. P. and Schweik, C. M. 2002. A Review

and Assessment of Land-Use Change Models: Dynamics of Space, Time, and Human

Choice. Gen. Tech. Rep. NE-297. Newtown Square, PA: U.S. Department of

Agriculture, Forest Service, Northeastern Research Station.

Alemayehu, B. 2015. GIS and Remote Sensing Based Land Use/Land Cover Change

Detection and Prediction in Fagita LekomaWoreda, Awi Zone, North Western

Ethiopia. MSc. thesis, Addis Ababa University, Ethiopia.

Alemu, B., Garedew, E., Eshetu, Z. and Kassa, H. 2015. Land Use and Land Cover Changes

and Associated Driving Forces in North Western Lowlands of Ethiopia. International

Research Journal of Agricultural Science and Soil Science 5(1): 28-44.

72

Amare, B. 2007. Landscape transformation and opportunities for sustainable land management

along the eastern escarpment of Wello (EEW). Doctoral thesis, University Bern.

Araya, Y. H., 2009. Urban Land Use Change Analysis and Modeling: a Case Study of Setúbal

and Sesimbra, Portugal. MSc. thesis, University of Münster.

Arsanjani, J. 2012. Dynamic Land Use/Cover Change Simulation: Geosimulation and Multi

Agent-Based Modelling. Springer-Verlag Berlin Heidelberg. Doctoral thesis, Ausrtia.

Bakker, W. H., Gorte, B. G. H., Janssen, L. L. E., Pohl, C., Prakash, A., Reeves, C. V., Weir,

M. J. C. and Woldai, T. 2001. Principle of Remote Sensing; ICT: The Netherlands, 180

P.

Bale Mountains National park general Management Plan (BMNP GMP), 2007. Bale

Mountains National Park General Management Plan. (Frankfurt Zoological Society,

eds.), Ethiopia.

Belay, T. 1995. Population pressure and problems of arable land degradation in Ethiopia.

Proceedings of the Workshop on Integration of Population, Environment Equitable and

Sustainable Development Issues into the Curriculum of the Demographic Training and

Research Centre of the Institute of Development Research at Addis Ababa University,

Wabe Shebelle Hotel, Addis Ababa.

Belay, T. 2002. Land cover/ use changes in the Derekolli catchment of the South Welo Zone

of Amhara Region, Ethiopia. Eastern Africa Social Science Research Review 18 (1): 1-

20.

73

Belayneh, A., Yohannes, T. and Worku, A. 2013. Recurrent and extensive forest fire incidence

in the Bale Mountains National Park (BMNP), Ethiopia : Extent, Cause and

Consequences 2(1): 29–39.

Bhawana, K. C. 2015. Land Use and Land Cover Change In relation to Internal Migration and

Human Settlement in the Middle Mountain of Nepal. MSc. thesis, University of

Twente, Nepal.

Briassoulis, H. 2011. Factors Influencing Land Use Land Cover Change. Journal of Land Use,

Land Cover and Soil Science1.

Brown, D. G., Pijanowski, B. C. and Duh, J. D. 2000. Modeling the Relationships Between

Land Use and Land Cover on Private Lands in the Upper Midwest, USA. Journal of

Environmental Management (59): 000–000, available online at

http://www.idealibrary.com.

Brown, D. G., Walker, R., Manson, S., Seto, K. 2004. Modeling Land-Use and Land-Cover

Change.

Burka, A. 2008. Land Use /Land Cover Dynamics in Prosopis juliflora invaded area of

Metehara and the Surrounding Districts Using Remote Sensing & GIS Techniques.

MSc. thesis, Addis Ababa University, Ethiopia.

Chang-Martínez, L. A., Mas, J. F., Valle, N. T., Torres, P. S. U. and Folan, W. J. 2015.

Modeling Historical Land Cover and Land Use: A Review from Contemporary

Modeling. ISPRS International Journal of Geo-Information (4): 1791-1812.

74

De Sherbinin, A. 2002. Land-Use and Land-Cover Change, A CIESIN Thematic Guide,

Palisades, NY. Center for International Earth Science Information Network of

Columbia University.

Desalegn, T., Cruz, F., Kindu, M., Turrión, M. B. and Gonzalo, J. 2014. Land-Use/Land-

Cover (LULC) Change and Socioeconomic Conditions of Local Community in the

Central Highlands of Ethiopia. International Journal of Sustainable Development &

World Ecology.

Duhamel, C. 2011. Land Use and Land Cover, Including their Classification. Journal of land

use, land cover and soil sciences I.

Eastman, J.R. 2006. IDRISI Andes tutorial. Clark Labs, Worcester, MA.

Ebro, A., Shenkute, B., Haile, G., Sauerborn, J. and Treydte, A. C. 2011. Land use / land

cover dynamics in the rift valley of Ethiopia: Implications on food security. Food

Security Center (FSC) brief No. 12.

Eshetu, Z. and Hogberg, P. 2000. Reconstruction of forest site history in Ethiopian highlands

based on C13 natural abundance of soils. Ambio (29): 83–89.

FARM Africa / SOS Sahel Ethiopia, 2007. The Key Steps in Establishing Participatory Forest

Management: A field manual to guide practitioners in Ethiopia, Addis Ababa.

FARM Africa, 2008. Bale Mountains Eco-Region Sustainable Development Plan: Report

Phase I and II planning Workshops, Addis Ababa: FARM Africa.

75

FDRE-CSA, 1996. The 1994 Population and Housing Census of Ethiopia, Result for Oromia

Region Volume I: Part I Statistical report on Population Size and Characteristics,

Addis Ababa, Ethiopia.

FDRE-CSA, 2013. Population Projection of Ethiopia for all Regions at Wereda Level from

2014 – 2017.

Fentahun, T. and Gashaw, T. 2014. Evaluation of land use/land cover changes of Bantneka

Watershed, Ethiopia. Merit Research Journal of Agricultural Science and Soil

Sciences 2(7): 081-085.

Garedew, E. 2010. Land-Use and Land-Cover Dynamics and Rural Livelihood Perspectives,

in the Semi-Arid Areas of Central Rift Valley of Ethiopia. Doctoral thesis, Swedish

University of Agricultural Sciences, Umeå.

Gashaw, T. and Dinkayoh, T. 2015. Land use/land Cover Dynamics in Hulet Wogedamea

Kebele, Northern Ethiopia. Current Research in Agricultural Sciences 2(1): 36-41.

Gebrehiwet, K. B. 2004. Land Use and Land Cover Changes in the Central Highlands of

Ethiopia: The Case of Yerer Mountain and its Surroundings. MSc. thesis, Addis Ababa

University, Ethiopia.

Gebrehiwot, S., Taye, A. and Bishop, K. 2010. Forest cover and stream flow in a headwater of

the Blue Nile: Complementing observational data analysis with community perception.

Ambio (39): 284–294.

76

Gebreslassie, H. 2014. Land Use-Land Cover dynamics of Huluka watershed, Central Rift

Valley, Ethiopia. International Soil and Water Conservation Research 2(4): 25-33.

Geist, H. J. and Lambin, E. F. 2001. What drives tropical deforestation? LUCC Report Series

No.4. Louvain-la-Neuve: CIACO.

Geist, H. J. and Lambin, E. F. 2002. Proximate Causes and Underlying Driving Forces of

Tropical Deforestation. BioScience 52(2): 143-150.

Geremew, A. A. 2013. Assessing the Impacts of Land Use and Land Cover Change on

Hydrology of Watershed: A Case study on Gilgel–Abbay Watershed, Lake Tana

Basin, Ethiopia. MSc. thesis, Universitat Jaume-I.

Girma, H. M. 2014. Drivers of Land Use Change and Forest Conservation Under Uncertain

Markets for Forest Ecosystem Services in Ethiopia. Doctoral thesis, University of

Pretoria, Pretoria.

Girma, H. M. and Hassan, R. M. 2014. Drivers of land-use change in the Southern Nations,

Nationalities and People’s Region of Ethiopia. African Journal of Agriculture and

Resource Economics 9 (2): 148-164.

Hailemariam, S. N., Soromessa, T. and Teketay, D. 2015. Non-carbon benefits for effective

implementation of REDD+: The case of Bale Mountains Eco-Region, Southeastern

Ethiopia. African Journal of Environmental Science and Technology 9(10): 747-764.

77

Hamza, I. A. and Iyela, A. 2012. Land Use Pattern, Climate Change, and Its Implication for

Food Security in Ethiopia: a review. Ethiopian Journal of Environmental Studies and

Management 5(1): 26-31.

Han, H., Yang, C. and Song, J. 2015. Scenario Simulation and the Prediction of Land Use and

Land Cover Change in Beijing, China. Sustainability (7): 4260-4279 .

Hillman, J. C. 1986. Bale Mountains National Park, Management Plan. Ethiopian Wildlife

Conservation Organisation, AddisAbaba, Ethiopia.

Houghton, R. A. 1994. The world wide of extent of land-use change. Bioscience (44): 305-

313.

Iacono, M., Levinson, D., El-Geneidy, A. and Wasfi, R. 2015. A Markov chain model of land

use change in the Twin Cities, 1958-2005. Tema. Journal of Land Use, Mobility and

Environment 8 (3): 49-62.

Jayasekara, R. S. 2012. Focus groups in nursing research: Methodological perspective.

Nursing Outlook 60 (6): 411-6.

Kindu, M., Schneider, T., Teketay, D. and Knoke, T. 2013. Land Use/Land Cover Change

Analysis Using Object-Based Classification Approach in Munessa-Shashemene

Landscape of the Ethiopian Highlands. Journal of Remote Sensing (5): 2411-2435.

Lambin, E. F. and Geist, H. J. 2003. Global land-use and land-cover change: what have we

learned so far? Global Change News Letter (46): 27- 30.

78

Lambin, E. F. and Geist, H. J. 2007. Causes of land-use and land-cover change. Retrieved

from http://www.eoearth.org/view/article/51cbed2f7896bb431f6905af.

Lambin, E. F. and Geist, H. J. (eds.), 2006. Land-Use and Land- Cover Change, Local Process

and Global impacts. pp.71-116.

Lambin, E. F., Geist, H. J. and Lepers, E. 2003. Dynamics of land-use and land-cover change

in Tropical regions. Annual Review of Environment and Resources 28(1): 205-241.

Lamichhane, B. B. 2008. Dynamics and Driving Forces of Land Use/Forest Cover Change and

Indication of Climate Change in a Mountain Sub-watershed of Gorkha. MSc. thesis,

Tribhuvan University, Institute of Forestery, Nepal.

Mani, N. D. and Krishnan, N. R. 2013. Assessment of Changes in Land Use/ Land Cover in

Tamil Nadu State in India Using GIS. African Journal of Science and Research 2(6):

01-06.

Muchen, M. 2012. Land Use Land Cover Change in the Ketar Catchment, Ziway Watershed,

Ethiopia. Georg-August-Universitat, Gottingen.

Melaku, B. 2003. Forest property rights, the role of the state and institutional exigency: The

Ethiopian experience. Doctoral thesis, Swedish University of Agricultural Sciences,

Uppsala.

Meyer, W. B. and Turner, B. L. II, 1992. Human Population Growth and Global Land-

Use/Cover Change. Annual Review of Ecology and Systematics (23): 39-57.

79

Meyer, W. B. 1995. Past and Present Land-Use and Land –Cover in the U.S.A. Consequences

PP. 24-33.

Minale, A. S. 2012. Population and Environment Interaction: the case of Gilgel Abbay

Catchment, Northwestern Ethiopia. Journal of Environmental Research and

Management 4(1): 0153-0162.

Ministry of Mines and Energy, 2003. Geological Survey of Ethiopia: Industrial Minerals and

Rocks Resource Potentials of Ethiopia. MoME: Addis Ababa.

Mirkatouli, J., Hosseini, A. and Neshat, A. 2015. Analysis of land use and land cover spatial

pattern based on Markov chains modelling. Open Journal of City, Territory and

Architecture 2(4): 1-9.

Molla, E., Gebrekidan, H., Mamo, T. and Assen, M. 2010. Patterns of Land Use/Cover

Dynamics in the Mountain Landscape of Tara Gedam and Adjacent Agro-Ecosystem,

Northwest Ethiopia. SINET: Ethiop. J. Sci. 33(2): 75–88.

Molla, M. B. 2014. Land Use/ land cover dynamics in the Central Rift Valley Region of

Ethiopia: The Case of Arsi Negele District. Academia Journal of Environmental

Sciences 2(5): 074-088.

Morie, N. D. 2007. Land Use and Land Cover Changes in Harena Forest and Surrounding

Area, Bale Mountains National Park, Oromia National Regional State, Ethiopia. MSc.

thesis, Addis Ababa University, Ethiopia.

80

Muluneh, A. 2010. Synthesis of Research on Land Use and Land Cover Dynamics in the

Ethiopian Highlands. UNU-Land Restoration Training Programme, Keldnaholt, 112

Reykjavik, Iceland.

Muluneh, W. 2003. Impact of Population Pressure on Land Use and Land Cover Changes,

Agricultural System and Income Diversification in West Gurague Land, Ethiopia.

Doctoral thesis, Norwegian university of science and technology (NTNU), Trondheim.

OFWE, FARM Africa, SOS Sahel Ethiopia, 2014. Bale Mountains Eco-Region Reduction of

Emission from Deforestation and Forest Degradation (REDD+) Project-Ethiopia.

Addis Ababa: FARM Africa and SOS Sahel.

Opeyemi, Z. A. 2006. Change Detection in Land Use and Land Cover Using Remote Sensing

Data and GIS: (A case study of Ilorin and its environs in Kwara State.).

Oumer, H. A. 2009. Land Use and Land Cover Change, Drivers and Its Impact: A

Comparative Study From Kuhar Michael and Lenche Dima of Blue Nile and Awash

Basins of Ethiopia. MSc. thesis, Graduate School of Cornell University.

Paul, W. B. and Lisa, A. S. 2011. Agricultural landscape change (1937–2002) in three

townships in Iowa, USA. Landscape and urban planning (100): 202–212.

Pellikka, P. 2008. Land Use and Land Cover Changes in Africa – from Local to Continental.

HERC seminar.

Rahdary, V., Soffianian, A., Najfabdai, S. M., Khajeddin, S. J. and Pahlavanravi, 2008. Land

Use and Land Cover Change Detection of Mouteh Wildlife Refuge Using Remotely

81

Sensed Data and Geographic Information System. World Applied Sciences Journal

3(1): 113-118.

Reis, S. 2008. Analyzing Land Use/Land Cover Changes Using Remote Sensing and GIS in

Rize, North-East Turkey. Sensors, www.mdpi.org/sensors.

Rimal, B. 2011. Application of Remote Sensing and GIS, Land Use/Land Cover Change in

Kathmandu Metropolitan City, Nepal. Journal of Theoretical and Applied Information

Technology pp. 80-86.

Robinson, N. 1999. The use of focus group methodology – with selected examples from

sexual health research. Journal of Advanced Nursing 29(4): 905-913.

Rosell, S. 2011. Regional perspective on rainfall change and variability in the central

highlands of Ethiopia, 1978-2007. Applied Geography (31): 329-338.

Sall, M. 2014. Land-based Adaptation and Resilience Powered by Nature.

Sang, L., Zhang, C., Yang, J., Zhu, D. and Yun, W. 2011. “Simulation of land use spatial

pattern of towns and villages based on CA-Markov model”. Mathematical and

Computer Modelling (54): 938-943.

Sayemuzzaman, M. and Jha, M. K. 2014. Modeling of Future Land Cover Land Use Change

in North Carolina Using Markov Chain and Cellular Automata Model. American

Journal of Engineering and Applied Sciences 7 (3): 295-306.

82

Sege, C. 1994. Population and Income. A Global perspective. pp 263-285. Cambridge

university press: Cambridge.

Seyoum, B. A. 2012. Land Use /Land Cover Change and its Effect on Existing Forest

Condition: The Case of Shakiso Natural Forest, Southeast Ethiopia. MSc. thesis,

Wondo Genet College of Forestery and Natural resource, Wondo Genet, Ethiopia

Shiferaw, A. 2011. Evaluating the Land Use and Land Cover Dynamics in Borena Woreda of

South Wollo Highlands, Ethiopia. Journal of Sustainable Development in Africa 13

(1): 87-107.

Shrestha, S. 2012. Drivers of Land Use Change: A Study from Sundarijal Catchment,

Shivapuri Nagarjun National Park. MSc. thesis, Kathmandu University.

SOS Sahel Ethiopia, 2010. Annual Accomplishment Report for 2009. Submitted to Charities

and Societies Agency (CSA), Addis Ababa, Ethiopia.

Tefera, M. M. 2011. Land-use/land-cover Dynamics in Nonno District, Central Ethiopia.

Journal of Sustainable Development in Africa 13(1): 123-141.

Tesfaye, S., Guyassa, E., Raj, A. J., Birhane, E. and Wondim, G. T. 2014. Land Use and Land

Cover Change, and Woody Vegetation Diversity in Human Driven Landscape of

Gilgel Tekeze Catchment, Northern Ethiopia. International Journal of Forestry

Research.

83

Tesfaye, Y., Roos, A., Campbell, B. M. and Bohlin, F. 2011. Livelihood strategies and the role

of forest income in participatory-managed forests of Dodola area in the Bale

Highlands, southern Ethiopia. Forest Policy and Economics (13): 258-265.

Teshoma, A. J. 2010. Determinating Factors for a Successful Establishment of Participatory

Forest Management: A Comparative Study of Goba and Dello Districts, Ethiopia.

MSc. thesis, University of Agder, Kristiansand, Norway.

Teshome, E., Kinahan, A. and Randall, D. 2008. Land Cover Change Study of the Bale

Mountain National Park. Frankfrt Zoological Society, Bale mountains National Park,

Robe, Ethiopia.

Tsegaye, D., Moe, S. R., Vedeld, P. and Aynekulu, E. 2010. Land-use/cover dynamics in

Northern Afar rangelands, Ethiopia. Journal of Agriculture, Ecosystems and

Environment (139): 174–180.

Tsegaye, L. 2014. Analysis of Land Use and Land Cover Change and Its Drivers Using GIS

and Remote Sensing: The Case of West Guna Mountain, Ethiopia. International

Journal of Remote Sensing and GIS 3(3): 53-63.

Turner, B. L., Skole, D., Sanderson, S., Fischer, G., Fresco, L. and Leemans, R. 1995. Land-

Use and Land-cover change Science/Research Plan. Joint publication of the

International Geosphere-Biosphere Programme (Report No. 35) and the Human

Dimensions of Global Environmental Change Programme (Report No. 7). Stockholm:

Royal Swedish Academy of Sciences.

84

Turner, II B. L. and Meyer, W. B. 1994. Global land-use and land-cover change: An

overview. In changes in land-use and land-cover: A global perspective. Cambridge:

Cambridge University Press. 3-10.

Veldkamp, A. and Lambin, E. F. 2001. Predicting land-use change. Agriculture, Ecosystems &

Environment 85(1): 1–6.

Verburg, P., Schot, P. P., Dijst, M. J. and Veldkamp, A. 2004. Land use change modelling:

current practice and research priorities. GeoJournal (61): 309–324.

Verheye, W. 2011. Land Cover, Land Use and the Global Change. Journal of Land Use, Land

Cover and Soil Science I.

Wakjira, D. T. 2015. Migration and Conservation in the Bale Mountains Ecosystem. IISD

report.

Walellegn, A. 2007. Assessments of land use land cover dynamics at Bale Mountains

National Park using GIS and remote sensing. MSc. thesis, Addis Ababa University,

Ethiopia.

Watson, C. 2013. Forest Conservation for Communities and Carbon: The Economics of

Community Forest Management in the Bale Mountains Eco-Region, Ethiopia.

Doctoral thesis, London School of Economics and Political Science.

Weng, Q. 2002. Land Use Change Analysis in the Zhujiang Delta of China Using Satellite

Remote Sensing, GIS and Stochastic Modeling. Journal of Environmental

Management (64): 273–284.

85

WIREs Climate Change, 2014, Interactions between Climate Change and Land use Change on

Biodiversity: Attribution Problems, Risks, and Opportunities (5): 317–335.

Woldeamlak, B. 2002. Land Cover Dynamics since 1950s in Chemoga Watershed, Blue Nile

Basin, Ethiopia. Mountain Research and Development 22(3): 263-269.

Worku, G., Amare, B. A. and Habtamu, T. H. 2014. Land use and land cover change in

Ameleke Watershed, South Ethiopia. Journal of Natural Sciences Research 4(14): 42-

47.

World Bank, 2007. Measuring the Economic Impact of Climate Change on Ethiopian

Agriculture: Ricardian Approach. Washington DC: World Bank.

Zeleke, G. and Hurni, H. 2001. Implications of land use and land cover dynamics for

mountain resource degradation in the northwestern Ethiopian highlands. Mountain

Research and Development 21 (2): 184–191

86

Appendix 1: Land Use/Land Cover Change Matrixes

Table 7: LU/LCC matrix between 1986 and 1996

Table 8: LU/LCC matrix between 1996 and 2006

LU/LC category Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/range

land

Total

Water 2338.56 97.83 1.62 6.48 6.66 3.42 2454.57

Agriculture/

settlement

20.88 214139.07 40739.49 38897.64 36262.89 11915.73 341975.7

Woodland 25.29 70902.72 265369.68 20436.93 34214.94 77083.02 468032.58

Forest 3.15 28648.89 12296.79 365633.19 3832.2 457.47 410871.69

Scrub/bush

land

1.62 32275.17 16096.05 7543.89 75922.56 8526.42 140365.71

Grassland/range

land

0.45 29315.79 56881.08 3434.04 16579.71 107156.16 213367.23

Total 2389.95 375379.47 391384.71 435952.17 166818.96 205142.22 1577067.48

LU/LC category Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/range

land

Total

Water

Agriculture/

settlement

2308.5 162789.39 58020.03 14238.54 22263.48 11356.02 270975.96

Woodland 61.74 69540.75 344180.34 9161.1 29208.87 112993.74 565146.54

Forest 53821.89 9482.49 386518.05 11250.99 2913.66 463987.08

Scrub/bush land 63.63 26077.77 8739.36 286.74 66548.25 8807.58 110523.33

Grassland/range

land

20.7 29745.9 47610.36 667.26 11094.12 77296.23 166434.57

Total 2454.57 341975.7 468032.58 410871.69 140365.71 213367.23 1577067.48

87

Table 9: LU/LCC matrix between 2006 and 2016

LU/LC category Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/range

land

Total

Water 1114.11 1189.53 35.19 1.35 14.85 24.67 2379.70

Agriculture/

settlement

5.94 231135.12 38836.98 30663.18 31252.59 43555.97 375449.78

Woodland 79315.56 162246.96 11691.36 23052.87 115027.80 391334.55

Forest 0.18 60393.15 21925.62 334540.98 14343.39 4734.89 435938.21

Scrub/bush

land

0.09 51106.86 7168.23 1415.07 87410.70 19746.38 166847.33

Grassland/range

land

0.27 21205.08 38241.81 490.68 7172.73 138007.34 205117.91

Total 1120.59 444345.30 268454.79 378802.62 163247.13 321097.05 1577067.48

Table 10: Transitional probability matrix derived from LU/LC map of 2006 and 2016

LU/LC category Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/range

land

Water 0.9996 0.0002 0.0000 0.0000 0.0002 0.0000

Agriculture/

settlement

0.0000 0.3330 0.0800 0.2517 0.1987 0.1366

Woodland 0.0000 0.3336 0.1188 0.2650 0.0391 0.2436

Forest 0.0000 0.4010 0.1170 0.2352 0.0998 0.1470

Scrub/bush

land

0.0000 0.4571 0.1223 0.1873 0.0909 0.1425

Grassland/range

land

0.0000 0.3294 0.2232 0.0395 0.0235 0.3844

88

Table 11: LU/LCC matrix between 1986 and 2016

LU/LC

category

Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassing/range

land

Total 1986 Loss

Water

Agriculture/

settlement

1033.20 157625.19 41285.61 16083.00 26283.51 28917.72 271228.23 113603.04

Woodland 32.40 150049.53 177652.71 12964.05 37294.29 189922.77 567915.75 390263.04

Forest 75989.34 19465.56 342949.77 18219.69 7362.72 463987.08 121037.31

Scrubland/bush

land

47.43 23228.55 3831.75 1169.37 71774.55 7450.20 107501.85 35727.30

Grassing/range

land

7.56 37452.69 26219.16 5636.43 9675.09 87443.64 166434.57 78990.93

Summary 837445.861

Total 2016 1120.59 444345.30 268454.79 378802.62 163247.13 321097.05 1577067.48

Gain 1120.59 286720.11 90802.08 35852.85 91472.58 233653.41

Net

change(NC)2

1120.59 173117.07 -299460.96 -85184.46 55745.28 154662.48

Net persistence

(NP)3

1.10 -1.69 -0.25 0.78 1.77

1 sum of diagonals and represents the overall persistence, 2 NC = gain−loss. 3 NP = net change/diagonals of each class

89

Appendix 2: Error Matrixes

Table 12: Error matrix for the LU/LC map of 1986

Reference

Data

Classified

Data

Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/range

land

Total Users

accuracy

--------------- --------------- --------------

-

--------------

-

--------------

-

---------------

Agriculture/

settlement

21 2 1 1 0 25 84%

Woodland 3 50 0 1 2 56 89.29%

Forest 2 1 44 0 0 47 93.62%

Scrub/bush

land

3 0 0 6 0 9 66.67%

Grassland/range

land

2 2 1 0 8 13 61.54%

Total 31 55 46 8 10

Producers

accuracy

67.74% 90.91% 95.65% 75% 80%

Overall Classification Accuracy = 86.00%

Kappa (K^) statistics

---------------------

Overall Kappa Statistics = 0.8065

90

Table 13: Error matrix for the LU/LC map of 1996

Reference

Data

Classified

Data

Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/

rangeland

Total User

accuracy

--------------- --------------

-

--------------

-

--------------

-

--------------

-

--------------

-

-------------

--

Water 1 0 0 0 1 100%

Agriculture/

settlement

0 30 4 2 1 0 37 81.08%

Woodland 0 2 45 0 2 1 50 90%

Forest 0 0 0 37 0 0 37 100%

Scrub/bush

land

0 1 1 0 8 0 10 80%

Grassland/range

land

0 0 6 0 0 9 15 60%

Total 1 33 56 39 11 10

Producer

accuracy

100% 90.91% 80.36% 94.87% 72.73% 90%

Overall Classification Accuracy = 86.67%

Kappa (K^) Statistics

---------------------

Overall Kappa Statistics = 0.8212

91

Table 14: Error matrix for LU/LC map of 2006

Reference

Data

Classified

Data

Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/range

land

Total Users

accuracy

--------------- --------------

-

-------------

--

--------------

-

--------------

-

--------------

-

---------------

Water 0 0 0 0 0

Agriculture/

settlement

0 35 2 6 2 0 45 77.78%

Woodland 0 2 32 0 0 0 34 94%

Forest 0 0 0 46 0 0 46 100%

Scrub/bush

land

0 0 0 0 9 0 9 100%

Grassland/

rangeland

0 1 3 0 0 12 16 75%

Total 0 38 37 52 11 12

Producers

accuracy

92.11% 86.49% 88.46% 81.82% 100%

Overall Classification Accuracy = 89.33%

Kappa (K^) Statistics

---------------------

Overall Kappa Statistics = 0.8576

92

Table 15: Error matrix for LU/LC map of 2016

Reference

Data

Classified

Data

Water Agriculture/

settlement

Woodland Forest Scrub/bush

land

Grassland/range

land

Total Users

accuracy

--------------- --------------

-

--------------

-

--------------

-

--------------

-

--------------

-

---------------

Water 0 0 0 0 0

Agriculture/

settlement

0 29 5 0 4 3 40 70%

Woodland 0 0 23 1 0 0 24 95.83%

Forest 0 1 0 28 1 0 30 93.33%

Scrub/bush

land

0 0 0 0 25 1 26 96.15%

Grassland/range

land

0 0 2 1 0 26 28 92.86%

Total 0 30 30 30 30 30

Producers

accuracy

93.33% 76.67% 93.33% 83.33% 86.67%

Overall Classification Accuracy = 86.67%

Kappa (K^) Statistics

---------------------

Overall Kappa Statistics = 0.833

93

Appendix 3: Field Observation Sheet Format

Field Observation Sheet

General Data

Observers Name Date and Time

Sample

ID Location Position Altitude (m) Photo taken

LU/LC

type

Region___________ Lat. (X)__________ Alt. (Z)_______ _________jpeg

Zone____________ Long. (y)___________

Wereda___________

Kebele____________

Appendix 4: Checklist for Focus Group Discussion (FDG) and Key Informant Interview

(KII)

1. Checklist for Focus Group Discussion (FDG)

Administrative Unit: Region: _______, Zone: ______, Wereda: _________

Name of Rural Kebele _________________________

Village Name _____________________

No. of participants____________________

Date ___________________________________

1. What are currently existing land use/land cover types in your locality?

2. What does the land use/land cover type of your locality looks like before 30 years, 20

years and 10 years?

94

3. Which land use/land cover type is increasing and which is decreasing starting from 1986

to this time, why?

4. On which period do you observed a rapid land use/land cover change, why?

5. What kind of land use/land cover change do you expect in the future? And why?

6. What are the direct/proximate drivers of land use/land cover change over the last 30

years, between 1986 & 1996, 1996 & 2006 and 2006 & 2016? (Options provided:

infrastructure development and urban expansion, forest encroachment for illegal and

legal settlement, overgrazing, agricultural expansion, occurrence of fire and

unsustainable harvest of forest products (like firewood, charcoal, logging)

7. Which land cover is highly affected by each proximate (direct) drivers of LU/LCC?

8. What are the underlining causes along each proximate driver?

2. Checklist for Key Informant Interview (KII)

1. Have you noted any change in the land use/land cover in your area over the past 30

years? A) yes B) No

2. If your answer to question number 2 is yes, what changes did you observed?

Increase/decrease in:

A) Agricultural land

B) Forest cover

C) Woodland

D) Scrub land

E) Bush/shrub lands

F) Grassland

G) Rangeland

H) Settlement and infrastructure

95

3. What are the causes behind their increase/decrease?

I. Direct causes

II. Indirecte (root) causes

4. Participation of the local communities, government and non-government organizations

in resource conservation and management activities and how they are participating?

5. How do you evaluate the livestock and human population size in your area in the last 30

years? What do you think the cause for population dynamics in your area?

6. What major technological change occurred in the area of infrastructural development

and farming activities in the last 30 years?

7. Do you think national policies and institution implemented starting from 1986 until

toddy have responsibility for land use/land cover change? If yes how?

8. What major natural calamities occurred in your area in the last 30 years?

9. Means of land acquisition (tenure) in your PA before 30 years, 20 years and 10 years?

10. What are the main economic activities for the local communities in your area before 30

years, 20 years and 10 years?