University Free State
IIII~I~IIlIffl III~IfllllllIlll~llllllllllllllllllllllllfllllllllfllll~1111
34300001320120
Universiteit Vrystaat
RANGELAND EVALUATION IN RELATION
TO PASTORALISTS PERCEPTIONS IN THE MIDDLE
AWASH VALLEY OF ETHIOPIA
ABULE EBRO GEDDA
RANGElLAND EVALUATION ]fN 1RElLATKON TO
lPA§TOIRAlLKSTS PERCEPTIONS ]fN T1H[EMKID>ID>lLE
AW AS1H[V AlLLE\' OF ET1H[KOPJLA
By
ABULE EBRO GEDDA
Submitted in partial fulfilment of the requirements for
The degree of
DOCTOR OF PHILOSOPHY
in the Faculty of Natural and Agricultural Sciences
Department of Animal, Wildlife and Grassland Sciences
(Grassland Science)
University of the Free State
Bloemfontein
Promoter: Prof. H.A. Snyman
Cc-promoter: Prof. G.N. Smit
January 2003
DEDICATION
This thesis is dedicated to the memory of my mother, Askale Yigelatu,
who had been fully committed to my success with strong prayer for the
betterment and success of my life as a whole.
TABLE OF CONTENTS
ABSTRA.CT i
DECLARA TJ:ON v
ACKNOWLEDG EMENTS vi
LIST OF TABLES viii
LIST OF FIGURES xi
LIST OF APPENDICES xv
ABlBREV][A TIONS xvi
CHAPTER li
1. ffiTRODUCTION 1
CHAPl'ER2
2.lLI1rlERA llJRE REVIEW 5
2.1. IN"TRODUCTION 5
2.2. RANGELAND AND JPASTORALISM IN" ETHIOP][A 6
2.2.1. The resource base 6
2.2.2. Past development interventions 8
2.2.3. Constraints related to the pastoral production system ,.l 0
2.2.4. Current attitude towards pastoralism 12
2.2.5. Development interventions undertaken in the Awash
Valley 13
2.2.5.1. Large scale irrigated agriculture 13
2.2.5.2. Nationally protected areas and conservation 14
2.2.6. Lake Beseka ........................................•.............•... 17
2.2.7. Livestock, feed and!rangeland resource base of the study
area 18
2.2.8. Migration and rangeland management in the study
tiistricts 21
2.3. WILDLIFE CONSERVATION AND PASTORALISM 22
2.4. PASTORALISM AND OJPPORTUNITIC MANAGEMENT 24
2.5. EQUILmRIUM VERSES NON-EQUILmRIUM DYNAMICS IN
PASTORAL PRODUCTION SYSTEM 26
2.6. RANGELAND CONDITION ASSESSMENT 28
2.6.1. Herbaceous vegetation 28
2.6.1.1. Significance and historical development of
rangenand condition assessment 28
2.6.1.2. Criteria used in rangeland condition
assessment 30
2.6.1.2.1. Botanical composition 31
2.6.1.2.2. Basal cover 31
2.6.1.2.3. Plant vigour 32
2.6.1.2.4. Biomass 32
2.6.1.2.5. Bush encroachment 32
2.6.1.2.6. Soil erosion 33
2.6.1.2.7. Soil compaction 33
2.6.1.2.8. Soil parameters 33
2.6.1.3. Techniques used in rangeland condition
assessment and their limitations 34
2.6.1.3.1. Subjective 35
2.6.1.3.2. Quantative 35
2.6.2. Woody vegetation importance and their assessment
techniq nes 39
2.6.2.1. The importance of woody vegetation ill]
the drylands 39
2.6.2.2. Browse production 41
2.6.2.3. The importance of trees in ecosystem functioning
and their role in rangeland condition 44
2.6.2.4. Bush encroachment 49
2.6.2.5. Common methods used in the determination of
browse production 52
2.7. THE CONCEPT OF RANGELAND GRAZING CAPACITY][N
PASTORAL PRODUCTION SYS 55
CHAPTER3
3. STUDY AREA 58
3.1. LOCATION AND AREA COVERAGE 58
3.2. CLIM.A TE 60
3.3. WATER RESOURCES 63
3.4. GEOLOGY AND soas 64
3.5. FAUNA AND PALENTOLOGICAL VALUE 64
3.6. VEGETATION 65
3.6.1. Grassland 57
3.6.2. Shrub/bushland 66
3.6.3. Riverine vegetation 66
3.7. PASTORAL GROUPS AND HUMAN POPULATION 66
3.7.1. The Afars 66
3.7.2. The Kereyu 67
3.7.3. The Ittu 68
3.7.4. Human population 60
3.8. RELATIONSlIllIJP AMONG THE TRmES AND OCCUJPATION ... 69
3.9. TERMmOLOGY 69
CHAPTER4l
4. PASTOlRALIlSTS PERCEPTIONS OF lRANGELAND
RESOURCE UTILISATION AS RELATED TO LIVESTOCK
PRODUCTION .. o 0000 •• 0.00 ••••••• 000 •••••••••••••• 0 •••• 0000.0.00 •• 0. 0 •• 000000.72
4.1. WTRODUCTION , 72
4.2. PROCEDURE 73
4.2.1. Sur-vey design .........••.•..........••••........•..•.•...........•.... 73
4.2.2. Data analysis 74
4.3. RESULTS 75
4.3.1. Demographic of the pastoralists surveyed! and main source
of income 75
4.3.2. Livestock ownership in the study districts .... 0 •••••••••••• 077
4.3.3. Pastoralists perceptions to rangeland resources
and utilisation 78
4.3.3.1. Rangeland plants 78
4.3.3.1.1. Poisonous plants 78
4.3.3.1.2. Grasses 83
4.3.3.1.3. Woody plants 85
4.3.3.1.4. Bush encroachment ... 0 •• 0 •••• 0 ••••••••••••••• 86
4.3.3.1.5. Rangeland 88
4.3.3.1.6. Perceptions towards rangeland
condition 91
4.3.3.2. Water resources 93
4.3.3.3. Natural minerals 97
4.3.4. Feed resources 99
4.3.5. Migration 102
4.3.6. Rangeland use conflicts I06
4.3.7. Resource utilisation from conservation area I08
4.4. D][SCUSSION 108
4.4.1. Demography of the pastoralists surveyed I08
4.4.2. Livestock ownership 109
4.4.3. Poisonous plants 110
4.4.4. Use of plants and Bush encroachment II2
4.4.5. Rangeland and perceptions towards rangeland
condition 113
4.4.6. Natura! minerals 117
4.4.7. Water resources 117
4.4.8. Feed resources 118
4.4.9. Migration 119
4.41.10.Rangeland resource use conflict I20
4.4.11. Resource utilisation in conservation area I2I
4.5. CONCLUSIONS AND RECOMMENDATIONS I22
CHAPTERS
5. RANGE LAND EVALUATION IN AWASH-FANTALE AND
KEREYU- FANTALE DISTRICTS 0.0000 •••••••• 0 •• 123
5.1. IN'TRODUCT][ON 123
5.2. PROCEDURES 124
5.2.1. Site selection 124
5.2.2. Field layout 125
5.2.3. Data collection 126
5.2.3.1. Floristic composition of the herbaceous layer .... 126
5.2.3.2. Basal cover 129
5.2.3.3. Grass dry matter yield 129
5.2.3.4. Soil erosion 129
5.2.3.5. Woody vegetation sampling 130
5.2.3.6. Soil sampling and chemical analysis 130
5.2.4. Data analysis 131
5.2.4.1. Grass species composition and other related
parameters 131
5.2.4.2. Woody layer 133
5.2.4.3. Statistical analysis 135
5.3. Results 135
5.3.1. Awash-Fantale district 135 .
5.3.1.1. Herbaceous ' 13 5
5.3.1.1.1. Grass species composition I35
5.3.1.1.2. Bare ground 136
5.3.1.1.3. Basal cover 136
5.3.1.1.4. Grass dry matter yield 136
5.3.1.1.5. Estimated soil erosion 139
5.3.1.1.6. Grazing capacity 140
5.3.1.1.7. Rangeland condition 140
5.3.1.2. Woody vegetation 141
5.3.1.2.1. Woody vegetation composition, density,
ETTE and the palatability of woody
plants 141
5.3.1.2.2. Browse production 143
5.3.1.2.3. Browsing capacity 148
5.3.1.3. Soil Parameters 152
5.3.1.4. Correlation matrix among the studied!
variables 154
5.3.1.5. Multiple regression between grass DM yield and
other parameters 154
5.3.2. Kereyu- Fantale district 156
5.3.2.1. Herbaceous layer 156
5.3.2.1.1. Grass species composition 156
5.3.2.1.2. Bare ground 156
5.3.2.1.3. Basal cover 159
5.3.2.1.4. Grass DM yield 159
5.3.2.1.5. Estimated soil erosion 159
5.3.2.1.6. Grazing capacity 159
5.3.2.1.7. Rangeland condition 160
5.3.2.2. Woody layer 161
5.3.2.2.1. Woody vegetation composition, density,
ETTE and palatability values 161
5.3.2.2.2. Browse production ..............•........... 164
5.3.2.2.3. Browsing capacity ....................•...... 168
5.3.2.3. Soil parameters 171
5.3.2.4. Correlation matrix among variables
studied and multiple regression of grass DM
yield in relation to other parameters 172
5.4. DISCUSSION 174
5.4.1. Grass species composition 174
5.4.2. Basal cover, bare ground and estimated soil erosion
values 178
5.4.3. Grass DM yield, grazing capacity and rangeland
condition 181
5.4.4. Woody vegetation composition 183
5.4.5. Browse production and browsing capacity 184
5.4.6. Bush encroachment 187
5.4.7. Soil 190
5.4.7.1. pH, electrical conductance and texture 190
5.4.7.2. Percent total Nitrogen 190
5.4.7.3. Organic carbon 191
5.4.7.4. Carbon Nitrogen ratio 192
5.4.8. Correlation among the variables studied and multiple
regression between grass yield and other parameters 192
5.5. CONCLUSIONS 193
CHAPTER6
6. CONDrrKON OF THE COMMUNAlL RANGElLANDS
iN RJElLATiON TO BENCHMARK
SiTES 195
6.1. ][NTRODUCTION 195
6.2. PROCEDURE 196
6.2.1. Identification of benchmark sites 196
6.2.2. Data collection and analysis 197
6.3. RESULTS 198
6.3.1. Rangeland condition assessment 198
6.3.2. Grass dry matter yield and basal cover 198
6.3.3. Grazing capacity ' 202
6.3.4. Soil parameters 202
6.3.4.1. Percent sand, silt and clay 202
6.3.4.2. Percentage total nitrogen 202
6.3.4.3. Percent Organic carbon (OC) ................•......• 202
6.3.4.4. Soil compaction 204
6.4. DISCUSSION 205
6.4.1. Rangeland condition 205
6.4.2. Grass dry matter yield, basal cover and grazing capacity205
6.4.3. Nitrogen, Organic Carbon and soil compaction 206
6.5. CONCLUSION 207
CHAPTER i
i.T]HJ:EIrNlFlLU1ENCE OlF TREES ANlIJ) lLlIVESTOClK
GIRAZIrNG ON GRASS SPECKES COMPOSKTKON,
VllELlIJ) A.NlIJ)SOHLo0 0 0000000000000000000000000000000000000000000000000000000000000208
7.1. ][NTRODUCTION 208
7.2. PROCEDURE 209
7.2.1. Site, trees and sub-habitats selection 209
7.2.1.1. Lightly grazed site 209
7.2.1.2. Medium grazed site 210
7.2.1.3. Heavily grazed site 210
7.2.1.4. Tree and sub-habitat selection 210
7.2.2.Sampling of herbaceous vegetation 211
7.2.3. Soil sampling and analysis 211
7.2.4. Data analysis 212
7.3. RESULTS 212
7.3.1. Grass species composition 212
7.3.2. Grass dry matter yield ..•.......•..•............•.....••..........•• 216
7.3.2.1. Heavy grazed sites 216
7.3.2.2. Medium grazed site 216
7.3.2.3. Light grazed site 216
7.3.3. Soil parameters 217
7.3.3.1. Heavy grazed site ~ 217
7.3.3.2. Medium grazed site 21 7
7.3.3.3. Light grazed site 217
7.4. DISCUSSION 219
7.4.1. Effect of grazing on grass layer and soil parameters 219
7.4.2. Effect of tree species on the grass layer and soil.. 220
7.4.3. Effect of sub-habitats on grass layer and soil parameters220
7.4.4. Effect of soil depth on soil parameters 221
7.5. CONCLUSION 222
CHAPTER 8
8. CONCLUDING REMARKS AND RJECOMMJENDATIONS .... 223
CHAPTER 9
9. SUMMARY 234
REFERENCES 237
AJPPEND!C1ES 285
RANGElLAND EVALUATKON IN RE1LAI10N TO
JPA§TOJRAlLliST§ JPERClE1PI10NS iN THlE Mli][))][))lLlE
AWASH VAlL lLlEY OlF lETHliOPiA
By
Abule Ebro Gedda
Promoter: Prof. HA Snyman
Ce-promoter: Prof. GN Smit
DEPARTMENT: Animal, Wildlife and Grassland Sciences
Degree: Doctor of Philosophy
ABSTRACT
Pastoralism is the most dominant land use form in the arid rangelands of Sub-
Saharan Africa in which Ethiopia is not an exception. However, in Ethiopia and
elsewhere, rangeland-based life-styles are in difficulty with the rangeland
environment under threat because of both external and internal constraints. The
spatial variability of the annual rainfall in these areas also has an affect on the
pastoralists livelihood. Accordingly, four studies were undertaken in two
neighbouring districts occupied by pastoralists of different ethnic groups living in the
middle Awash valley of Ethiopia with the objective of evaluating the condition of the
rangelands, which was related to the perception of the pastoralists in order to come
up with possible recommendations to minimize further degradation.
The pastoralists perceptions of the rangeland resource were studied through group
discussions and by using a structured questionnaire where each household was taken
as a unit of analysis (90 households from Oromo living in Kereyu-Fantale district
and 55 households from Mar living in Awash-Fantale district). The data was
analysed using Statistical Package for the Social Sciences (SPSS). The result showed
that the average family size per household was about 6.74, with nearly 80% of the
people without any kind of education. The main source of income for both pastoral
11
groups was from the sale of animals. The second source of income to the Oromo and
Afar pastoralists was from the sale of crops and milk and milk by-products,
respectively. Both pastoral groups reported that woody species like Cryptostegia
grandiflora, Capparis fascicularis, Erythrina abyssinica and Flueggea virosa) and
herbaceous species like Tribulus terrestris, Tephrosia subtriflora and Cynodo are
sources of poisons which affect their livestock production. Ninety seven and 3% of
the Oromo respondents use Cymbopogon commutatus and Chrysopogon plumulosus
for house roofing respectively, while 38.1%, 23.0%, 10.6% and 28.3% of the Afar
pastoralists use C. commutatus, C. exacavatus, Enterpogon and Sporobolus ioclados,
respectively for a similar purpose. Seventy six percent of the Oromo and 77 % of the
Afar respondents do not harvest grasses from the rangelands and the primary use of
woody plants in both pastoral groups was for livestock feeding. It was indicated that
the grazing lands were bush encroached notably with Acacia senegal, A. nubica and
Prosopis juliflora (Awash-Fantale district only) and the condition of the rangeland to
be in poor condition. None of the Afars and only 12% of the Oromo pastoralists had
private grazing lands. The majority of the respondents chose to continue with
communal type of ownership in the grazing lands and a shortage of water was a
critical constraint to the Oromo pastoralists. There is a critical shortage of livestock
feed during the dry season and the first measure taken to solve feed shortage is
migration. Unfortunately, 90% of the Oromo and 60% of the Afar respondents
replied that migration is a bad practise. The Afar pastoralists (cattle = 20; sheep = 12;
goats = 26; Camels = 15) had a higher number of livestock owned per household
than the Oromo pastoralis (cattle = 10; sheep = 8; Goats = Il; Camels = 5).
Rangeland condition in terms of grass, browse and soil parameter was studied at Il
sites in Awash- Fantale district and 10 sites iri Kereyu -Fantale district using
techniques and/or methods mostly developed in South Africa. Grazing and browsing
capacities were also calculated for each of the rangeland sites. The most dominant
grass species-in the study districts was Chrysopogon plumulosus followed by
different species of Sporobolus. The percentage bare ground as estimated by the
point method varied from 0.33 to 10.79 with a mean value of 5.27. The basal cover
in both districts was low, averaging 3.35%. The DM yield of the grass ranged
111
between 168.52 kg ha" to 832 kg ha-I. The grazing capacity varied from as low as
54.14 ha LSU-I to as high as 7.06 ha LSUI. The results of the evapotranspiration tree
equivalent (ETTE ha") showed that the study districts were bush encroached with A.
senegal, A. nubica and P. juliflora. In both districts, the browse production (total leaf
DM) ranged from as low as 194 kg ha" to 3 311 kg ha-I, with most of the leaf dry
mass found above the height of 1.5 m. In both districts, the highest browsing capacity
(ha BU-I) was contributed by A. senegal and A. nubica.
The condition of the communal grazing lands was also assessed m relation to
benchmark sites. Basal cover and the DM yield of grasses was higher in the
benchmark sites (basal cover= 5.3% and DM yield of grasses = 985.7 kg ha") than
the sample sites (basal cover = 3.3% and DM yield of grasses = 447.2 kg ha"), which
indicated that given proper management, there is ample room to improve the grazing
capacity of the rangelands.
With the objective of studying the effects of tree species' on grass species
composition, yield and some soil parameters under different grazing gradients (light,
medium and heavy) in two sub-habitats (under canopy and open grassland), two tree
species (Acacia tortilis and Balanites aegyptica) were identified. The data was
analysed using DECORANA and SAS (Statistical Analysis System). The results
showed that the grass species found at the heavily grazed sites were mostly annuals
and less desirable species. The major difference between the medium and lightly
grazed site in grass species composition was the presence of Panicum maximum
under the canopy of trees in lightly grazed condition. The DM yield of grass
improved substantially as the grazing intensity decreased (heavy = 31l. 9 kg ha",
medium = 1 607 kg ha-I and light = 2 737.5 kg ha"). At the medium and lightly
grazed sites, the DM yield of grass was higher (P<O.OOl) under tree canopies than
the corresponding open grasslands. Soil nutrient status increased as the grazing
pressure decreased from heavy to light grazing. Electrical conductance, percentage
nitrogen and organic carbon increased (P<O.Ol) under tree canopies compared to the
corresponding open grasslands whereas they decreased with an increase in the depth.
of soil.
IV
In conclusion, all studies with different objectives and arguments clearly indicate that
the condition of the rangelands IS poor, requiring careful and participatory
interventions. Future studies need to distinguish between climate and man-made
droughts although droughts are a normal phenomenon in these drier areas.
Rangelands in poor condition increase the intensity and frequency of climatic
droughts.
v
DECLARATION
I declare the dissertation hereby submitted by me for the partial fulfilment of the
requirement of the degree Doctor of Philosophy (Grassland Science) at the university
of the Free State is my own independent work and has not previously been submitted
by me at another university/Faculty. I furthermore cede copyright of the dissertation
in favour of the University of the Free State.
Abule Ebll"oGedda
VI
ACKNOWLEDGEMENTS
First and foremost I would like to thank the heavenly father God for his love,
kindness, mercy and for the opportunity He gave me to continue my study. I have no
words of thanking Him except saying what King David has said in 2 Samuel 7: 18 "
Who am I, 0 Lord God? And what is my house, that You have brought me this far?"
My sincere appreciation and thanks goes to my supervisors' proff. H.A. Snyman and
G.N. Smit for their tireless guidance, advice and encouragement in designing the
study, preparation and write up of the thesis. Their suggestions, supervision and
constructive criticisms during the study contributed much to the successful
completion of the study. I also highly appreciate both of them for the kindness with
which they attended my personal problems. I would also like to extend my special
thanks to the other staff members of the Grassland Science Discipline (UFS) for their
strong support and assistance through out the study period. The suitable working
atmosphere offered to me by all members of the Discipline is greatly appreciated. I
would also like to thank dr. Johan Du Preez (Department of Plant Science, UFS) for
his kind assistance in data analysis.
I am deeply indebted to the Oromiya Regional Government and the Ethiopian
Agricultural Research Organization (EARO) for the privilege given to me to
undertake my study. I would like to express my deepest gratitude to mr. Aliye
Hussein (Head of the Oromiya Research Institute) and the staff members of the
Institute (mr. Geremaw, mr. Lemma and mrs. Felekech) for their strong moral
support and facilitating the study processes in time: It is also a pleasure for me to
express my sincere thanks to mr. Tefaye Alemu Aredo for his love, moral support
and his clear commitment of my success. I am very thankful to Tesfaye for allowing
me unreserved access to all the facilities in the research center including the timely
provision of vehicles, finance and technical personnel. Special mention must be
made of Assefa Haile Selassie, Bedru, Kebede, Abebe Mebrate, Basiznew, Nesru,
Gebre Medhin, Abebe Gelan and Mengistu Kebede for their contribution in the
fieldwork. I would also like to thank dr. Nesru Hussein and mr. Nigatu Alemayehu
VB
for their encouragement and assistance. I am also indebted to dr. Abera Dheressa for
his encouragement and provision of vehicles and other facilities. I also extend my
thankful gratitude to dr. Azage Tegegne for his encouragement and permitting me to
use facilities within the !LRl Debre Zeit station. I would like to express my deepest
gratitude to Etenesh Yitnaw, Amare Atale and Aklilu Bogale for their unlimited help
throughout the study period. I am very much obliged to drr. Zinash, Kidane, Demel
and Getachew Gebru) for their encouragement and provision of materials. I would
also like to thank the Ethiopian Wildlife Conservation Organization for permitting
me to to use their facilities.
I would like to thank mr. Beruk Walkeba and mr. Mussa, heads of the study districts
agricultural offices, for all kinds they did for me, while collecting data in the field.
The help of Abera, Abomassa, Barudin, Tsigaye, Genet, Fantale, Dahaba, Tagas,
Elias, Qutubi, Kemal, Mohammed and all the others is worth mentioning. I sincerely
appreciate the help I was given from CARE Awash in the form of a vehicle, other
facilities and personnel. I am also grateful to dr. Shimeles Beyene for permitting me
to use their facilities. I would like to thank the Mathara Estate Sugar (Research
Department) for their technical help and permission to use their facilities. They were
very instrumental to my fieldwork. Much of my thanks go to Tadesse Nege and Taye
Eshete for their support and encouragement. I am also indebted to dr. Gezahegn
Ayele and mr. Haile Mariam Twolde for their kind assistance in analysing the
household data. I am very thankful to both pastoral groups for sharing me there time
and knowledge. I would like to express my deepest gratitude to the members of the
Christian Revival Church (Bloemfontein) for there love, encouragement and making
my stay in Bloemfontein very blessing and unforgetable. My special thanks goes to
Pastor At Boshoff for his prayer and teachings. I also thank Tilahun Seyoum, Ketema
Tilahun, Solomon Tefera, Amare Gizaw and Adugna Wakjera for their
encouragement. I remain grateful to my parents, brothers (Girma, Muluken,
Abebayehu, Mandfero and Ketema), sisters (Genet, Kidest, Tsiga, Azeb, Zinash and
Worke) and all others whose list would be too long to enumerate them all, for their
contribution to my success. Finally, I would like to thank my wife Misrak, my son
Krubel and daughter Aden for their full support through out the study period.
VUl
LIST OF TABLES
CJHIAPTE1R2
Table 2.1. Land use practices in the study districts 18
Table 2.2. Tropical livestock unit (TLU) equivalents for Sub-Saharan Africa
(ILCA 1990) 20
Table 2.3. Rate of browse production for a variety of savanna woodland types .43
CHAPTER 3
Table 3.1. Classification of seasons according to the level of precipitation by the
different pastoral groups (Ayele 1986) 63
CHAPTER 4l
Table 4.1. Profile of respondents by family size (mean ± SD), ethnic group, tribe and
.educational background (Respondents: Oromo = 90 and Afar = 55) 76
Table 4.2. Mean number oflivestock species owned per household in the study
districts (Respondents: Oromo = 90; Afar = 55) 77
Table 4.3. Percentage of respondents indicating common plants as poisonous in the
study districts (Respondents: Oromo = 82 and Afar = 54) 79
Table 4.4. Poisonous plants affecting the different livestock species as ranked by the
two pastoral groups in Kereyu-Fantale and Awash-Fantale districts (1 =
animal species highly affected by the specific poisonous plant; 4 = animal
species least affected by the specific poisonous plant; Respondents:
Oromo 82 and Afar = 54) 79
Table 4.5. Percentage of respondents practising different traditional methods of
treating animals poisoned with different poisonous plants (Respondents:
Oromo = 82; Afar = 54) 82
Table 4.6. The uses of woody plants in the study districts as ranked by the
pastoralists (l= highest use and 6= least use; Respondents: Oromo = 89
and Afar = 51) 86
Table 4.7. Reasons for the increase of bushes and shrubs ranked as percentage of the
respondents (1= highest reason 6=lowest reason; the number in
parenthesis indicate % respondents; Respondents: Oromo = 80 and Afar =
49) 87
IX
Table 4.8. Percentage of respondents by type of reasons for the choice of
communal type of ownership in the grazing lands (Respondents: Oromo =
62 and Afar= 47) 89
Table 4.9. Reasons contributing to the poor rangeland condition, as ranked by the
pastoralists (1= Most important reason; Il= Least important reason;
Respondents: Oromo= 87 and Afar= 51) 92
Table 4.10. Pastoralists suggestions on how to improve the rangelands (percentage
of the respondents; (Respondents: Oromo= 70 and Afar= 43) 93
Table 4.11. Type of supplement offered during the dry season by the different
pastoral groups (1= Most used; 5= Least used; Respondents: Oromo= 52
and Afar= 15 101
Table 4.12. Type of feed conserved during the dry season in the study districts
ranked as percentage of the respondents (1= the highest; 4= Lowest).102
Table 4.13. Common diseases observed during migration by livestock species
(percentage of the respondents; Respondents: Oromo= 75 Afar= 42) ..... 106
CHAPTERS
Table 5.1. Rangeland sites identified for the study in Awash-Fantale and Kereyu-
Fantale districts 127
Table 5.2. Desirability, life form and frequency (%) of the different grass species in
the Awash-Fantale district 137
Table 5.3. Mean and standard deviation of bare ground (%), basal cover (%), grass
yield (kg ha-I), estimated soil erosion values and grazing capacity (ha
LSU1 and ha TLU1) for the rangeland sites studied in Awash-Fantale
district (N= 10; HL = Herbaceous layer WL = Woody layer) 139
Table 5.4. Palatability of woody plants as rated by the pastoralists in Awash-Fantale
district : 143
Table 5.5. Physical and chemical characteristics of soils of the rangeland sites in
Awash-Fantale district. 153
Table 5.6. Standard coefficient, rvalues, R2 increment (%), R2 values (%)
and significance of multiple regression with dry matter yield as a
dependent factor 155
Table 5.7. Desirability, life form and frequency (%) of the different grass species
x
in the Awash-Fantale district 157
Table 5.8. Mean and standard deviation of bare ground (%), basal cover (%), grass
yield (kg ha") and estimated soil erosion values for rangeland sites in
Kereyu-Fantale district. 160
'fable 5.9. Palatability of woody plants as rated by the pastoralists in Kereyu-Fantale
district 163
Talble 5.10. Physical and chemical characteristics of soils of the rangeland sites in
Kereyu-Fantale district 173
Table 5.11. Standard coefficient, rvalues, R2 increment (%), R2 values (%) and
signficance with dry matter yield as a dependent factor (Multiple
regression) 174
CHAPTER 6
Table 6.1. Rangeland condition scores in the first group of sample sites (Kachachilo,
Madala 1 and Harole 1) 199
Table 6.2. Calculation of rangeland condition in the study districts (Harole 2, Aleka,
Madala2 and top of bud) 200
CHAPTER 7
Table 7.1. Heights and canopy diameters of the tree species studied 211
Table 7.2. Grass species (yield kg ha") occurring under tree canopies
and in open grasslands in three sites having different grazing
intensities in and near Awash National Park 213
Table 7.3. DM yield of the natural pasture (Mean & SE) under two sub-habitats
along different grazing intensities in and near Awash National Park
(NS= Non-significant; *** = P<O.OOl) 216
Table 7.4. Chemical properities of soils occurring under tree canopy and
open grasslands at two depths at different grazing intensity gradients
in and near the Awash National Park of Ethiopia
(NS= non-significant; * = P<0.05; **= P<O.Ol; ***= P<O.OOl) 218
Xl
LISTS OF FIGURES
CHAPTER 2
Figure 2.1. The pastoral areas of Ethiopia 7
Figure 2.2. Schematic illustration of an ideal tree, its measurements and structure
(Smit 989a) 54
CHAPTER3
Figure 3.1. Location of the study districts in Ethiopia 58
Figure 3.2. Main areas in the study districts 59
Figure 3.3. Annual rainfall (mm) by year in the study districts (1966-2000) 60
Figure 3.4. Summary of mean monthly rainfall (mm), minimium and maximium
Temperature (OC) in the studyarea 61
Figure 3.5. Mean monthly rainfall for the year 2001 for Awash 7 kilo and
Mathara National meterological stations 62
CHAP1rER4
Figure 41.1.Loss of range land due to volcanic eruption in Kereyu-Fantale district and
the plants grown on the erupted volcanic soil are not grazed by
animals 73
Figure 4.2. The main sources of income in the study districts
(Respondents: Oromo= 89 and Afar = 53) 76
Figure 4.3. Example of Cryptostegia grandflora as poisonous plant : 80
Figure 4.4. An Afar carrying his young camel on his back for treatment at a clinic.83
Figure 4.5. Number of respondents giving various reasons for not harvesting grasses
from the rangelands (Respondents: Oromo= 53 and Afar = 31) 84
Figure 4.6. Purpose of using rangeland by the different pastoral groups in the last 10
-20 years (Respondents; Oromo = 69; Afar = 47) 88
Figure 4.7. Type of ownership preferred by the pastoralists for managing their
rangelands in the future 89
Figure 4.8. Percentage of Afar and Oromo respondents utilising different water
sources by season (Respondents Oromo= 88 and Afar= 49) 95
XlI
Figure 4.9. Livestock and human beings utilising the same sources of water at a bore
hole (Oromo) 95
Figure 4.10. An Ororno woman fetching water from the surface ofa shallow
pond 96
Figure 4.11. Calves watered from an irrigation canal (Afar) 96
Figure 4.12. Saline water under preparation for camels (Oromo) 98
Figure 4.13. First measures taken by the different pastoral groups (percentage of
respondents) to solve feed shortage; Respondents: Oromo= 86 and
Afar=52) 100
Figure 4.14. Condition of the browse in the dry season with only Grewia species
retaining some yellowing senescing leaves 103
Figure 4.15. Condition of the browses in the dry season with only Cadaba
rotundfolia retaining leaves 103
Figure 4.16. Perceptions ofOromo and Afar pastoralists towards migration
(Respondents: Oromo= 85; Afar= 47) 105
Figure 4.17. Pastoralist' s perceptions to the intensity of conflict in the last 3-4 years
(Respondents: Oromo= 85 and Afar=51) 107
Figure 4.18. A grass harvested from Awash National Park by an Oromo woman and
transported with a donkey 108
CHAPTER5
Figure 5.1. The most dominant grass species in each of the rangeland sites in
Awash-Fantale district. 138
Figure 5.2. Rangeland condition of the rangeland sites of the Awash-Fantale district,
based upon desirability of species 140
Figure 5.3. Abundance (%) of woody vegetation in Awash Fantale district (based
upon woody plants ha-I) 142
Figure 5.4. Species composition of woody vegetation based on ETTE (%) in
Awash-Fantale district ·.. · 142
Figure 5.5. Browse production stratified into height strata across the rangland sites
Xlll
in Awash-Fantale district 144
Figure 5.6. Estimates of the leaf dry mass (kg ha") at peak biomass, with
subdivision into height strata, of woody plants in Awash-Fantale
district 145
Figure 5.7. Browsing capacity (ha BUl) across the rangeland sites in Awash-
Fantale district 148
Figure 5.8. Browsing capacity (ha BV) for the most important species
based upon their contribution to the browse production in
Awash-Fantale district 149
Figure 5.9. The most dominant grass species in each of the rangeland
sites in Kereyu-Fantale district 158
Figure 5.10. Rangeland condition of the rangeland sites in Kereyu-Fantale
district based upon desirability of species 161
Figure 5.11. The density of woody plants expressed in abundance
(%) in Kereyu-Fantale district.. 162
Figure 5.12. Woody vegetation expressed in terms of% ETTE in Kereyu-Fantale
district. 162
Figure 5.13. Total and stratified leafDM (kg ha-I) in the rangeland sites in
Kereyu-Fantale district. 164
Figure 5.14. Estimates of the leaf dry mass (kg ha") at peak biomass,
with the subdivision into height strata, of woody plants
contributing more than one percent to the total leaf dry
mass in each of the rangeland sites 165
Figure 5.15. Approximate browsing capacity (ha BUl) across the rangelands
studied in Kereyu-Fantale district .168
Figure 5.16. Browsing capacity (ha BV) for the most important
individual woody plants by season and height strata in
Kereyu-Fantale district 169
CHAPTER6
Figure 6.1. Vegetation condition of the junction area between Afar and Ororno
ethnic groups : 196
XlV
Figure 6.2. Grass dry matter yield, basal cover and grazing capacity of sample
sites and the respective benchmark sites 201
Figure 6.3. Soil characteristics at the sample and benchmark sites in the study
districts .203
Figure 6.4. Soil organic carbon and compaction values in the sample and benchmark
sites in the study districts .204
CHAPTER 7
Figure 7.1. Ordination of sample sites by grazing intensity and sub-habitats 214
Figure 7.2. Ordination of the different grass species by grazing intensity and sub-
habitat 215
xv
]LKS]' OF AlP]PlENIlJ)KC}ES
Appendlix 4.1. Questionnaires used in the household survey 285
Appendix 5.1. Ordination of the different grass species using DECORANA in
Awash-Fantale district .294
Appendix 5.2. Ordination of the different grass species using DECORANA in
Kereyu-Fantale district.. .295
Appendix 5.3. Correlation matrix among variables studied in Awash-Fantale district
(*=p <0.05; ** =PO.01; *** =P ::;0.001) 296
Appendix 5.4. Correlation matrix among the variables studied in Kereyu-Fantale
district. .297
XVI
AJBJB:REVIATIONS
ADB African Development Bank
ANP Awash National Park
BTE Browse tree equivalent (g, kg ha -I)
BU Browsing unit
CANVOL Canopy volume
DM Dry matter
EARO Ethiopian Agricultural Research Organization
ETTE Evapotranspiration tree equivalent (500 cnr' , ha-I)
EWCO Ethiopian wildlife conservation organization
FAO Food and Agricultural Organization of the United Nations
!FAD International Foundation for Agricultural Development
LMAS Leaf dry mass (g, kg ha-I)
masl Meters above sea level
MOA Ministry of Agriculture
TE Tree equivalent
UNESCO United Nations Educational, Scientific and Cultural Organization
1
CHAPTER ].
INTRODUCTION
Ethiopia's ruminant livestock population is the largest in Africa and tenth in the
world. Economically, the livestock sub-sector accounts for about 40% of Ethiopia's
Agricultural Gross Domestic Product (GDP), 16% to the total GDP and has
generated an estimated 31% of the total agricultural employment (Mengistu 1996).
Hides and skins accounted for between 12% and 16% of the total export value during
1984-1988 (Addis 1992). In addition, more than 90% of the croplands are cultivated
with the aid of 6 million draft animals (Gryseels & Anderson 1983). Annually, a total
of 0.74 million tonnes of milk and 0.23 million tonnes of beef are produced (FAO
1995). In the pastoral areas, livestock is the sole source of livelihood and accounts
for more than 60% of the household revenue (Zinash et al. 1998). Despite the huge
livestock population, productivity per animal is low for all classes of livestock. This
low level of productivity is caused by poor husbandry and management systems, as
well as the prevalence of diseases and malnutrition (AAMC 1984).
Livestock production in Ethiopia is mainly based on grazing and/or browsing by
cattle, sheep, goats and camels. In the highlands, crop residues and agro-industrial
by-products augment natural pasture. In the pastoral system, livestock production is
almost totally dependent on native pasture and woody plants (Daniel & Tesfaye
1996; Zinash et al. 1998). This can be ascribed to insufficient rainfall to sustain
arable production and the practice of nomadic grazing. The pastoralists have been
able to survive in the harsh conditions of the dry land for centuries because of their
traditional coping mechanisms and development of risk management strategies.
Pastoralism is the most dominant land use form in the dry lands of Sub-Saharan
Africa (SSA) in which Ethiopia is not an exception. There are more than 25 million
people in the SSA to whom pastoralism is not only a system of livestock production,
but also a way of life (Lane 1998). The majority of the pastoralists in Ethiopia
consist of the Somalia's, Borana and Afar living in the southeast, southern and
2
northeast rangelands, respectively. In addition, other ethnic groups include the
Hamors, Arbores, Dannans, Neurs and the Kereyus (EARO 2001). In Ethiopia,
pastoralists and agro-pastoralists inhabit a significant portion of the more arid areas.
An estimate by FAO (1992) put the area covered by the rangelands at 78.1 million
hectares or some 70% of Ethiopia's land area. Of the total livestock population of the
country, the Ethiopian lowlands is estimated to raise 40% cattle, 75% goats, 25%
sheep and 100% camels. The rangelands are not only known for livestock, but there
are many wildlife parks, sanctuaries and reserves (Schloeder & Jacobs 1993; Beruk
& Tafesse 2000; Dawit 2000; MOA 2000; EARO 2001).
The condition of the rangeland areas of Ethiopia and present productivity as
compared to the potential is variable but generally low. The low condition in
productivity resulting from mismanagement is generally manifested by ecological
deterioration of vegetation, soil erosion, lowered fertility, reduced soil water
availability for plant growth and decreased nutritive value of available forages (MOA
1996). The loss of biological productivity is largely due to over-use of shrinking
rangelands by a growing number of pastoralists. Their potential impact can be more
severe than that of farmers because the ecological impact of a 2% increase in pastoral
population is equivalent to a 4% increase of farmers (Williams & Balling 1996;
Lusigi & Acquay 1999). In addition, the spread of cropping, particularly in areas
critical to dry season grazing, uncontrolled grazing around water points and near
villages, as well as cutting of trees for fuel are leading to rangeland degradation.
Despite the vast area of rangeland in Ethiopia, rangeland research and development
are limited only to that of the Borena rangelands (Daniel & Tesfaye 1996).
Accordingly, large-scale development projects related to the pastoral societies were
carried out by government and non-government organizations in southern (mainly),
south-eastern and north-eastern rangeland development units. On the other hand,
some of the areas of the country including the study districts, which are
predominately pastoral, remained without any large-scale pastoral development
projects (Alemayehu 1998). Even in areas with development projects, where millions
of dollars have been spent by international institutions, the general impression is
. rather negative. Very little seems to have been achieved as production has not
3
increased, natural resources continue to deteriorate, the social structures have broken
down and livelihoods have declined (MOA 1996).
The failures in pastoral development projects, be they in Ethiopia or Africa, are
partly attributed to ignorance of the indigenous knowledge of the pastoralists.
However, traditional experiences, skills and strategies accumulated by pastoralists
over the centuries should instead complement modern scientific knowledge rather
than be ignored (MOA 1996; Haan 1999; Muller 1999; Tsundle 1999). Based on an
overview by Haan (1999), it is stated that driven by the political objectives of
providing cheap meat to the urban areas, earlier range livestock development
concentrated on increasing meat output from the pastoral systems. As this was often
contrary to the objectives of the pastoral population, adoption of those technologies
to promote increased meat production almost always failed. Understanding the
production objectives of a particular system also explains the different supply
response to price changes, i.e., whether a system is mostly trade or mostly
subsistence oriented (Pratt et al. 1997) and is thus a critical element in determining
the economic feasibility of any investment.
In comparison with the rangeland resource base of the country, studies with regard to
the condition of the rangeland ecosystem have not received the attention they
deserve. In addition, previous studies have focused mainly on the Borena rangelands
(e.g., Oba 1998; Ayana & Baars 2000) and very few in the others. Furthermore,
numerous advancements have been made in the area of both rangeland assessment
techniques and factors to be used in the assessment, which better aid in the
understanding of changes in rangeland ecosystem dynamics. In any assessment of a
rangeland ecosystem composed of different vegetation components, rangeland
monitoring must incorporate three tiers of assessments i.e., the soil, herbaceous layer
and the tree-shrub layer (Friedel 1991). Despite the significant contribution of woody
vegetation both as a source of feed for domestic and game animals, most of the
assessments fail to consider the contribution of the woody vegetation, not only in
terms of feed resources but also their role in general ecosystem stability.
Unfortunately, there is no detailed rangeland condition assessment considering the
4
tiers of the rangeland resource for grazing lands commonly grazed by the animals of
the pastoralists living in Awash-Fantale and Kereyu-Fantale districts identified for
this study.
Based on a conference held in Ethiopia on pastoralism (Ali 1993; Desta 1993) and
studies conducted by Tibabu (1997); Mudris (1998); Sharon (2000) and a review of
the strategy prepared by EARD (2001) the following points characterise the study
area. The rangeland is mainly used for livestock, wildlife, crop production and
government owned development activities. Accordingly, formerly open access
rangelands have gradually become the focus of conflict between the different forms
of land use. With the added pressure of population growth decreasing the available
land for grazing, losses of useful fodder/pasture species and herbs is quite evident.
There is no adequate database with respect to the vegetation cover, composition of
plant community and their current status particularly for those of the commonly
grazed rangelands by the pastoralist s livestock. To this is added the competition for
scarce resources by different pastoral groups. With increasing human and livestock
populations and the growth in resources being relatively static, the problems
associated with the carrying capacity of the Awash basin are likely to become more
serious, further aggravating the problems of the survival of the different pastoral
groups and the other development activities. Studies undertaken elsewhere (e.g., Ellis
& Swift 1988; William et al. 1990; Behnke et al. 1993; Ekaya et al. 2001) and long
term studies in Ethiopia (Coppock 1994) have clearly indicated that household use of
rangeland resources and associated constraints vary between years and between
seasons of the same year depending on the spatial variability of rainfall.
The main objectives of this study were, therefore, to evaluate the condition of the
rangelands and assess the perceptions of the pastoralists with regard to rangeland
resources, their utilization, constraints and possible solutions in two neighbouring
districts (Awash-Fantale and Kereyu-Fantale districts) inhabited by different pastoral
groups (Afar and Oromo pastoralists) living in the middle Awash valley of Ethiopia
and to investigate possible solutions to minimize further degradation of the rangeland
ecosystem.
5
ClHLAJPTER 2
]L ITTERA TlURE REVITEW
2.1. ][NTRODUCTION
Rangelands are defined as those areas of the world, which by reasons of physical
limitation, low and erratic precipitation, rough topography, poor drainage, or cold
temperatures are unsuited for cultivation and which are a source of forage for free
ranging native and domestic animals, as well as a source of wood products, water
and wildlife. This definition includes grassland as well as savanna and forest areas
often used by grazing animals (Stoddart et al. 1975). Almost half of the land area in
the world, covering about 6.7 billion ha can be classified as pastures and rangelands
(Singh & Ghosh 1993). If all the uncultivated land of the world with potential to
support grazing or browsing animals is taken into account, 70% of the world's land
area can be classified as rangeland (Holechek et al. 1989). More than 200 million
people use rangelands worldwide for some form of pastoral production. About 30-40
million nomadic people in developing countries are wholly dependent on livestock
(WRI, lIER 1989). African rangeland, extending over 3 million km2 support a
pastoral population of 16-20 million (Widstrad 1975) and nearly 500 million head of
livestock (FAO 1975). The majority of the world's pastoralists are in Africa (55%),
Asia (29%), the Americas (15%) and Australia (10%) (Child et al. 1984).
Rangelands are an important renewable resource, in addition to performing a number
of ecological and economical functions, which include provision of humans with
consumable products such as red meat, fibre and water and non-consumptive
services such as recreation and wildlife viewing (Batabyal & Godfrey 2002). Of
these, extensive livestock production is the major land use on rangelands with large
areas of land required per head of livestock (Zhou et al. 1998). Furthermore,
rangelands provide habitats for a wealth of wild and domesticated plant species.
Many plants are of medicinal value and other species may provide important genetic
6
material for future economic use. The genetic pool of species found, may hold
important keys for improving livestock, developing new crop varieties, cunng
disease and numerous other benefits to mankind as yet undiscovered. Thus,
conserving the rich biological diversity of these rangelands is crucial for sustainable
economic development (Miller 1997).
Based on the above facts, this chapter gives a description of the rangelands and
pastoralism in Ethiopia, major production constraints, development interventions
undertaken in the middle Awash Valley of Ethiopia, pastoralism and the different
arguments revolving around their livelihood. Furthermore, a review of literature is
also done on rangeland condition assessment techniques for both herbaceous and
woody vegetation; browse production, the importance of woody vegetation in
ecosystem functioning and their role in rangeland condition and bush encroachment.
2.2. JRANGELAND AND lPASTORALISM IN ETHIOPIA
2.2.1. The resource base
There is no accurate and reliable data indicating the area coverage, human and
livestock population of the pastoral and agro-pastoral production systems in Ethiopia.
However, extensive review works by Coppock (1994); Beruk & Tafesse (2000);
Dawit (2000); MOA (2000); Tafesse (2000); EARO (2001); Getachew (2001)
indicated that it is estimated to cover 61% to 67 % of the country usually below 1
500 meters above sea level (m.a.s.l.) The pastoral areas of the country are divided
into five, i.e., the northeast, eastern, southern, southeast and extreme southern part of
the country (Figure 2.1). Recent estimates by the same reviewers indicated that the
human population is in the range of 7 to 8 million, which is about 13% of the human
population in the country and are composed of 29 Nilotic and Cushitic ethnic groups.
Of this, it is estimated that 93% are considered to be pastoralists or agropastoralists,
while the remaining are either hunter-cultivators or pure cultivators (Beruk &
Tafesse 2000). Furthermore, the density of people km" in the lowlands is about 10 as
opposed to 90 for the highlands (EPA 1998), indicating the potential for future
7
development. Using rainfall and temperature regime, the climate is divided into arid
(64%), semi-arid (21%) and the remaining dry sub-humid (EPA 1998).
The estimates of the eSA (1996) indicates that there are 32 million cattle, 13 million
sheep, 10 million goats, 4 million equines, 1.0 million camels and 33 million poultry
in Ethiopia .
.....,.,.,.,=.,.".".........,.",. ..,.,.',> •••••••••• ", '" .
Limited pastoral use due to trypanosomiasis
Predominantly pastoral but without large scale pastoral
development
111111111111 Predominantly pastoral including large scale development
programmes
1. NERDU (Northeastern Rangelands Development Unit)
2. SERP {Southeastern Rangelands Project}
3. SORDU (Southern Rangelands Development Unit)
Figure 2.1. The pastoral areas of Ethiopia.
Gadamu (1990) estimated that of the total livestock population in Ethiopia, the share
of pastoral and agro-pastoral people is about 40%, 25%, 75%, 100% and 20% of the
cattle, sheep, goats, camels and equine population, respectively. Recent review work
8
by MOA (2000) indicated that the share of pastoral and agro-pastoral areas in terms
of livestock ownership changed to 28% for cattle, 26% for sheep, 66% for goats and
100% for camels. According to this estimate, the lowlands carry about 26% of the
total livestock population of Ethiopia. Although, the lowlands have lower number of
livestock than the highlands, the most important breeds of livestock like the Borana
cattle are found in the lowlands and lowland breeds of cattle and sheep made up over
90% oflegal export of live animals (Coppock 1994).
Not only livestock, but also the presence of more national parks and wildlife
sanctuaries in the lowlands is a clear indication of the potential of the rangelands for
conservation activities. Review work by Beruk and Tafesse (2000); MOA (2000);
Tafesse (2000); EARO (2001) indicated that, except for the Bale and Simen
mountains, the rest of the national parks are situated in the lowlands. Out of the 24
endemic bird species, the lowlands share 19 species with the highlands. In addition,
the same review indicated that, in terms of plant biodiversity, the rangelands are well
known for their diverse plant species. For example, Ogaden region is one of the
richest areas considered for endemic flora, characterized by a high diversity in
Acacia, Baswellia and Commiphora species. The area has about 25% of the plant
species in the country. There are actual or potential mineral resources, such as gold
in the Adola of Borena, natural gas in the Ogaden, salt mines in Afar and Soda ash in
the rift valley. The range1ands are also rich in surface, underground water and large
perennial rivers. The rich deposit of natural gas, geothermal energy is a clear
indication of the rich wealth potential of the pastoral areas. These areas are also of
prime interest for archaeological and socio- anthropological studies (Getachew 2001;
Yacob 2001).
2.2.2. Past development interventions
Many projects were undertaken in the pastoral areas beginning with the 1960's. They
are the Arero rangeland pilot project and the second, third and fourth livestock
development projects. A brief review of the projects is presented in this section.
9
The Arero Rangeland Pilot Project was initiated in 1965 with the aim of improving
rangeland use efficiency through pond construction and controlled grazing. The
project area covered about 2 400 km2 of the rangelands near Yavello. This project
attracted settlement around the new water points, resulting in severe overgrazing. On
top of this, a subsequent drought in northern Kenya led to the migration of Kenyan
Borana pastoralists into the project area and the collapse of the rotational grazing
program. Furthermore, there was little local support for the project (Solomon 1993;
Tafesse 2001; World Bank 2001).
The World Bank as of 1973 funded the second livestock development project
(SLDP). It was initiated to develop an integrated market and stock route system in
order to improve livestock off-take rate. The SLDP created some infrastructure that
contributed to the opening up of pastoral areas and facilitated the expansion of
livestock marketing. Most project-constructed slaughterhouses are still in use and
have been recently privatised. Most terminal markets are also in use, but under
municipal management. However, the stock routes have not been used. In addition,
many of the facilities were destroyed in the Ethio-Somalia war of 1977. This project
did not take into account the traditional stock routes, which would have been
successful otherwise (Solomon 1993; Tafesse 2001; World Bank 2001).
The third livestock development project (TLDP) was established in 1975 with a
blend of World Bank, International Foundation for Agricultural Development
(!FAD) and African Development Bank (ADB) funding. It was the first large-scale
development project and sponsored studies in the Borena, Afar and Somalia
rangelands, implemented by the International Livestock Center for Africa (Il.Cá),
which was also contracted to monitor project impacts (World Bank 2001). The
project was meant to rehabilitate and develop the three areas through its sub-units
namely, North East Rangeland Development Unit (NERDU), Jijiga Rangeland
Development Unit (JIRDU) and Southern Rangelands Development Unit (SORDU)
(Figure 2.1). The project aimed to raise the standard of living of pastoral people
through the restructuring of the traditional system of extensive livestock production.
The project provided, among other things, veterinary services, water and
10
infrastructural development. It continued to provide services with funding from the
Ethiopian Government until 1996. It was abandoned due to the government's
decentralization and regionalization policy.
The fourth livestock development project (FLDP) funded by the World Bank and
IFAD, was focused on upland livestock systems, but included linkages to the pastoral
areas. It was started in 1988 and the project included a comprehensive forage testing
and seed multiplication program, a livestock credit program, vaccine production,
drug supply and epidemiology program and substantive capacity building activities.
It differed from the previous projects in that it took into account the traditional
organizations and indigenous knowledge of the pastoralists. The FLDP, though
promising, collapsed as a result of Ethiopian political turmoil in 1990.
In conclusion, despite these development interventions, human life in the pastoral
areas still continues to be full of uncertainties. The benefit generated has not justified
the investment (Solomon 1993).
2.2.3. Constraints related to the pastoral production system
The major constraints facing the pastoral production systems have been reviewed by
different authors (e.g., Coppock 1994; Alemayehu 1998; Oba 1998; Beruk & Tafesse
2000; MOA 2000; Zinash et al. 2000; EARO 2001; Getachew 2001; Yacob 2001).
Some of the constraints indicated by the different reviewers are presented in the
following paragraphs.
Periodic drought is a characteristic of the lowland pastoral production systems. Even
in normal rainfall years, there are localized parts of the lowlands, which suffer from
drought. As a result of increasing human and livestock population pressure, the
capacity to cope with drought has declined to the point that there is a growing threat
to the survival of viable pastoral production system (World Bank 2001).
The pastoralists living in the rangelands are also threatened by bush encroachment.
Bush encroachment is the process of open grassland savannas being transformed into
Il
thick bush (Barnes 1979) and is one of the major problems threatening the
rangelands (Coppock 1994; Alemayehu 1998; Beruk & Tafesse 2000; EARO 2001).
According to Coppock (1994) cited from Hacker (1990) roughly 15 species of
woody plants are thought to be encroachers in the Borena plateau. Of these, Acacia
drepanolobium and A. brevispica are the most important. The lack of prescribed
burning, accompanied by severe overgrazing and the expansion of farming in the
dryland were among the main problems associated with bush encroachment.
According to Oba (1998), 40% of the Borena rangelands were estimated to be bush
encroached. Beruk & Tafesse (2000) with reference to the Mar region indicated that
an increase of Acacia seyal, A. mellifera, A. senegal, and Prosopis juliflora are of
major concern.
Because of the diverse culture of the people, the abundance of wildlife and plant
resources, scenic value and the existence of archaeological sites, pastoral areas are
well known as tourist attractions. Accordingly, a total land area of 0.4 7 million ha of
the rangeland has been converted to wildlife sanctuaries, i.e., 0.35 million ha (Mar
region), 0.06 million ha (Southern region), 0.051 million ha (Gambella Region)
(Beruk & Tafesse 2000). Studies in other countries have clearly indicated that,
depending on marketing arrangements; wildlife can generate greater wealth at lower
economic and environmental costs than livestock and arable agriculture and thus be
profitable to rural sectors (Kiss 1990; Cumrning & Bond 1991; Jansen 1992).
Furthermore, Cole (1990) indicated that wildlife can increase the revenue earned per
kg of animal in the rangeland and wildlife ranches in dry regions can earn three times
more per ha than cattle ranches. Research shows that wild animals gain weight more
quickly than domestic stock, breed faster and have leaner meat. Unfortunately,
because of poor tourism development, wildlife conservation capability, the lack of
the involvement of the pastoralists and the absence of benefit sharing from
conservation activities, the wildlife enterprise is blamed for not benefiting the
pastoralists living nearby the conservation areas. The resultant effect being severe
ecological deterioration affecting both livestock and wildlife development activities.
12
In addition to the above reasons, the expansion of large-scale commercial farms
without due consideration to the benefits of the local pastoralist is considered a threat
to the livestock production system. These interventions can help the country in many
ways. However, great concern must also be given to the pastoralist's welfare and the
ecology of the rangelands.
Livestock are the principal and most productive investments of the pastoralists with
few alternative investment opportunities. However, the difficulties of access to
markets and the high costs of marketing and other related problems have hindered
proper livestock marketing in the pastoral system (Sintayehu 1993; World Bank
2001). Moreover, inadequate delivery of veterinary services both in terms of area
covered and supplies is among the major problems of the pastoralists (Wario 1993;
Tefesse 2001).
Inter and intra clan conflict over the rangeland resources, mainly grazing land and
water points especially during the dry season, has contributed to the decline of the
resources. This is a common feature of the pastoralists of the Afar, Somalia, Boran,
Neurs, Kereyu and ethnic groups in the South Omo. Human and livestock losses are
inevitable. In most cases, the ultimate result will be a reduction in the overall size of
the traditional pastoral territory (Desta 1993; Tibabu 1997; Mudris 1998; Beruk &
Tafesse 2000; Getachew 2001; Yacob 2001).
2.2.4. Current attitude towards pastoralism
In the past, Ethiopian government policy towards pastoralists had never been spelt
out clearly in any pastoral or arid-land policy document or strategy. It was only
implicit in the kinds of projects that were implemented in pastoral areas (Tafesse
2000). Currently, the government is taking measures that indicate the concern
towards the welfare of the pastoralists, like the establishment of national pastoral
extension team, the emphasis given in the research system to pastoral areas and the
issues of proclamations regarding pastoral welfare (Beruk & Tafesse 2000).
13
2.2.5. Development interventions undertaken in the Awash Valley
Natural resources of the rangelands of the study districts are subject to three
competing claims: development to generate revenue for the state, conservation of
wildlife and wilderness areas as well as use for local production. Furthermore, the
expansion in the size of Lake Beseka is further decreasing available land in the area.
In the following sections, the development interventions undertaken are reviewed.
2.2.5.1. Large scale irrigated agriculture
The history of large scale irrigated agriculture in Ethiopia dates back to the beginning
of the 1960' s when the potential of the Awash River for irrigated agriculture was
realised. In order to speed up the development of irrigation agriculture, the Imperial
Ethiopian Government established the Awash Valley Authority (AVA), which had
little concern for the need of the pastoralists (UNDPIRRC 1984). The Awash river
basin is the most developed part of Ethiopia in terms of commercial agriculture,
having 0.07 million ha of land already under cultivation and an estimated potentially
irrigable land close to 0.21 million ha (Halcrow 1989). The valley is divided into
uplands, upper, middle and lower basins. Of the basins, it is the upper and part of the
middle Awash that is of concern to this study. The dominant perennial crop in the
upper valley is sugar cane while fruits, vegetables and cotton are produced in the
middle valley. Past irrigation developments in both the upper and the middle basin
have been responsible for the loss of the Kereyu and Mar peoples most critical dry
season grazing habitat, which forced the pastoralists to move their settlements to the
outer margins of the irrigation schemes, in those areas that were once seasonally
exploited (Schloeder & Jacobs 1993). The state owned farms that are operating in the
study area are Methara Sugar Estate, Nura Era State Farm, Awara Melka
Agricultural Development and Middle Awash Agricultural Development Enterprises
(MAADE).
The Methara Estate Sugar was established based on the agreement of the Ethiopian
government and a Dutch company called H.V.A. (Handels Vereniging Amsterdam).
It covers an area of Il 000 hectares of land (Waktola & Micheall 1999). Studies by
14
Tibabu (1997) and Mudris (1998) indicated that of the total Il 000 employees in
1996 the contribution of Methara Sugar Estate in terms of employing the local
pastoralists is Il % and they are mainly guards. Admission to the school is only
allowed to the children of the employees. Irrigation by the Methara sugar plantation
has denied the Kereyu and Ittu pastoralists access to critical watering points along the
river banks and has almost eliminated all traditional dry season forage sites (
Schloeder & Jacobs 1993).
The other farm operating in the area is the Nura Era State Farm. This farm is located
in the vicinity of the Kereyu and Ittu pastoralists. Nura Era fruit and vegetable farm
covers a total land area of 3 277 ha of which 2 900 ha are in actual production
(Waktola & Micheal 1999). Studies by Tibabu (1997) and Mudris (1998) indicated
that of the 4 141 workers employed in the farm in 1996, the Oromo pastoralists were
3.6%.
The two farms operating In the Afar area are the Awara Melka Agricultural
Development Unit and the Middle Awash agricultural development enterprises. The
Awara Melka agricultural development is a fruit and vegetable farm occupying 1 200
ha of which 820 ha is under cultivation. Tibabu (1997) studied the number of people
employed in this farm in 1996 and indicated that of the total employee, l3% were
Afars, which are working mainly in guarding, cotton picking, weeding and some
other activities (Waktola & Micheal 1999). In addition, the same authors indicated
that the relationship between the farm and the local communities appeared to be
better than that with other state farms. Communities benefit from the farm through
social and technical services and they get free health care at the clinic. According to
Lane et al. (1993), the middle Awash development enterprise has a land area of 30
400 ha which is a major loss of dry season grazing lands. Moreover, it has denied
pastoralists access to about 30 km of the Awash River.
2.2.5.2. Nationally protected areas and conservation
In Ethiopia, awareness of wildlife as a valuable resource began as early as 1940' s
(Negarit Gazeta 1944). With the growing awareness on the importance of wildlife,
15
there came a need to manage wildlife on a sustainable basis. Wildlife conservation in
the "modern sense" and as a state owned enterprise started during the regime of His
Imperial Majesty Haile Selassie I. The United Nations Educational, Science and
Cultural Organisation (UNESCO) supported the idea of the government and assisted
in the identification of those areas with potential for wildlife development. One of
these, the Awash conservation area is located at the point where the Ethiopian Rift
Valley joins the Afar triangle.
The Awash National Park (ANP), which is located in the centre of the conservation
area, is approximately 756 km2 in size (EWCO 2000). According to UNESCO
(1964) the justifications for the choice of Awash as a national park rested on its extra
ordinary interesting feature from both the physiographic and geological point of
view. lts proximity to the capital city of Ethiopia (Addis Ababa) also strengthens the
justification. Furthermore, the area had already been protected as a private hunting
reserve for His Majesty Emperor Haile Sellasie I and there was an abundance of
game in the area (Petrides 1961).
The establishment of the park placed it within the classification of a "strict
conservation area" defined as excluding all kinds of human use of the area like
settlement, exploitation of natural resources and grazing" (Moore 1982). Currently,
in the park there are more than 81 species of mammals, 453 species of birds (6 of
them endemic), some reptiles and invertebrates. In addition, the existence of wild
animal species such as Beisa oryx, Soemmerrings Gazelle, Greater Kudu, Lesser
Kudu, Salts dikdik, Deferssa waterbuck, Warthog, Mountain reedbuck, Klipspringer,
Grevys zebra, Bushbuck and others add a beauty to the park (EWCO 2000).
The establishment of the park required the resettlement of the people and their
livestock from their original places and it also coincided with the establishment of
the sugar cane plantation leaving little alternative to the local community (Schloeder
& Jacobs 1993). Some efforts to resettle the people were unable to solve the
problem. The attempts made were primarily to buy 1 500 ha of land in the Borchata
area, which was out. of the traditional territory of the pastoralists. His Imperial
16
Majesty also granted 3l.1 ha of land from the king's private reserve ofland along the
Kesem River, which was in the traditional boundary of other tribes. Tribal conflicts
accompanied by the lack of water in the new sites blocked the use of the land given
to the Kereyu pastoralists. In general, the efforts made were short lived and there was
no compensation made there after (Schloeder & Jacobs 1993; Tibubu 1997; Mudris
1998).
With the subsequent changes of government in Ethiopia, conservation activities have
been affected. Accordingly, in 1975 following the change in government, the
pastoralists challenged the management policies of the Awash National Park (ANP).
The reasons for this were primarily in response to the settlement and grazing
exclusion practices, which the ANP staffs were trying to put into effect. The absence
of rainfall for two consecutive years resulted in the permission of access to the
different sections of land occupied by the park (Schloeder & Jacobs 1993). The
challenge from the side of the pastoralists continued further with 67% of the park's
land being effectively settled on or used by the various pastoralist groups, either
seasonally or permanently. Furthermore, several of the conservation areas were
overrun and the wild animals poached. Such changes and challenges have seriously
affected the development of the wildlife industry in the area.
Throughout this long period negotiations have been undergoing between the
pastoralists and the management of the park at different times to solve the problems
of the conservation activities and the pastoralists livelihood (particularly on issue of
grazing lands). At times, the issues move to Central Government both directions
presenting their problems. A new management plan is now under revision as former
plans (Robertson 1970; Jacobs & Schloeder 1993) have proved inappropriate and it
is hoped that it will be ready within three years.' It is likely that zonation of the park
will occur in a similar manner that was suggested in the 1993 management plan
(Sharon 2000). Furthermore, efforts to solve the problems of the pastoralists as well
as the conservation activities are being made by non-government organizations
notably the Awash Conservation and Development Project (CARE Awash).
17
In general, state intervention to create exclusive wildlife reserves without the
pastoralists obtaining viable alternatives has been the greatest failure in the
development of the wildlife industry. Furthermore, the lack of participation of the
local community in the conservation process and its further development just served
to further isolate the communities and has resulted in negative impacts to both human
society and the conservation areas (Schoelder & Jacobs 1993; Tibabu 1997; Mudris
1998). The studies undertaken by Tibabu (1997) and Mudris (1998) showed that
government owned interventions caused a change in land use practices of the
pastoralists from pastoralism to crop farrning, charcoal and firewood selling and
others forms of living.
2.2.6. Lake Beseka
In the study area, land is further decreased due to expansion of lake Beseka. It
occupies a small rift between two opposing sets of faults and as a result, water is able
to percolate up through the porous lava blocks to form a large standing body of saline
water rendering its use as a source of water for livestock and human beings
(Schloeder & Jacobs 1993). Moreover, due to poor irrigation infrastructure, practices
and over grazing in the catchments area in the last 20 years, the lake has increased
from 3.3 km2 to 35 krrr', with an average annual increase in height of 0.12 m
(EVDSA 1990). The effect of this being a decrease in the grazing land of the Kereyu
pastoralists. Currently, the government has planned and budgeted to let the lake flow
into the Awash River.
In general, there is a lack of accurate and reliable data on the size of the land
occupied by the different enterprises and the reports are also contradicting. The
available information is shown in Table 2.1.
18
Table 2.1. Land use practices in the study districts.
Existing land use Area (ha)
Methara sugar plantation (Waktola & Micheal 1999) Il 000
Nura Era state farm (Waktola & Micheal 1999) 3277
Awara Melka state farm (Waktola & Micheal 1999) 1 200
Middle Awash agricultural development enterprise 30400
(Lane et al. 1993)
Lake Beseka ( Schloeder & Jacobs 1993) 3500
Awash National Park (Schloeder & Jacobs 1993) 75600
2.2.7. Livestock, feed and rangeland resource base of the study area
Animal production plays a major role in the economy of pastoral society (Singh &
Ghosh 1993). Unlike the other pastoral areas, there are a number of additional factors
that affect the livestock, feed and rangeland resource base of the area and the
associated use of these resources. Even in a normal pastoral production system, there
are several changes in management practices between the different seasons of the
same year and between different years because of change in the amount of
precipitation (Ellis & Swift 1988; Behnke et al. 1993; Coppock 1994).
When raising the issue of livestock in the study districts, it is not only the pastoral
groups that are the utilizers of the rangeland resources, but also people living in the
town and the different plantations own a considerable amount of livestock that
heavily depend on the rangeland and feed resource of the area. With an increase in
human population, both in the town and in the different plantations, there is an
accompanied increase in the need for more land for crop production and forage
resources, which in turn, have created severe competition even among the
pastoralists of the same ethnic group, which at times result in heavy dispute and
conflict. The rangeland resource is also shared between the various components of
the ecosystem including both livestock and wild animals. Added to the different
interacting factors operating in the study districts, over the last 5 years there was a
19
general decline in the amount of rainfall, which might affect the rangeland resource
base of the area (Methara Estate Sugar, meteorological data, 1966-2001). The
influence of rainfall on herbaceous production in semi-arid areas is well documented
(Shiflet & Dietz 1974; Warren et al. 1984; O'Connor 1994; Ekaya et al. 2001;
O'Connor et al. 2001). Based on this account, the livestock feed and rangeland
resource base of the area need to be assessed to indicate the changes in space and
time.
The pastoralists in the study area are transhumants keeping cattle, sheep, goat and
camels like the rest of the pastoral society in Ethiopia. A transhumant pastoral are
those types of pastoralists which are not nomadic, but instead maintain permanent
settlements. They do, however, move their livestock seasonally in order to exploit
areas away from the permanent settlement sites. The entire village rarely moves with
the herders in these instances (Halcrow 1989). Like all pastoral areas in Ethiopia, the
animals owned are used for milking, slaughtered for meat, sold for cash or bartered
for other commodities.
Similar to the general situation in Ethiopia, there is no accurate and reliable data
regarding the livestock population of the area. The available information is also
limited in time to indicate changes that can occur under the conflicting resource use
pattern of the study area and a decline in the amount of precipitation in consecutive
years. Livestock populations for the Afar, Kereyu and Ittu pastoralists living in and
around the park were 35 139 cattle, 61 804 sheep and goats, 8 282 camels and 1 076
donkeys (Schloeder & Jacobs 1993). This when converted to Tropical Livestock Unit
(TLU) equals 24 597 for cattle, 6 180 for sheep and goats, 8 282 for camels and 538
for donkeys. This conversion is based on the TLU conversion for Sub-Saharan
Africa, which is shown in Table 2.2. A Tropical Livestock Unit is the equivalent of
one bovine animal of 250 kg live weight. In the rest of the thesis this conversion
factor is used unless specified.
Mudris (1998) showed a livestock ownership of 13, 10, 10,3 and 1 for cattle, sheep,
goat, camel and equines per Kereyu and Ittu household, respectively. For the Afars
20
living in Awash- Fantale district, Tibabu (1997), reported an average livestock
ownership per household of 23.4 for cattle, 47 for sheep and goats, 19.0 for camels
and 0.23 for donkeys.
Table 2. 2. Tropical livestock unit (TLU) equivalents for Sub-Saharan Africa (Il.Cé,
1990).
Species Average biomass (kg) TLU equivalents
Camels 250 1.0
Cattle 175 0.7
Sheep and goats 25 0.1
Horses and mules 200 0.8
Donkeys 125 0.5
To the pastoralist, his domestic animals are considered family and are also an
important medium of social exchange. A pastoralist with many animals can be
generous to his friends and relatives giving them animals during ceremonies, when
they are ill, or purely as a sign of friendship. He can help poorer households by
giving or lending them animals. A man with many animals can afford to marry more
wives and have more children. He can also take in impoverished friends or relatives
as dependants, adding to his prestige and his labour force. The Maasai say a
successful man is like a tree on a hot sunny day, he shelters many people under his
shade (Solomon et al. 1991). Thus, the study of livestock ownership in pastoral
communities is very important, though the pastoralists are relactant to tell the exact
number of their livestock.
Poor nutrition caused by inadequate amounts and poor quality of feed is one of the
major causes of low livestock productivity in tropical areas (Kaitho 1997).
Subsequently, a detailed assessment of the available feed resources, their extent of
usage by the categories of users, season of usage, constraints that are faced in the
usage and alternative measures taken during feed shortage need to be clearly known
in such a way that it can assist in solving the problems associated with the
degradation of the rangeland. To this effect, there is a limitation of information
21
regarding the utilisation of the different feed resources in the area. The Methara
Estate Sugar produces annually a total of 0.37 million sacks of fine molasses and
some other by-products, which could have helped much in alleviating the further
degradation of the available rangeland (Mathara Estate Sugar 2001). Unfortunately,
most of the molasses is used for export, ethanol production and for the production of
alcoholic drinks. It is also used for cattle feeding but mainly by enterprises outside of
the study districts.
2.2.8. Migration and rangeland management in the study districts
Intact relationships exist between the pastoralist and his environment based on the
availability of rainfall and the resulting availability of grazing. The seasonal
availability of green pasture and its rapid exhaustion owing to heavy grazing are the
main reasons for the migratory movements of the pastoralists in the study area. The
timing, direction and the extent of their movement are primarily dictated by rainfall.
Although previous studies indicated the existence of migration in the study districts
(Desta 1993; Tibabu 1997; Mudris 1998) there is no detailed information to describe
the problems associated with migration, livestock disease problems linked with
migration and pastoralists perceptions on migration which are key elements m
rangeland resource use pattern and decisions regarding their future situation.
Land in Ethiopia was formally brought under State ownership in 1975. However,
most state owned rangeland is in practice managed as common property and is
usually controlled by an ethnic group, clan or other social units. They in turn apply
rules and traditional social control, which regulate the access and coordinate, the
action of individual managers in the utilization of the natural resources (Simonsen &
Mitiku 1998). In both pastoral groups, kinship is concerned with rights over property
and succession in a local community. Like the rest of pastoralists in Ethiopia, these
pastoralists have their traditionally set social organization that is involved in
rangeland management. In both cases, village heads, sub clan leaders and clan
representatives, elders (Gada leaders in the Oromo groups) play a major role in the
protection and utilization of the rangeland (Desta 1993; Tibabu 1997; Mudris 1998).
22
2.3. Wll.,DLIFE CONSERVATION AND PASTORALISM
The problem associated with state owned conservation and pastoralism is not unique
to Ethiopia. Although the problem is not wide spread globally, it exists in a much
smaller area in Afiica, namely part of the rangelands of eastern and southern Africa.
Therefore, a brief review of the situations in these countries will be important in
order to have a wider outlook of the problem and the interventions undertaken to
solve the problem.
Livestock and wildlife co-existence means the joint occupation and utilization of the
same general land area for any period during the course of the year (Bourn & Blench
1999). On the basis of this definition, pastoralists and their livestock have probably
co-existed with wildlife in Afiica for some 7000 years. It may be difficult to
perfectly establish the relation between wildlife and human beings at that time, yet it
is likely that few of the tensions evident today were present.
In East Africa, nationally protected areas cover some 193 000 km2 or 12.7% of the
region with the largest owned by Tanzania, followed by Kenya and Uganda. Wildlife
is a major tourist attraction and foreign exchange earnings of countries such as
Botswana, Kenya, South Africa, Tanzania, Zambia and Zimbabwe (Pearce 1997).
Hence, extensive land areas have been set-aside in these countries for some form of
conservation management (e.g., as much as 25% of Tanzania, 30% of Zambia, 12%
of Zimbabwe and 10% of South Africa). In South Africa, private reserves and
integrated livestock-wildlife ranches are common, accounting for almost half the
total conservation areas. Game ranching is also rapidly increasing in Zimbabwe and
Nambia (Cumming 1990a,b,c; Cumming & Bond 1991; Winrock international 1992;
Child B. 1995; Child G. 1995a, b).
Extensive reviews of the problems currently facing pastoralism and conservation
were made by various researchers (Schulz & Skonhoft 1996; Bourn & Blench 1999).
They indicated that human population growth, agricultural expansion, deforestation,
hunting and the ramification of economic development have had profound,
cumulative impacts on the environment, natural habitats and wildlife population all
23
over the world. For example, the wildlife population of Kenya has fallen by a third
over the last two decades and wildlife has also been eliminated from much of the
Uganda, including many protected areas.
The establishment of national parks and game reserves excludes and has often
directly displaced rural communities from their land, which traditionally was theirs.
It is argued that, due to implicit notions of nomadic pastoralism as uneconomic and
environmentally destructive deeply founded in Western agricultural mentality, the
pastoral people have been caught in an accelerating poverty process (Johnsen 1998).
Similar problems have been noted in Kenya and Tanzania (Lane & Swift 1989; Coe
& Goudie 1990; Enghoff 1990; Homewood & Rodgers 1991; Prins 1992; Kamugisha
& Stahl 1993; Lane & Moorehead 1994; Mustafa 1997).
The general trend over much of the East African regions has been for livestock and
game to be managed separately. Pastoral livestock are excluded from the great
majority of such areas in Africa today, although multiple uses are allowed in some
protected areas, such as Amboseli National Park and Ngorongoro Conservation
areas.
Conservationists argue that livestock tends to displace wildlife from grazing areas
and convert habitat to less productive and biologically diverse states (Coe et al.
1976). Contrary to this idea, eliminating livestock grazing and pastoral management
practices, such as burning, from areas in Ngorongoro Conservation area, has
detrimentally altered vegetation for wildlife, favouring less nutritious and less
palatable species (Homewood & Rodgers 1987, 1991). Livestock and wildlife exploit
different (but overlapping) ecological niches in time and space and have evolved
different physiological and behavioural strategies to reduce competition (Infield
1996). At the same time, there has also been a realisation that relying solely on
protected areas to conserve biodiversity is insufficient. Wildlife is a fugitive
resource, with some species migrating seasonally, ignoring boundaries delimited by
protected areas. For example, in Kenya three-quarters of all large mammals are
24
found outside protected areas for at least part of the year. Conservation and wildlife
management must, therefore, be extended beyond the protected areas.
In order to solve the above-mentioned problem many countries have tried different
methods. Strategies for the integration of livestock and, wildlife rely on the benefits
of diversifying livelihood through non-consumptive and consumptive sustainable
use. Among them are commercial game ranching, sharing resources with the local
communities around national parks and community based wildlife management
practices such as the communal areas management programme for indigenous
resources (CAMPFIRE) in Zimbabwe. Although many problems are associated with
the programme it is now being used as a model for similar programmes in a number
of other countries in SSA (Bourn & Blench 1999).
2.4. PASTORALISM AND OPPORTUNISTIC MANAGEMENT
Recent research works undertaken In communal grazing lands, has shown that
fluctuation in yield of natural forage in drylands areas is affected far more by
variability in rainfall than by grazing pressure (Warren & Khogali 1992; Behnke et
al. 1993). Thus, the marginal nature of pastoral environments has imposed certain
constraints to livestock production and settlement patterns. Livestock are bred for
their resilience to drought and disease rather than their productivity. Human and
livestock movement is a response to seasonal variations in forage availability. Herd
diversification is common, with many herd owners herding a variety of different
stock in different areas. Not only do different animals have different niche
specialisations, but also have different vulnerabilities to drought. Such diversification
helps reduce overall vulnerability to drought and disease and herd growth tends to be
opportunistic rather than conservative. Thus, communal ownership of the rangeland
is an adaptation to variability (Hogg 1997; UNDP 1997).
Movement of animals in response to spatial and temporal variation In resource
availability is perhaps the most classic of all the tracking strategies (Swallow 1994).
It is used as a means to adjust local imbalances in stock numbers and to make use of
25
the seasonal availability of pasture and water (Smith 1992; Behnke et al. 1993), thus
indicating that pastoralists are strategists (Dyson-Hudson 1980). Rather than
manipulating herd numbers in response to climatic variability, as would a rancher
operating in an enclosed area, pastoralists move and so shift their resource
endowments (Oba 1998). In addition, pastoralists almost invariably have some
access to crop residues and other agricultural by-products. Catch cropping by
pastoralists often results in more fodder than grain. In addition, grazing arrangements
between pasoralists and agriculturalists have long been a route for pastoralists to gain
access to farm resources (Toulmin 1992; Powell & Williams 1993).
International development agencies and African governments have devoted
considerable effort to the suppression of pastoral techniques of land and livestock
management. These programmes were undertaken on the presumption that
pastoralism was inherently irrational, ignorant, unproductive, ecologically
destructive and largely or solely responsible for poverty, overgrazing and the general
degradation of the land. It is also associated closely with the failure of development
schemes (Hardin 1968; Picardi & Seifert 1976; Livingstone 1977; Sandford 1983).
The main reasons associated with these ideas are that communal farming inevitably
results in over-grazing because it is totally unregulated free-for all which Hardin
(1968) named the "tragedy of the commons". There are no controls to ensure the
continued viability of land held in communal ownership. In addition, according to
Sandford (1983) decisions are motivated by tradition, rather than rationale or
scientific knowledge.
Contrary to the above assumptions, with regard to herd productivity, comparative
studies of ranch and pastoral herd output in West Africa (Breman & De Wit 1983),
Southern Africa (De Ridder & Wagenaar 1984) and East Africa (Western 1982;
Cossins 1985) demonstrate that pastoralism either equals or exceeds the productivity
per unit land area of commercial ranching in comparable ecological environments.
Furthermore, many studies of traditional communal farming do not support the free-
for-all assumption (Netting 1976; Gilles & Jamtgaard 1981; Sinclair & Fryxell
1985). The existence of complex social mechanisms, as well as ecologically based
26
management strategies which regulate the use and distribution of resources by the
pastoralists is well documented (Schapera 1943; Bodley 1985; Ellis & Swift 1988).
In fact it has been shown that many pastoral groups have through their indigenous
organization, been able to control access to rangeland and that some of the longest-
used areas are least degraded. A wealth of anthropological studies over the past few
decades have convincingly demonstrated that most such practices are part of an
internal rationality or logic which permit people to optimally utilize available
resources. Perhaps more importantly, they support social systems of reciprocity,
which have become adapted to the local conditions over time (Bodley 1985).
2.5. EQUamRIUM VERSES NON-EQUamRlUM DYNAMJ[CSIN
PASTORAL PRODUCTJ[ONSYSTEM
The terms equilibrium and non-equilibrium as used in rangelands, are points of
strong debate among scientists working in rangeland ecology, particularly when
applied to African pastoral ecology. The central aspect of this debate is the definition
of the degree to which climate or consumers influence environmental trends. One
view is that consumers reach densities that degrade environments from a previous
condition of equilibrium (Lamprey 1983) and the other side is that the dynamics of
pastoral systems are non-equilibrium and primarily dictated by variability in rainfall
(Ellis & Swift 1988).
Based upon a review (Scoones 1995b) indicated that the equilibrium system assumes
that range-livestock systems operate in environments that are stable or equilibrium in
nature, where climate variability is not important, i.e., plant-growing conditions are
relatively invariant over time. Moreover, vegetation change is gradual, following the
classical climax concept of Clementsian succession (Clements 1916; Stoddart et al.
1975). Livestock populations are in turn limited by available forage in a density-
dependent manner, so that excessive animal numbers, above a carrying capacity
level, result in negative effects on the vegetation (Miller 1997). The concept of an
equilibrium system first developed and applied in North America became the basis
for planning livestock use on rangelands in pastoral systems (Miller 1997). Lamprey
27
(1983) from the long-term studies in Tanzania, Coppock (1994) from Ehtiopia and
Meadows & White (1979); Evangelou (1984); Grandin (1987) from Kenya Maasai
land support the concept of equilibrium ecosystem dynamics.
The concept of equilibrium has been challenged both from a theoretical and practical
point of view (Hyder et al. 1966; Glieck 1987; Horowitz & Little 1987; Ellis & Swift
1988; Nicolis & Prigogine 1989). Ecological research in the last decade in semi-arid
rangelands, where climatic variability is high and ecosystem functions very dynamic,
suggests that most arid and semi-arid rangeland ecosystems function as non-
equilibrium systems. In these areas, plant growth and rangeland productivity were
found to be more functions of climate than of livestock stocking rates, and the effect
of livestock on rangeland vegetation being more sporadic than continuous. Given the
array of publications in widely circulated literature, the new science is rapidly being
established as a dogma and could subsequently be incorporated into policy decisions
(Cousins 1996; Dikeni et al. 1996). However, recently Campbell et al. (2000) from
Zimbabwe examined one tenet of the new science that pastoral systems give higher
economic returns than other systems. They compared the economics of four cattle
management practices. They reported that strategies based on conservative stocking
rates would have higher net present values than strategies based on opportunistic
stocking rates or management. They further identified several serious flaws in the
papers that elevate opportunistic pastoral systems as giving higher economic returns
than other systems, opening more room for continuation of further research In
equilibrium and disequilibria concepts as related to the pastoral production system.
Researchers differ in terms of opiruon regarding as where the non-equilibrium
dynamics occur. Some relate it to the coefficient of variation (CV) of annual rainfall,
i.e., when the CV is greater than 35%, the ecosystem tends to be non-equilibrium.
Others, on the other hand, suggest that areas that receive less than 300-400 mm of
rainfall annually will operate as non-equilibrium systems (Miller 1997).
28
2.6. AANGELAND CONDITION ASSESSMENT
2.6.1. Herbaceous vegetation
2.6.1.1. Significance and historical development of rangeland condition
assessment
Since the inception of the word rangeland condition, different people have tried to
give different kinds of definitions. However, it is usually concerned with the "state of
the health" of the vegetation, i.e., what the vegetation should look like under normal
climate and optimum management. In a more complete sense, it includes the direct and
indirect changes in vegetation composition land productivity and land stability over
time under various regimes of livestock production (Bailey 1945; Parker 1951; Short
& Woolfolk 1956; Tueller & Blackburn 1974; Pratt & Gwynne 1977; Tainton 1981;
Wilson & Tupper 1982; Trollope et al. 1990; NRe 1994; Tainton 1999).
Rangeland condition and trend are perhaps the most important concepts III the
management of renewable resources in arid and semi-arid lands. There are three
objectives in rangeland condition assessment namely; evaluation of rangeland condition
relative to its potential in that ecological zone (Tainton 1999); evaluation of the effects
of current management on rangeland condition and monitoring changes over time
(Hacker 1974; Trollope 1990; Tainton 1999) and classification of the different
vegetation types on the farm and quantification of their condition (Tainton 1999). This
is mainly done based on a concern for the long-term productivity and stability of the
rangelands (Wilson & Tupper 1982). As pointed out by Tainton (1988) and Stuart-
Hill & Hobson (1991) interpretation of the condition and grazing capacity assessments
allows the user to define possible management options to improve the condition and
therefore grazing capacity of the rangeland. It is, therefore, not surprising that a lot of
effort has been put in the ways of developing appropriate techniques for rangeland
condition assessment.
It is widely believed that overgrazing is the most important cause of rangeland
degradation (Fourie & Fouche 1985; Fourie et al. 1985; Danckwerts & Tainton 1996;
David et al. 2000). When the production potential of rangeland is over-estimated the
29
resultant overgrazing will cause a decrease of the palatable perennial plants in favour
of the less palatable, undesirable vegetation. Such changes will also influence the
hydrological status (Snyman & Fouche 1991, 1993), stability (Snyman 1998), quality
(Potgieter 1991), productivity (Kirkman & Moore 1995a; O'Connor & Bredenkamp
1997; Snyman 1997), and utilization potential of the rangeland. This can have major
economic implications (Van der Westhuizen 1994; Snyman 1998; Van der
Westhuizen et al. 1999) and therefore to ensure sustainable utilization and
production of the grassland ecosystem, the determination of rangeland condition and
trend is essential.
Smith (1895), Griffith (1903) and Wooten (1908) were some of the first authors to
recognize the need for rangeland assessments as an input for management decisions
(Foran 1976). In the early 1900's, Clements (1916) proposed that vegetation was
dynamic and is constantly changing. This theory contrasted to the then-prevalent
view that vegetation was static and unchanging (McIntosh 1985). Clements (1916)
successional theory rested upon the assumption that vegetation could be classified
into "formations" representing a group of plant species acting as an organic entity.
This theory of vegetation dynamics has been referred to as the mono-climax theory.
While Clements (1916) theorized that climate alone controlled the climax vegetation,
other scientists questioned the narrowness of this mono-climax description of a stable
community. Clements (1916) identified livestock grazing as a factor initiating plant
invasion and the processes of plant succession. In the late 1930's Stoddart used the
term "vegetative condition class" to describe successional stage (Spence 1938).
Attempts to officially define methods to determine rangeland condition were made
by the Interagency Range Survey Committee in 1937.
Dyksterhuis (1949) described a quantative approach for assessing rangeland
condition. The standard for comparison was developed by establishing 3 classes of
plant species (decreasers, increasers and invaders) based on their response to grazing.
Relative coverage of plant species within these 3 classes was estimated in the field
and compared to standard cover values for that location. He broadened the idea of
climax beyond that of Clements (1916) to include edaphic or physiographic
30
climaxes. Publications after 1949 and into the late 1960' s explored the relationship
between rangeland condition and insects (Nerney 1958), runoff (Leithead 1959),
plant vigour (Goebel & Cook 1960), livestock production (Cook et al. 1962), grazing
management (Valentine 1967) and wildlife management (Berg 1966).
The first International Rangeland congress, which was held in 1978, paved the way
for enhanced communication between international scientists and presented a general
critique of the rangeland condition concept. Since then, critiques have attacked 3
assumptions that were outlined by Smith (1989) as (1) climax is pristine vegetation
in equilibrium with climate and soil; (2) climax is the only vegetation furnishing
adequate soil protection and, therefore, is the most productive species composition
for a given site, and (3) retrogression due to grazing stress and succession after
removal of that stress are viewed as opposing and reversible linear responses.
Scientists presented anomalies in very arid areas with highly erratic precipitation
(Westoby 1980), areas dominated by exotic species (Westoby 1980), annual
grasslands (Dyksterhuis 1949; Laycock 1991), forested lands (Dyksterhuis 1949;
Hall 1978), areas having a long history of use where "climax" vegetation is not
known (Westoby et al. 1989), and areas which do not revert to climax when
management is removed (Slayter 1975). In addition, scientists faulted the rangeland
condition concept for its total dependency on botanical composition (Wilson &
Tupper 1982). Foran et al. (1986) described the ecological theory underlying the
rangeland condition concept as inadequate. Since then, non-equilibrium community
dynamics, alternative steady states and transition thresholds have been offered as
components in alternative theoretical models for rangeland vegetation dynamics
(Archer & Smeins 1991; Friedel 1991; Laycock 1991).
2.6.1.2. Criteria used in rangeland condition assessment
Wilson & Tupper (1982) extensively reviewed and categorised the factors involved in
the concept of rangeland condition into three. Accordingly, changes in vegetation such
as the botanical composition of the herbage, its total quantity or cover and its invasion
or emergence from a dominance by inedible shrubs or trees are considered as primary
31
attributes, whereas the secondary changes may occur to soil attributes, such as
infiltration rate, nutrient content or soil stability. Changes that may occur in production
characteristics of the land, such as animal production, water yield, wildlife habitat and
amenity value are taken as the last stage. The common factor in these is that in all cases
a change in condition refers to a change in status, relative to the potential of each
parameter within a particular land class. Therefore, a number of factors have been used
in singly or combination and the most common once are discussed below.
2.6.1.2.1. Botanical composition
Rangeland condition is measured in terms of vegetation because a change in vegetation
is most easily measured and it is the primary factor which leads to a change in other
attributes such as erosion or reduced secondary productivity (Wilson & Tupper 1982).
Furthermore, grass species composition is important as species may vary significantly
in their acceptability to grazing .herbivores, not only due to differences in palatability
but also due to phenological differences. These are all determinants of habitat, which
is especially important in the case of multi-species ecosystems (Smit 1994).
2.6.1.2.2. Basal cover
Basal cover is the proportion of the soil surface covered by vegetation and it is
mostly used for herbaceous vegetation. This might be near to 0 % in deserts or in an
unplanted cultivated field and greater than 12% in dense grassland (Baars et al.
1997). Limitations to the use of basal cover were indicated by Wilson & Tupper
(1982), as it is a relative measure used to determine mainly the density and
composition of perennial grass species. Mentis (1983) further added that, it is
difficult to measure basal cover accurately, precisely and cheaply and it is not an input
in common models of soil losses or rangeland production. Despite these limitations,
basal cover is still used in determining rangeland condition even m recent years
(Shackleton et al. 1994; Baars et al. 1997; Amsalu 2000; Ayana & Baars 2000;
Dahlberg 2000).
32
2.6.1.2.3. Plant vigour
Grass vigour can be defined as the potential or ability of a grass plant to regrow during
the season following defoliation. This measure has direct bearing on management, as it
gives an indication of management effects on the production of grazing during the
following year and it also serves as a short-term measure of the effect of grazing on the
health of the individual species (Kirkman 1995). In general, many researchers have
observed the influence of grazing on plant vigour and plant vigour as one important
parameter (Johnson et al. 1992; O'Reagain & Turner 1992; Morris & Tainton 1993;
Kirkman 1995; Kirkman & Moore 1995a,b,c; Peddie et al. 1995; Lutge et al. 1996;
Mckenzie 1996; Muir & Abrao 1999).
2.6.1.2.41. Biornass
From a grazing viewpoint, production or yield is one of the most important measures
in assessing rangeland. Biomass or standing crop usually refers to the weight of
organisms present at one time (Pieper 1978). Most estimates of plant biomass or
standing crop include only that above the soil surfaces. This material is commonly
available to large herbivores. Below ground biomass is very important for plant
functions, but is difficult to measure and generally not included in inventory or
monitoring procedures (Jerry et al. 1989). Direct harvesting is considered the most
reliable method of determining aboveground biomass. Furthermore, the use of
equations and models are also developed (e.g., Deshmukh 1984; De Leeuwet al
1991; Snyman & Fouche 1993).
2.6.1.2.5. Bansh encroachment
The encroachment of inedible shrubs and trees into semi-arid rangelands represents a
community change that may be viewed either as a change in the botanical
composition or as a separate vegetation attribute; the herbaceous -woody species
balance. Such indices of shrub encroachment may be an incomplete measure of the
situation, particularly when the shrubs are not fully established and represent an
incipient rather than a current problem. Separate measure of shrub seedling density
33
may be required and those would be considered as an assessment of the trend in
condition, rather than condition itself (Wilson & Tupper 1982).
2.6.1.2.6. Soil erosion
Soil erosion is a natural process but the quantity and the rate of surface runoff and
sediment yield may be altered through land use and management practices (Gifford
& Hawkins 1978; Blackburn et al. 1982; Thurow et al. 1986; Weltz & Wood 1986;
Snyman 1999a). Physical erosion or more frequently, a decline in the physical
characteristics of the soil surface, is the most serious manifestation of a decline in
rangeland condition because of its long lasting and progressive impact on production
attributes (Wilson & Tupper 1982). Ellison et al. (1951) noted that condition should
be based primarily on soil stability and secondarily on forage value. He indicated that
condition is always unsatisfactory unless the soil is stable and that the forage value is
only considered when the stability is assured.
2.6.1.2.7. Soil compaction
Compaction is the most complex soil feature having significant inter-relationships
with most of the recognized physical, chemical and biological properties of soils as
well as with environmental factors such as climate (Mckibben 1971). Compaction of
the soil leads to physical soil degradation and modification of the pore structure.
Examination of the soil matrix reveals a reduction in size and number of macropores
and a change of shape and continuity of pores. Associated with these changes is
increased soil strength and bulk density and a reduction of conductivity, permeability
and diffusivity of water and air through the soil pore system that all have an
important bearing on plant growth (Soane et al. 1981).
2.6.1.2.8. Soil parameters
Research on rangelands has historically focused on the effects of vanous
management practices on forage production and animal response. Little attention has
. .
been given to the impact of grazing on the nutrient dynamics of the soils. Recent
34
interest in soil health or soil quality has directed attention to grazing impacts on soil
parameters that make up soil health/quality (Du Preez & Snyman 1993; Manley et al.
1995;Lavadoetal.1996;Demeretal. 1997).
2.6.1.3. Techniques used in rangeland condition assessment and their limitations
An extensive review on the assessment of rangeland condition and trend was made by
many researchers (Wilson & Tupper 1982; Lauenroth & Laylock 1989; Hardy & Hurt
1989; Friedel 1991; Hurt & Bosch 1991). They indicated that it remained a source of
controversy despite its use for a long period. The limitations refer both to the concept of
rangeland condition and its assessment. Moreover, there is confusion in determining
which factors are of particular importance for meaningful rangeland condition
assessment and how these factors should be collected or determined whilst the
interpretation of the results often varies from one researcher to another, depending on
the person's original objective.
In general, the data collected and the production simulation models (stochastic) are
exceptionally limited, localized and in most cases derived somewhat subjectively
(Mentis 1983; Bosch & Booysen 1992). The extrapolation potential of the techniques
is, therefore, limited and in most cases a particular technique can only be of real
value in the area where it was developed. These anomalies indicate that there is a
mismatch between the advance in the understanding of ecological processes and the
developments in the theory and practise of rangeland assessment. Procedures
therefore differ considerably in their objectivity. The method to be used in rangeland
condition assessment should be simple, inexpensive, efficient and adopted over a wide
area within a minimum of time. Moreover, the method should be objective (Wilson &
Tupper 1982; Mentis 1983; Heard et al. 1986).
There are a number of steps involved in the development of any method of rangeland
condition. These include the choice of vegetation measure, the species to be included
and their classification as well as the type of index to be constructed. At each step, the
35
choice made from the available methods and systems is dependant on the current land
use and the attribute of condition to be assessed.
In general, there are two basic approaches to rangeland condition rating, I.e.,
ecological and the other is based on production. Both approaches depend on
assessment in relation to the potential or capability of the ecological site and on the
amount and composition of the vegetation (Smith 1988). Moreover, the techniques
have a common factor, namely the creation of a basis from which grazing capacity
can be calculated (Heard et al. 1986; Hardy & Hurt 1989) and attempt to provide a
measure of monitoring condition trends with time (Heard et al. 1986). The available
techniques are reviewed in more detail, as well as their shortcomings.
2.6.1.3.1. Subjective
The method developed by Roberts (1970) used plant density, species composition,
vigour, surface soil condition and insect or rodent damage on a 3-point scale to index
rangeland condition. Such ratings were combined to form a single numerical rating
for the site whereby rangeland could be described in one of the four condition classes
(very poor to good). The technique only tells the state of health of the rangeland and
it cannot be used as a basis to estimate grazing capacity. Later, Van Zyl (1986) used
a method based on a complete survey of the species composition and linked that with
rainfall. The same was done by Snyman (1993) for a semi-arid rangeland. Jordaan et
al. (1997) indicated that the subjective technique is most frequently used by both
extension staffs and farmers.
2.6.1.3.2. Quantative
In the quantative approach there are two general methods, i.e., agronorruc
(production) and ecological. In the agronomic approach, it is important to obtain
measures of rangeland composition, at least in terms of palatabity classes, which are
representative of the rangeland being grazed (Bames et al. 1984). This approach has
been widely used in the United States (Humphrey 1945, 1946, 1949, 1962) and South
Africa (Edwards & Coetsee 1971; Barnes et al. 1984). Species are allocated palatability
36
ratings, which signify their forage production potential. Three palatability classes were
developed for the purposes of classifying grassland species: class I - highly palatable,
class II - intermediate and class ill - unpalatable (Barnes et al. 1985, 1987; Rethman &
Kotze 1986). Multipliers, i.e., weightings of 3, 2 and 1 are used for classes I, II and ill
respectively, to derive a palatability composition rating for each sample site. It is
calculated as the sum of the products of the relative abundance of each species and its
weightings, and is expressed as a percentage of the maximum palatability composition
(viz 300) to produce a scale ranging from 33.3 (all species in class ill) to 100 (all
species in class I). These palatability composition (PC) values are converted to
weighted palatability composition (WPC) values by means of the formula developed by
(Barnes et al. 1984), namely:
WPC= (pC-33.3) x 100766.7
Under the ecological group, there are many methods such as the benchmark,
ecological index method, the key species, weighted key species and the degradation
gradient techniques.
A benchmark is an example of vegetation that is considered to provide the highest
possible sustainable animal production for the rangeland type under consideration
(Danckwerts 1982). Benchmarks provide a baseline against which the benefits and
harmful effects of domestic grazing can be evaluated and a standard against which
the effects of man's intervention in the natural environment can be assessed. A
benchmark will not necessarily be the climax vegetation for an area and is often
some form of sub-climax. However, limitations to the benchmark method have been
suggested by Mentis (1983), Barnes et al. (1984) and Martens et al. (1996).
Benchmarks may be in atypically favoured locations, they may experience different
rainfall or have a different fire history, or they may simply not exist and must be
theoretically _ constructed. Differences due to variability within a site may be
inappropriately attributed to management. Multiple benchmarks have been proposed
(Wilson 1984; Bosch et al. 1987) to represent the variability within a rangeland site
or land unit. However, recent works indicated that the benchmark does not provide
37
an estimate of the botanical composition over relatively large areas. For successful
rangeland condition assessment, the technique must be simple and fast to execute,
ecologically meaningful and scientifically sound. The commonly used rangeland
benchmark technique does not fulfil these requirements (Martens et al. 1996).
Therefore, efforts continued for more refined techniques.
The ecological index method was developed by Vorster (1982) for the Karoo biome
and is based on the principle that rangeland, in a reasonably homogenous farming
area, is compared with a reference point (benchmark), which has the same
topography and occurs in the same area as the survey site. The method combines
cover with botanical composition to provide a single indicator of veld condition that
is quick and easy to measure. The plant species are divided into a decreaser group,
three increaser groups and an invader group, based on specialist knowledge. Relative
index values of 10, 7, 4,1 and 1 are allocated to the five groups, respectively. A
rangeland condition index is obtained by the sum of the respective groups. The latter
is obtained by multiplying the sum of the frequency of the species, within each
group, with the relative index value. The derived rangeland condition index is used to
calculate grazing capacity or to monitor rangeland condition trends. A major
shortcoming of this approach is the subjective allocation of species to categories and
the assumption that all species are equally sensitive to utilisation.
In an attempt to develop a more rapid and simplified technique for use in large scale
field surveys (Tainton 1988) and the apparent insensitivity of the benchmark and the
ecological index method (Hardy & Hurt 1989), a number of attempts have been
made to develop an assessment technique based on the relative abundance of only a
limited number of key species within any ecological region. Foran (1976) compared
indices developed from the abundance of only Themeda triandra with those derived
from the full proportional species composition of a site and found such indices to
provide a relatively precise indication of the condition of the community.
Heard et al. (1986) used an adaptation of the key species technique to deduce
rangeland condition indexes from weighted species occurrences. Hardy & Hurt
38
(1989) and Hurt & Hardy (1989) make use of an ordination technique to identify
those species, which react to grazing. Species weights are based on the positioning of
these key species on the grazing gradient. The sum of the key species and their
associated weights only are used to calculate the rangeland condition index. Since all
the key species react to this grazing impact, the final score is thus an exact indication
of the site's position on the grazing gradient (Hurt & Bosch 1991).
The lack of sensitivity in monitoring rangeland condition changes in the above
methods paved the way for the development of the degradation gradient technique
(DGT). The DGT was developed in the climatic climax grasslands using a multi-
variate procedure (Bosch 1989: Bosch et al. 1987; Bosch & Gauch 1991; Bosch &
Kellner 1991). Degradation gradients represent a suite of sample sites in
deteriorating ecological condition along a grazing gradient, and were constructed for
certain relatively homogenous grazing areas in the climatic climax grasslands. These
gradients were then used as. a basis for objective and quantative condition
assessments of new sites by incorporating the new sites into the old ordinations
(Bosch & Gauch 1991). 'Interpretation of the rangeland condition index is conducted
in conjunction with a description model derived from the ordination studies (Bosch
& Kellner 1991).
The use of ordination to objectively identify species responses to grazmg has
received considerable attention in recent years (Mentis 1983; Bosch 1989; Hardy &
Hurt 1989; Hurt & Hardy 1989; Bosch et al. 1992; Hurt et al. 1993; Martens et al.
1996). Work done by Jordaan et al. (1997) has indicated the advantage of the
degradation gradients over the other methods. Similarly, Van der Westhuizen et al.
(1999) used the degradation gradient technique in semi-arid sweet grassland of
Southern Africa. They reported that the advantage of the technique lies in its
simplicity and ease of use. Another advantage is the fact that only the most important
indicator and dominant species in the study area are used to determine rangeland
condition. Not only the advantages, but they also indicated the limitations. Though
the degradation gradient method is seen as objective and scientifically correct, the
calculation of the ordination on which this method is based is complicated and only
39
possible with a computer. Furthermore, Hurt & Bosch (1991) pointed out that
relatively large sets of data are required to construct the gradients.
In conclusion, in response to the need of developing objective methods of monitoring
changes in rangeland condition, many methods have been developed. The general
agreement moves towards the more scientific and objective method of assessing
change in rangeland condition, i.e., the degradation gradient technique. Although this
is true, there are many limitations to use this method in least developed countries like
Ethiopia particularly in large-scale rangeland assessment. Some of the limitations are
the lack of a large set of data in vegetation, differences in the production systems
(nomadic pastoralism vs privately owned commercial farms), the severe degradation
of the rangelands, shortage of technical capability both in terms of manpower and
advancements in range science. Therefore, the choice of the method to use should
depend on the local conditions and it is preferable to use different techniques for
different objectives. Moreover, for countries like Ethiopia it is preferable to use the
simpler techniques used in developed countries and later develop techniques that best
fit local condition.
2.6.2. Woody vegetation importance and their assessment techniques
2.6.2.1. The importance of woody vegetation in the drylands
The crucial role of woody plants as a source of feed for both domestic animals and
game, firewood, mulch and soil conservation and as well as vegetation improvement
in semi-arid regions has long been recognized (Scholes & Walker 1993; Breman &
Kessler 1995; Smit 2002) indicating that they have multiple roles in farming systems.
In many cases they provide long-term stability or successful conservation of the
production systems especially in an environment of extreme oscillations (Sankary &
Ranjhan 1989). Branches from spiny woody species like Acacia tortilis and A.
erubescens are used for the construction of fencing kraals where livestock can be
protected from predators (Smit 1999a). With the expansion of the tourism industry
the market for wooden carvings from the indigenous tree species also became
popular.
40
Trees play an important role in the diets of browsing ungulates living in arid zones
(Walker 1980) and are often browsed or causally lopped and fed (Skerman 1977; Le
Houerou 1980). They provided valuable forage to herbivorous animals probably
since the time of their domestication (Skerman 1977). Due to the highly irregular
nature of rainfall in the drylands and virtual disappearance of nutritious grasses
during the dry seasons, trees and shrubs are an essential part of the pastoral
environment (Le Houerou 1980).
In Africa, at least 75% of the trees and shrubs serve as browse plants and many of
these are leguminous. In northern Africa, browse forms 60-70% of rangeland
production and 40% of the total availability of animal's feeds in the region (Nott &
Savage 1985, Milton 1988). In India, browse is the principal feed for goats and meets
over 60-70% of their forage requirements, of which the leguminous types of browse
are especially important. In Mediterranean and arid environments shrubs act as
nutritional reserves during the summer drought (Azocar 1987; Azocar et al. 1991).
In the mid hills ofNapal (Thapa et al. 1997) more than 75% of the tree fodder that is
fed to livestock is used during the dry period. In these areas, browse trees and shrubs
often have a higher crude-protein and mineral content and some times higher dry
matter digestibility than associated grasses, particularly during the dry season. An
important attendant advantage is a lowered cost of feeding due to a reduced
dependence on purchased energy and protein supplements. In parts of Africa, animal
production is heavily dependent on the availability of browse legumes, while in
Australia, interest in fodder trees has focused mainly on native species, particularly
the Acacia species, but novel browse legumes such as Leucaena leucocephala have
also been introduced with some success.
Browse legumes are more important than simply being a source of feeds. In many
situations, they form complex interactions between plants, animals and crops, the
positive aspect of which help to balance a plant-animal ecosystem and from which
there is a sustainable source of feeds. The foliages of some are used as vegetables by
humans while the roots, bark or stem and leaves of others are used for medicinal
purposes (Kaitho 1997). Many pastoral and agropastoral peoples throughout the
41
drylands areas of Africa have traditionally used and managed woody plants to
produce fuel wood, fodder, building poles and other products for sale and domestic
use. Diress et al. (1998) in their study of the influence of land use on woody
vegetation in the semi-arid area of Abala district, North Afar identified 43 woody
species and indicated their use as fodder for livestock, construction, fuel wood, shade
and shelter, human food, fence, medicine, farm implements and household utensils.
A non-quantified preliminary study in villages adjacent to the Awash National park
of Ethiopia showed a similar pattern of usage (Sharon 2000).
2.6.2.2. Browse production
Compared with the research effort allocated to the study of grasslands in East Africa,
the browse component of savanna ecosystems has for long been neglected (Pellow
1983). It is only recently that biomass or productivity measurements have been
published and still these are mainly from arid-lands in Kenya (Oba & Post 1999;
Rosenschein et al. 1999). Recently, Oba & Post (1999) studied the browse
production of Acacia tortilis under browsed (goat) and unbrowsed conditions. They
reported that goat browsing did not result in different tree growth rates between the
browsed and unbrowsed treatments. However, browsed trees produced more twigs
than unbrowsed trees.
Even in semi-arid savannas, there are limited published results of browse production.
Even these are very limited in the number of tree species they have considered for
their study. Rutherford (1982) reported some quantative data of aboveground
biomass of Burkea africana- Ochana pulchra savanna. He estimated the mean aerial
biomass to be 16 273 kg ha", of which 236 kg ha-! comprised the current season's
twigs and 1 100 kg ha-! the leaves. Of the potential browse, only 3.8 % was estimated
to be within reach of impala, 5.1% within reach of kudu and 67% within reach of _-
giraffe. Above ground standing crop of Colophospermum mopane in the Klaserie
private natural reserve, at the peak biomass, has been estimated at 20 840 kg ha" of
wood and 801 kg ha-! of leaves (Scholes 1987). Kelly & Walker (1976) estimated
. the mean standing crop of Colophospermum mopane in Zimbabwe to be 19 940 kg
42
ha-I, of which 1 506 kg ha-I, comprised the current season's shoots. Walker (1980)
quoted seasonal production estimates of woody species leaves and twigs, from a
Colophospermum mopane arid shrub and tree savanna in Zimbabwe that ranged from
594 kg ha" to 2 121 kg ha-I and demonstrated production varies strongly owing both
to inherent site differences (mainly soil depth) and to management. Some other
reports of browse production are given in Table 2.3.
Dekker & Smit (1996) studied the browse production and leaf phenology of different
plant communities in a semi-arid savanna in South Africa. These estimates included
the totalleafDM, as well as estimates of the leafDM at browsing heights of 1.5,2.0
and 5.0 m. They reported that the total leaf DM in the different plant communities
ranged between 1 224 kg ha" and 2 672 kg ha". Colophospermum mopane
contributed substantially to the total leaf DM in all communities. Their results also
indicated variability in the total and season availability of leaves to vary between the
different vegetation types and can therefore be considered important factors that
influence the distribution of large herbivores.
Foliar phenology is important because of its relation to processes and factors such as
tree growth periodicity, flowering and fruiting, plant herbivore interactions and
ecosystem properties (Borchert 1980; Reich & Borehert 1982, 1984; Aide 1988;
Reich et al. 1991, 1992; Wright 1991). In tropical savannas, the temporal patterns in
growth and reproduction are linked to the rhythms of the various seasons. Numerous
woody species, both dominant and sub-dominant, are deciduous with leaf fall
occurring during the dry season. Deciduousness is a feature of trees within savannas
across the world, with pronounced seasonal reduction in canopy cover as the dry
season progress. There appears to be considerable variation in ecosystem-level
patterns of deciduousness in the world's savannas. African savannas are mainly
deciduous (Mensat & Cesar 1979; Chidumayo 1990), as are those of India (Shukla &
Ramakrishhnan 1982; Yadava 1990). In the Ilanoes of South America, evergreen
species predominate (Monasterio & Sarmiento 1976; Sarmiento et al. 1985),
although forest patches within the Ilanoes are dominated by semi-deciduous or fully
43
deciduous species (Medina 1982) as are the seasonally dry forests of Costa Rica
(Borchert 1994).
Table 2.3. Rate of browse production for a variety of savanna woodland types.
Woodland type Plant part Production Study Author Remark
(kg ha" location
annum")
Acacia xanthophloea New shoots 5000 Serengeti Pellow Production
riverine woodland and leaves National (1983) of browse
park below 5.75
m only
Acacia New shoots 1 750 Serengti Pellow Production
ridge-top regeneration and leaves National (1983) of browse
thicket park below 5.75
m only
Colophospermum Twigs and 1510 southeast Kelly Range 590-
mopane leaves Zimbabwe (1973) 2120 kg ha"
Annum "
Dense shrub Green 1490 southwest Kennan 0.2 ha
Colophospermum leaf Zimbabwe (1969) sample area
mopane and Grewia
species
Rutherford (1979a) stated that it is important to have a clear understanding of what is
meant by browse and available browse. He described browse as the sum total of that
material on the woody species that is potentially edible to a specific set of animals
and that browse is most commonly regarded as the current season's growth of both
leaves and twigs. Available browse on the other hand is usually a more restricted
quantity than browse and in most studies available browse is simply determined on
the basis of maximum height above ground to which a specified animal can utilize
browse. The availability of browse below a specified browse height may be reduced
by obstruction of browse material towards the centre of the plant by dense branch
entanglements (Rutherford 1979b), while leaf senescence of winter deciduous
species will lower available browse during certain periods (Styles 1993; Smit 1994).
Possibly more important than the amount of available browse at peak biomass is the
continuity of browse availability over the season. A state where leaves are retained
on the trees in younger phenological state followed by the early emergence of new
44
season's leaves, is likely to be of greater value to browsers than a state where trees
carry no leaves for some time with only dry leaves available on the ground.
Since woody species growth is terminal and greater at the top of the plant
(Kozlowwski & KeIler 1966), greater productivity occurs in the upper canopy than
the lower canopy. Therefore, if the upper canopy is above the browse limit
(approximately 1.5 m) of the browsing animal, the plant will grow away from the
animal leaving access to only the less productive lower canopy. In order to achieve
the highest productivity and keep control on the browse plants therefore,
management must aim at maintaining browse at or below the available height.
2.6.2.3. The importance of trees in ecosystem functioning and their role in
rangeland condition
There are contradictory views on whether trees at low density improve or degrade the
condition of the rangelands. In order to examine the influence of trees at low-density
different researchers used many criteria. Among them, the most commonly used are
plant species. composition, forage yield, soil physical and chemical characteristics
and forage quality. In this section, the influences on plant species composition,
forage yield, soil physical and chemical characteristics will be reviewed.
Commonly, the influence of trees kept at low densities is studied under light grazing
conditions. Therefore, published information on the effects of different grazing
gradients for trees kept at low densities is scarce. The only work that can be cited
along this line is that of Belsky et al. (1993). Belsky et al. (1993) studied the effects
of widely spaced trees of Acacia tortilis and Adansonia digitata on their under storey
environments in four savannas located along a gradient of increasing livestock
utilization (light, medium and heavy). They reported that a unique under storey flora
and higher herbaceous yield under tree crowns in the lightly and moderately grazed
sites than in the corresponding open grasslands. There was no effect of sub-habitat
on the herbaceous layer at the heavily grazed site which was dominated by
agricultural weeds whereas, there were differences in the sets of species found under
-----------------------------------------------------~
45
tree crowns and in open grasslands at the lightly and moderately grazed sites.
There is also a lack of consensus in the literature regarding the effects of tree
canopies on herbaceous vegetation composition. Studies conducted by Kennard &
Walker (1973) in Zimbabwe; Stuart- Hill et al. (1987) in South Africa; Belsky et al.
(1989) in Kenya; Veenendaal et al. (1993) in Botswana, Smit & Swart (1994) in
South Africa, and Asferachew et al. (1998) in Ethiopia revealed a grass species
composition turnover from under-storey to open areas. Guevara et al. (1992)
examined floristic composition and structure of vegetation in three tree specïes
(Ficus insipida, F. colubrinae and Nectandra ambigens) under the canopy, directly
under the canopy perimeter and beyond the canopy in the open pasture. In agreement
with the above researchers, they reported that mean species richness per quadrat was
significantly higher under the canopy than at the canopy perimeter and in the open
pasture sites.
Asferachew et a!. (1998) investigated the influence of the canopy features of three
indigenous woodland type species (Acacia tortilis, A. senegal and Balanites
aegyptica) under light grazing condition in the Awash National park of Ethiopia.
They reported that the highest under-canopy vegetation diversity was recorded for A.
tortilis, followed by A. senegal and Balanites aegyptica. In all studies mentioned
here, there was no difference among the tree species in terms of their effect on the
herbaceous layer composition.
Contrary to the above findings, De Ridder et a!. (1996) and Moyo & Campbell
(1998) reported that there was no effect of canopy on grass species. De Ridder et al.
(1996) did not find the distinct species suite but did find grass species abundance
differences for the under-storey and open area environments based on their dry
weight. This finding was further strengthened by Moyo & Campbell (1998) from
Zimbabwe. They reported that ordination techniques using grass density and biomass
as indices separated quadrats according to soil type but not grass species according to
under storey or open areas or according to tree species.
46
Similar to the herbaceous vegetation composition, published literature presents
contradictory results on the effects of tree canopies on the yield of the herbaceous
vegetation. Thus, there can be an increase or decrease or no effect of the canopies on
herbaceous production. Accordingly, Kennard & Walker (1973); Stuart-HilI et al.
(1987); Belsky et al. (1989); Frost & McDougald (1989) and Smit & Swart (1994)
reported higher grass yield in areas directly below tree crowns compared to their
corresponding open areas despite the fact that overhead tree crowns intercept and
reduce the amount of rain-water and solar radiation reaching the ground and
competition for water and nutrients between tree and grass roots in the under- storey
areas (Amundson et al. 1995).
The increase in the grass yield under tree canopies is attributed to the improvement
in soil physical properties such as water infiltration, water holding capacity, aeration,
bulk density (Elwell 1986) and exchange capacity of the surface soil due to the
improvement in soil nutrient status as a result of high volume of leaf fall under trees
(Kennard & Walker 1973; Campbell et al. 1994), droppings from birds and
mammals (Belsky et al. 1989) and the concentration of nutrients from areas beyond
the crowns by means of lateral tree roots (Tiedemann & Klemmedson 1973). Other
possible reasons include lower under storey solar radiation and soil and plant
temperatures (Tiedemann & Klemmedson 1973) resulting 10 reduced
evapotranspiration, which creates mesic under-storey patches which are colonized by
palatable perennial grass species that have a higher water use-efficiency (Amundson
et al. 1995). For example, Belsky et al. (1993) reported an increase in the herbaceous
vegetation by 95.3% and 64% due to a canopy effect when compared to the open and
root zones, respectively.
Contradicting the above results, Grunow (1980); Grossmann et al. (1980); Dye &
Spear (1982) observed a decrease in herbaceous layer productivity under tree
canopies when compared to nearby open grasslands. According to Kay & Leonard
(1980); Monk & Gabrielson (1985); Pieper (1990) and Burrows et al. (1990) forage
production is often reduced by trees that compete with the under storey herbaceous
species for water, nutrients and light. In contrast to most published reports, Moyo &
47
Campbell (1998) reported similarity in grass yield under trees and open areas in both
years of their experiment.
Smit (2002) in his extensive review of the importance of ecosystem dynamics in
managing the bush encroachment problem in southern Africa indicated that nutrients
such as nitrate, phosphorous, a series of anions and cations and various trace
elements are essential to the nutrition of plants (Bell 1982) and act as determinants of
the composition, structure and productivity of the vegetation. Even though the base
richness of the parent material is initially important in determining soil fertility,
biological activities are important in the creation and maintenance of localized areas
of soil fertility often on base-poor substrates (Scholes 1991). Trees may act as
biological agents, creating islands that differ from those in the open. The importance
of canopies in soil enrichment is well documented (Bosch & Van Wyk 1970;
Kennard & Walker 1973; Tiedemann & Klemmedson 1973; Kellman 1979;
Bernhard-Reversat 1982; Belsky et al. 1989; Young 1989; Smit 1994; Smit & Swart
1994; Hagos 2001).
Kellman (1979) compared open savannna with canopies under Brysonima crassifolia
and Pinus caribaea in terms of their effect on soil nutrient status. Soils under the
canopy of Brysonima had a higher calcium (Ca), magnesium (Mg), potassium (K)
and base saturation than the soils in the open savanna. There was no difference
between the two sub-habitats in terms of sodium (Na), available P04, organic carbon
and moisture content. Studies conducted by Bosch & Van Wyk (1970) found a
higher content of nitrogen (N), phosphorous (P), K, Mg and Ca in soil from under
Boscia albitrunca, Combretum apiculatum, Acacia torti/is and A. senegal in
comparison with soils from the open areas. Another study by Bernhard-Reversat
(1982) showed that organic carbon and N content of soils of savanna in North
Senegal were concentrated in the first few centimetres of soil and increased under
tree canopies (A. senegal and B. aegyptica). A study by Palmer et al. (1988) in the
Eastern Cape Province indicated that soils of the grasslands were found to be poorer
than those of the bush clumps in Ca, Mg, K, and Na. Soils of the grasslands had
lower conductivity and pH values than soils of the bush clumps. Bush 'clump soil
48
contained higher percentages of organic material with a higher moisture, Mg and Ca
content.
Belsky et al. (1989) reported that mineralizable N and microbial biomass were
significantly higher in soil from canopy zone than from the root and grassland zones,
whereas organic matter, P, K and Ca (but not Mg) declined in soil from the base of
the trees towards the open grasslands. Smit and Swart (1994) investigated the
influence of leguminous and non-leguminous woody plants on the herbaceous layer
and soil in three sub-habitats (between trees, under leguminous and non-leguminous
trees) under varying competition regimes. They reported that the soil both under
leguminous and non-leguminous trees was richer in nutrients (% total nitrogen, %
organic carbon, electrical resistance, pH (H20), Ca, K, Mg) than between tree
canopies, whereas there was no significant difference in clay (%), phosphorous and
sodium between the two sub-habitats. Moreover, they indicated that soil enrichment
differed between tree species that grew in the same environment. Similarly,
Asferachew et al. (1998) reported that surface soil organic carbon, total nitrogen,
available phosphorous, exchangeable potassium and electrical conductivity were
higher under tree canopies for all the study trees compared to outside canopy soils.
Mayo et al. (1998) examined whether differences in soil nutrient status and
micro climate exist between under storey and open areas and whether such
differences are affected by soil type and tree species (Terminalia sericea and Acacia
karoo). Their results when pooled across sites revealed a non-significant difference
in pH, P202, Mg, and Ca between open and tree crowns, while soils under crown had
a higher K and total N. On the other hand, when the data is examined based on the
soil types, there was no difference in pH, P20S, Ca, K and total N in the analysed
parameters on sandveld. On red and vertisols, soils under tree crown had a higher K
and total N (%), respectively than their corresponding open lands. Very recently,
Wezel et al. (2000) from Niger reported significantly higher concentrations between
38-51% for carbon, Nand Pand 22% on ECEC (exchangeable cation exchange
capacity) for K in soil under shrubs than nearby open areas. The pH showed only
slight, but significant differences.
49
There is a lack of consensus in published literatures on the influence of trees on pH.
While Bosch & Van Wyk (1970); Kennard & Walker (1973); Palmer et al. (1988)
and Young (1989) reported a higher pH under tree canopies, Belsky et al. (1989) and
Hagos (2001) recorded a lower pH at the base of Acacia trees than further from the
trunk. In another study by Falkengren-Grerup (1989) claimed that stem flow
appeared to have decreased pH and base saturation in the topsoil of Swedish forests.
However, based on the positive association between increases in exchangeable
cations and soil pH (higher base saturation) (Barnard & Felseher 1972; Kennard &
Walker 1973; Hatton & Smart 1984), a higher pH under canopies of savanna trees
conforms more logically with the higher content of exchangeable cations in this sub-
habitat.
2.6.2.4. Bush encroachment
Bush encroachment is the process of open grassland savanna being transformed into
thick bushes and shrubs (Barnes 1979; Smit 2002) and is commonly seen as an
indicator of rangeland degradation in semi-arid savannas (Skarpe 1991; Perkins &
Thomas 1993a,b; Coppock 1994). It is a well-known phenomenon in savanna areas
(Archer et al. 1988). It involves indigenous woody species occurring in their natural
environment and is a widespread phenomenon occurring in many parts of the world,
including North America (Hobbs & Mooney 1986; Archer et al. 1988; Archer 1989),
Australia (Walker & Gillison 1982), South America (Schofield & Bucher 1986),
India (Singh & Joshi 1979) and Africa (Van Vegten 1983; Belsky 1990; Perkins &
Thomas 1993 a, b; Coppock 1994; O'Connor 1995; Smit 2002).
There is some controversy concerrung the driving factor behind the shift in
vegetation in grasslands and savannas. However, it is commonly explained in the
context of climatic change, fire suppression, the elimination of mega herbivores and
overgrazing (Archer 1989; Skarpe 1990a,b, 1991, 1992; Teague & Smit 1992;
Scholes & Walker 1993; Smit 1994; Smit & Rethman 1998; O'Connor & Crow
1999; Moleele et al. 2002; Smit 2002). For example, in Southern Africa, Ethiopia
50
and Kenya, the shift is associated with anthropogenic activities, especially high cattle
densities in communal grazing areas (Coppock 1994; Ringrose et al. 1996).
However, Abel (1993) challenged this theory of high cattle density. In Ethiopia and
Kenya, another factor that has undoubtedly facilitated woody encroachment has been
the national policy, which has banned burning of grazing and agricultural lands since
the rnid-1970s (Coppock 1994).
According to Smit (1994) in a broader sense the increase in the density of woody
plants is a natural process in response to certain changes. While the process is
natural, the change in secondary determinants, resulting in an accelerated rate of
woody plant increase, is not natural. Moreover, the same researcher discussed that
tree thickening is primarily a function of two processes, i.e., by the increase in
biomass of already established plants (vegetative growth) and by increases in tree
density, mainly from newly established seedlings (reproduction). In general, an
increase in the woody plant density beyond a critical density limit IS normally
accompanied by a decrease in herbaceous production and undesirable shifts in
composition (Archer 1990) mainly due to severe competition for available soil
moisture with a resultant effect of reduction in herbaceous production and grazing
capacity (Dye & Spear 1982; Moore 1989; O'Connor 1991; Smit & Rethman
1998b).
Moore et al. (1985) reported reduced production of the herbaceous layer with
increasing tree abundance in the Kalahari Thornveld and Shrub bushveld of the
Molopo area. Treatment with aerial applications of an arborrocide (Tebuthiuron)
resulted in increased grass production by between 220 and 740% with a subsequent
death of the woody plant. A density of up to 200 tree equivalents (TE) did not show a
decrease in grass production, while grass production declined linearly with further
increases in tree density. A density of up to 2 000 TE ha" almost completely
suppressed-grass production. The range of decrease in grazing capacity (ha-I Large
Stock Unit) -1 fall between 8.7 ha LSu-I (1 200 kg grass DM ha") to 45.8 ha LSU-I
(230 kg grass DM ha -1) over the 200 to 2 000 TE ha" density gradient. A similar
report was made' by Richter (1991) and Richter et al. (2001) from other parts of the
51
Molopo area of the North Cape in South Africa. Richter et al. (2001) indicated that
bush encroachment (2 500 tree equivalents ha-I) decreased the potential grazing
capacity by as much as 331%, 149%, 58% in the Molopo Thornveld, the mixed
Vaalbos Thornveld and the Eastern Grass Bushveld of South Africa, respectively, in
comparison to sites with tree densities of less than 400 tree equivalents per ha. On
the other hand, Donaldson (1978) obtained small improvement in grazing capacity
(9.1 ha AUI to 7.3 ha AUI) due to clearing woody plants in mixed savanna
dominated by Combretum apiculatum and Acacia tortilis. The decline in herbaceous
biomass due to woody cover is offset somewhat by increased browse production, the
magnitude of which depends on species composition and degree of woody cover.
Smit (2002) on his extensive review of the problem indicated that removal of some
or all of the woody plants will normally result in an increase of grass production and
thus also the grazing capacity. However, the results of woody plant removal may
differ between vegetation types, with the outcome determined by both negative and
positive responses to tree removal. This is due to the fact that the physical
determinants, biological interactions and individual species properties are unique to
each spatial and temporal situation. In addition, past management practices have
added to the complexity by bringing about different kinds and degrees of
modification (Teague & Smit 1992).
In general, Smit (2002) concluded that any bush control programme (chemical,
mechanical or biological) should focus on tree thinning rather than clearing of all
woody plants. Moreover, in making decisions on the intensity of tree thinning, the
sizes of the trees, which should be removed and the species to be thinned, cognisance
should thus be taken of the balance between the need to reduce the competitive effect
of the trees on the herbaceous layer and the positive influences, which the trees may
have. The aridity of the area should also be considered and effective management of
bush encroachment should not be considered a once-off event, but rather a long-term
programme.
52
2.6.2.5. Common methods used in the determination of browse production
An area of constant debate in any rangeland management programme is the ability to
determine the capability of a given area to sustain browsers (Hughes et al. 1987;
Melville et al. 1999). Estimates of shrub production, however, have been laborious
and less reliable than those for herbaceous vegetation for three main reasons. They
require manual separation of current growth from past year's growth, difficulty
associated with shrub density measurements and the variability of plant form, both
inherent and as a result of other influences (Hughes et al. 1987). Accordingly, a
number of methods have been proposed for the estimation of browse production.
Hughes et al. (1987) reviewed some of the techniques used in shrub production and
suggested that plots, commonly used in herbaceous production, give variable
production results with shrubs and are labour intensive. The weight estimation
technique (Pechanec & Piekford 1937), though fast and relatively accurate, required
a subjective determination by the person sampling and results in mental fatigue after
several hours of use (Cabral & West 1986). Plant density methods may yield
differing results (Beasom & Haucke 1975), so methods must be used which are
appropriate for the distribution, size and character of the shrubs measured.
Researchers have examined various plant measurements to determine their
usefulness in estimating production.
The most widely employed method for estimating browse involves the use of
standard statistical least square regression analysis, based upon the correlation of
some easily measured plant parameter with biomass, determined by destructive
harvesting. Rutherford (1979b) demonstrates thirty-five such equations, the simplest
of which use regressions based on plant height (Kelly & Walker 1976), stem
diameter (Barnes et al. 1976; Dayton 1978), a combination of plant height and stem
diameter (Rutherford 1979b) and canopy volume (Kelly & Walker 1976). In an effort
to improve the technique, Teague et al. (1981) introduced a tree equivalent for
compensating the difference in tree heights. Accordingly, he defined a tree
equivalent as a tree or shrub with l.5 m height. Smit (1989a) discussed the
53
limitations of TE, as the method does not compensate for structural differences
between tree species. Furthermore, TE values increase arithmetically with an
increase in tree height, while tree volume increases exponentially.
To estimate the browsing value of the savanna, Teague et al. (1981) introduced the
use of the browse unit (BV). It represents the total browseable length of palatable
trees and shrubs within l.5 m of the soil surface (the browsing height of goats) and is
derived by adding together those portions of all palatable trees and shrubs within the
l.5 m stratum (Tainton 1999). This is an example of quantifying non mass browse
data in which browse is not expressed in mass per unit area. Rutherford (1979b) cited
several more such techniques as number of twigs per unit ground area (Bookhout
1965; Halls et al. 1970; Knierim et al. 1971); leaves counted in permanent plots
(Crouch 1968); leaf counts per twig (MacOnochie & Lange 1970); elongation of
twigs (Halls & Alcaniz 1972) and several others.
Taking into account the ecological implications of trees in savanna areas, Smit
(1989a) indicated three important aspects from an agro-ecological point of view:
competition with herbaceous vegetation for moisture, food for browsers and creation
of sub-habitats suitable for desirable grass species. Accordingly, Smit (1989a)
developed three quantative descriptive units, which formed the basis of the Biomass
Estimation from Canopy Volume (BECVOL) model (Smit 1989a,b). The three
descriptive units are the evapotranspiration tree equivalent (ETTE), browse tree
equivalent (BTE) and canopied sub-habitat index (CS!). The BECVOL uses a
regression analysis approach using standard statistical least square regression
analyses. Various equations relating tree dimensions to leaf mass have been
presented, either for complete woody plants (Mason & Hutchings 1967; Barnes et al.
1976; Rutherford 1979a) or woody plant proportions (Barnes et al. 1976). Smit
(1994) indicated the advantage of this technique over the others is the ability to
estimate both partial and whole plants. The calculation of the ETTE and BTE is
based on the relations between the spatial volume of a tree and its true leaf dry mass
and true leaf volume, respectively. The following measurements of trees are used for
the calculation of the spatial canopy volume, i.e., tree height (A), height of maximum
54
canopy diameter (B), height of first leaves or potential leaf bearing stems (C),
maximum canopy diameter (D) and the base diameter (E) of the foliage at the height
of the first leaves (Figure 2.2) .
-....cCl)
E F
C'I
Cl)
(I)
.Nc...
Cl)
E G A
C'I
Cl)
(I)
B
Figure 2.2. Schematic illustration of an ideal tree, its measurements and structure
(Smit 1989a).
The modern methods all use the canopy volume of the woody species as the basis for
the estimation of available browse. Studies by Orban (1995) gave a clear indication
that canopy volume calculations lead to reliable estimates of available dry material.
Accordingly, Melville et al. (1999) compared the BECVOL with ARBORSTRQU
(Arbour structure method) developed by Caldwell (1998), which requires only three
measurements for each tree. They found that while the BECVOL method was more
comprehensive package than the ARBORSTRQU, the ARBORSTRQU was more
55
users friendly than the BECVOL. Therefore, the choice of which method to use
should depend on the person's objective.
2.7. THE CONCEPT OF RANGELAND GRAZING CAPACITY IN
PASTORAL PR.ODUCTION SYSTEM
Grazing capacity is defined as the maximum possible stocking of herbivores that
rangeland can support on a sustainable basis (FAO 1988). Its estimation is based
commonly on the assumption that livestock require dry matter intake equivalent to
2.5% to 3.0% of their body weight day". De Leeuw & Thotill (1990) reviewed two
major approaches to determine grazing capacity, i.e., plant or animal oriented.
However, both approaches pose problems, which are associated with characteristics
of the African environment and the production systems found in the various
ecological zones. These problems relate to scale (the area of assessment), species
mix, mobility, land tenure and the production goals of the actual producers.
Bartels et al. (1993) critically examined the concept of grazing capacity as used in
nomadic pastoralism by reviewing different literatures. They indicated that grazing
capacity is one of those terms in rangeland management that has the appearance of
objectivity but in practice means different things to different people, depending on
whether their primary interest is livestock production, wildlife management
(Caughley 1976; Macnab 1985; Dhondt 1989) or resource conservation. Even with
those involved in livestock production it may mean one thing to a commercial
rancher in the developed world and something quite different to a subsistence
pastoralist in sub-Saharan Africa. lts use also differs among people depending on
their objectives. In addition, inconsistencies are also observed in the methods
employed to estimate grazing capacity. Not only this, Scoones (1989a); De Leeuw &
TothilI (1990) discussed the difficulty of determining available forage per animal
owing to high annual and spatial variability in plant production, seasonal changes in
forage quantity and quality, livestock species mix, and the use of supplemental feeds
in Sub-Saharan Africa. The use of supplemental feeds indicates that rangelands are
.not necessarily the limiting resource for livestock production, except perhaps during
56
a severe drought. Breman et al. (1983) pointed out that species composition and
forage quality (FAO 1988) aspects are often neglected in calculations of carrying
capacity; they present an example where these omissions resulted in a gross over
estimation. Pratt & Gwynne (1977) and Purnell (1983) stressed the importance of
water point distribution in determining grazing capacity in East Africa.
The scale of the grazing capacity assessment has an important bearing on its validity
and usefulness. If direct interrelationships between animal output and feed supplies
are aimed at, resources within the grazing orbit of the individual producer need to be
determined. This is more easily achieved for individual mixed farmers than for
nomadic or transhumant pastoralists. For these latter groups, communal use of
grazing resources implies that only aggregate values of grazing capacity are
meaningful, the size of the area being dependent on the boundaries of common use
(De Leeuw & Thotill 1990). Consequently, a calculation of grazing capacity has to
include different land units sometimes many kilometres apart (De Leeuw & Thotill
1990).
Application of the grazing capacity concept on North American rangeIands assumes
that one decision-maker or a cohesive decision-making body has sole control of the
grazing resource. In systems where land tenure is communal or open access,
numerous households, each making more or less independent management decisions,
are using the same grazing resource. Another issue, which brings into question the
effectiveness of the grazing capacity concept in pastoral systems, is the investment
role of livestock. In Western ranching systems, livestock generally offer a low return
on capital (Workman 1986) and any excess capital created through livestock
production flows into more lucrative investments. In most of rural Africa, on the
other hand, a livestock enterprise can offer the greatest return on investment. For this
reason, capital flows into livestock rather than out of it. In the developed world, the
concept of grazing capacity has been applied mostly to cattle production enterprises.
Even if more than one species of livestock are kept, they are often allocated separate
areas of grazing land. In the pastoral production system, multi-species enterprises are
the rule. Consequently, several livestock enterprises are combined within the same
57
management unit. As feed preferences between species differ, assessing the fraction
of edible feed becomes more complicated.
58
CHAPTER3
STUDY AREA
3.1. LOCATION AND AREA COVERAGE
The study districts, Awash-Fantale and Kereyu-Fantale are located in the Southern
part of the north-eastern rift valley of Ethiopia, in the middle Awash valley that has
taken its name from one of the biggest rivers (Awash) in Ethiopia. The two districts
are located adjacent to each other (Figure 3.1) and some of the main areas in the text
are shown in Figure 3.2. Most of the area falls within an altitude of 800 to 1 100
m.a.s.l. However, there are higher peaks such as the Fantale mountain reaching up to
2 007 m.a.s.l. The study districts are found at a distance of 200 to 250 km from the
capital city of Ethiopia, Addis Ababa and occur at the border between two Regional
National States, i.e., the Afar and the Oromiya Regional National States.
ERITREA
Study districts
1. Awash-Fantale
2. Kereyu-Fantale
• Addis Ababa
KENYA
Figure 3.1. Location of the study districts in Ethiopia.
0\
tri
~~
., .,.
SHOP ,. WEST,,." ... ,...... , . HARAGE..
I
.....,I
--;-,::...;,' '.': 'Nura·.::; ... ............ _---',' ':(Hlra ><.. , ...r . \,, ,,I
I
,.. I •' '....- .... la'( ,~,,' Bu/ga ,, '. N
.,) \, , !"J, , o alO/no"". .s
~~ ,', ,I
-- .... - ... _-- I I
c ~" I,
= Awash National Park Boundary
Figure 3.2. Main areas in the study districts.
60
The area under study is situated along the main transport system in the country, i.e.,
the highway from Addis Ababa to Assab port and the railway line connecting
Ethiopia and Djibouti passes through the study area. Moreover, the area is also
crossed by one of the biggest rivers called Awash, which starts from mount Werqa
about 150 km west of the capital.
The total land area ofKereyu-Fantale district is 1 169.85 km', which is about 8.59 %
of the East Shoa zone of Oromiya (eSA 2000), while the area occupied by Awash-
Fantale district of Afar is 1 080.0 krrr', which is about 6.43 % of zone III in the Afar
Regional National State (Wolde Gebriel 2001). In general the study districts, cover
part of the upper and the middle Awash Valley.
3.2. eLmA TE
One of the environmental factors that leads to uncertainty in pastoral and agro-
pastoral envirorunent is climatic variability. As a result, it is difficult or impossible to
predict the levels of production that the system might produce from one year to the
next, or how ecosystems structure may change over time. Therefore, it is essential to
look into the nature of the climatic pattern to relate it to its influence on rangeland
dynamics.
lOOO
'ê' 800
'-' 600
~
,il 400
0:::
200
0
Years
Figure 3.3. Annual rainfall (mm) by year in the study districts (1966-2000).
61
The study districts fall in a semi-arid zone, a zone that receives an annual rainfall
between 400 to 700mm. The summary of the mean monthly meteorological data
(rainfall and temperature) and the annual rainfall obtained from Methara sugar
factory is presented in Figures 3.3 and 3.4, respectively (Mathara Estate Sugar
2001). The nature of rainfall in these areas is weakly bimodal. The main rainy season
is from July to September, while there is also a shorter rainy period from February to
April. According to the meteorological data obtained from Methara Estate Sugar
(1966 to 2001), the annual rainfall ranged from as low as 323 (1984) to as high as
863 (1982) mm, with a mean annual rainfall of550.9 mm (Figure 3.3).
Furthermore, the 35-year record indicates that the rainfall was above the average for
14 years (4l.2%), while in the rest of the years it was below the average (58.8%).
About 46% (253.3 mm) of the rainfall falls in July and August (main rainy season)
(Figure 3.4). Very little rainfall is received between November to January (5.27%).
Mean monthly rainfall ranges from 5.1 mm in December to 132.2 mm in August.
The amount of rainfall received during the short rainy period (Febrruary to April) is
about 120.3 mm (2l.8%). Moreover, the same data source indicated that the lowest
mean minimum temperature was in December (12.9°C), while the highest in June
(21.2°C).
__ Rainfall --- Minimium temperature
-Ir- Maximium temperature
140.- __ ~================================~~40
120 35
100 30
E 25 ~80 o5 !3
~ 20 ~
.c; 60 c"..
IX 15 §!-
40 10
2.0 5
Months
Figure 3.4. Summary of mean monthly rainfall (mm), minimium and maximium
temperatures (OC) in the study area.
62
The mean nurumum and maximum temperatures are 17.4°C and 32.7°C,
respectively. The lowest mean maximum temperature is 30.1°C (December) and the
highest 35.9°C (September). The number of rainy days per month is highest in
August (15.3 days) and the mean maximum and minimum relative humidity are
89.6% and 33.4%, respectively. The mean daily pan evaporation varies between 6.3
mm (December) and 7.7 mm (May). Data from two sources, namely Awash 7 kilo
and Methara national meteorological stations were taken to look at the nature of the
rainfall during the study period (Figure 3.5). The mean annual rainfall for Awash 7
kilo was 471 mm, which was 85.7% of the long-term average, whereas, that of the
Methara national meteorological station was 421 mm (77% of the long term
average).
250.-----------------------------------~
200 Awash 7 kilo
.§_ ISO
~ 100
~ 50
Figure 3.5. Mean monthly rainfall for the year 2001 for Awash 7 kilo and Mathara
national meteorological stations.
63
According to Ayele (1986), both the Afars and the Oromo pastoralists divide the year
into 5 major seasons depending on the level of precipitation (Table 3.1).
Table 3.1. Classification of seasons according the level of precipitation by the
different pastoral groups (Ayele 1986).
Afar naming Oromo naming Months Description
Karima Genna July to September Main rainy season
Hagai Hageya May to June Hot spill period
Dedaa Furmata January A period of light
to February showers
Sugum Abrasa March to April Short rainy season
Gilaal Deda mora October Cool season
to December
3.3. WATER RESOURCES
The Awash, Bulga Rivers, Lake Beseka and the hot spring at Filwoha dominate
water resources in the study areas. The Awash River is very important to the local
communities and the country at large. It is the main source of hydroelectric power to
the country and is intensively utilized by several large-scale irrigation schemes along
the basin (Halerow 1989; Lane et al. 1993).
Another river of importance is the Kesem river on the northern border of the park. lts
watershed is limited and large sections of the river regularly dry up during the dry
season. Contrary to what one expects, despite the proximity of the Awash and Kesem
rivers, the pastoralists have been denied access for long periods because of the
establishment of the park and the irrigation schemes (Lane et al. 1993). In villages
adjacent to Methara sugar plantation the water is considerably polluted and is
thought to be the main reason for human diseases in the area (Lane et al. 1993;
Sharon 2000).
64
Boreholes and ponds are the other water sources in the study area. Currently, there
are many boreholes and ponds established by government and non-government
organizations. Temporary water sources like ponds and rock water catchments serve
as possible water source in the rainy season and then dry up rapidly in the dry period
due to high evaporation rates.
3.4. GEOLOGY AND SOILS
The study area falls in one of the most geologically active regions of the world. It is
in the zone of interphase between two tectonic mega structures, the African and the
Somalian plates, which are pulling apart and away from the Arabian plate. These
activities played a major role in the formation of both the structure and hydrology of
the study areas (Schoelder & Jacobs 1993).
The soils of the study area are classified into three types (FAO 1965) according to
the parent material from which they are derived. They are the ancient alluvial and
colluvials, recent alluvial and volcanic material. In the ancient alluvial and colluvials
soils there are histosols and solonchaks. The regosols and andosols, which fall within
the volcanic material are derived from basalt gravel of colluvial origin and are the
result of the eruptive nature of the parent material. They are found at the base of
Fantale mountain and throughout the Mathara area. The recent alluvial soils have
developed from recent alluvial deposits and are termed fluvosols. They are found
along the banks of the Awash River, which are cultivated as irrigated farms and
deposited on recent very slightly, or non-calcareous deposits and saline soils near the
hot spring and the Mathara area.
3.5. FAUNA AND PALENTOLOGICAL VALUE
The Awash area is naturally endowed with diverse terrain, which gives opportunity
for wildlife viewing, walks and hikes, swimming (hot springs) and museum visits.
The East Arican rift valley is best known for the abundance of fossil fauna and flora.
Ethiopia is no exception to this since it was here that the oldest hominid remains
65
were discovered (Johannson & White 1979). This site is known as Hadar, located in
the Afar triangle just 220 km from the Awash National Park. Here the remains of
Australopithecus afarensis, more commonly known as "Lucy" and estimated to be
2.9 to 3.3 million years old, were discovered. The same site has yielded several
remains of Bovidae as well (Gentry 1981). Closer to Awash National park evidence
of hominid remains have also been discovered.
3.6. VEGETATION
Vegetation is one of the rangeland resources that are particularly important for the
survival of the different livestock species and wild animals scattered over the
rangelands. Both grazers (cattle, sheep and other wild herbivores) and browsers
(camel, goat and other wild herbivores) utilize the available rangeland resource
through continuous grazing without any form of rangeland improvement. Both
herbaceous and woody vegetation contribute to the forage required by the different
browsers and grazers. There is variability in the density and species composition of
both herbaceous and woody vegetation in the grazing land owing to the difference in .
biotic and abiotic factors of which human induced change being the most prominent.
The grass cover of the area consists predominately of Chrysopogon plumulosus,
while different Acacia and Grewia species dominate the woody vegetation. Studies
related to the nature of the vegetation in the area, are done primarily in relation to the
development activities of the different plantations and the Awash National park.
Accordingly, the vegetation type within or immediately surrounding the Awash
National Park has been classified by different authors FAO (1965); Robertson
(1970); Hillman (1993); Schloeder & Jacobs (1993). Vegetation types are generally
classified as grassland, shrubland/ bushland and riverine vegetation.
3.6.1. Grassland
The grassland areas are mainly found in Ilala Sala (found in the interior of the park),
Sabober, Arole, scattered grasslands at the base of Mount Fantale. The dominant
grass species is Chrysopogon plumulosus, while other grass species like Hypharrenia
66
hirta, Cymbopogon commutatus, C. excavatus, Aristida adscensionsis, Heteropogon
contortus, Ischaemmum afrum, Lintonia nutans, Bothriochloa radicans and Themeda
triandra are also found.
3.6.2. Shrub/bushland
These are predominately found on shallower alluvial and colluvial soils. The
predominant shrub/bush species are Acacia mellifera, A. senegal, A. nubica,
Dichrostachys cinerea and Grewia species.
3.6.3. Riverine vegetation
The riverine vegetation is dominated by A. nilotica, A. seyal, A. tonilis and A.
mellifera along the Awash River.
3.7. PASTORAL GROUPS AND HUMAN POPULA 'nON
This study addressed three pastoral groups; namely, the Afars, Kereyu and the Ittu.
The latter two fall within the same ethnic group and live in the same district.
3.7.1. The Afar
The Afars are one of the largest pastoral groups in Ethiopia inhabiting the vast
rangelands of northeastern Ethiopia, most of the middle and almost the entire lower
Awash valley. They occupy the northern part of the Rift valley, which runs through
the eastern horn of Africa to form a geographical entity that is accurately known as
the" Afar triangle" .
Today, they have lost access to much of their land due to irrigation (Helem 1993;
Waktola & Micheal 1999). They are politically dispersed in Ethiopia, Eriteria and the
Republic of Djibouti. Afar is the name the people speak of themselves rather than
"Dankil" or "Adal" by which their traditional adversaries know them (Harbeson
67
1978). Over the last 50 years, the Mar clans have moved in a generally westerly
direction as the military strength of the Issa has increased in the east. The Argoba
tribe prevented movement further west. To the south, there has been a period of
fighting between the Mar and the Kereyu and Ittu, Argoba on the western border,
with territorial boundaries shifting back and forth (Tibabu 1997). Currently however,
the relationship between the Mars and the Argoba is peaceful. In the district
considered for this study, the Mars are sub-divided into two tribes, the Debene and
the Waima. The Debene are the larger and more widely dispersed tribe and they are
said to have been in the area for at least 200 years (Schloeder & Jacobs 1993). They
are mainly found around Filwoha hot spring, Sabure, Boliyata and Kebena villages.
The Waimas have moved to the area from the east some 30-50 years ago when the
Issa took over their territory. The drought in 1970' s and 1980's also contributed to
their expansion. This put the land of Debene under pressure, which has already been
shrinking due to the expansion of state farms and conservation areas. Conflicts
between the two clans forced the majority ofWaima to move eastwards to the Awash
town areas (Tibabu 1997). However, currently their relationship is peaceful and they
share grazing lands. The Mars are generally Muslims.
3.7.2. The Kereyu
The Kereyu and Ittu pastoralists belong to the same ethnic group and they speak the
same language. The Kereyu people are believed to have their origin in southern
Ethiopia with the Borana people (Wilding 1985). They are thought to have arrived in
the area around 200 years ago establishing themselves around Fantale mountain, lake
Beseka, Sabober plains and Metahara. The dominant inhabitants and the users of the
resources in Kereyu-Fantale district until early 1950 were the Kereyu pastoralists. In
the early 1950's, the Ittu from western Hararghe came around Fantale and north of
the Awash river because of the competition within their tribe as well as with the
Issa's. This, together with, the establishment of National park, state farmers and the
increase in the size of lake Beseka has substantially decreased their traditional
grazing land.
68
There are two sub-tribes within the Kereyu people namely, Baso and Dulcha. The
Baso are mainly found along the Awash river, close to the large scale irrigated farms
around Methara town. The Dulcha are settled far out. in the rangelands. Basos are
mostly Muslims, while Dulachas are mostly Wakefata (Mudris 1998). Traditional
rule of the social system known as the "Gada" system regulate cultural events and
economic and social decisions. The Gada system is an institution that represents an
extreme development of a type of social structure known as age set. The basis of the
social organization is a kinship system and the organization is referred to as lineage.
The social relations are based on genealogical reckoning and tracing ancestors
through the father's line of descent (Asmarom 1973).
3.7.3. The Ittu
The Ittu were pushed out of their traditional area by the expansion of the Issa tribe in
the early 1950's (Schloeder & Jacobs 1993). The number of immigrants has
increased and now the number of Ittu is thought to exceed that of the Kereyu
(Futterknecht 1995). During the establishment of the Awash national park, some Ittu
were resettled south of the Awash river. However, several fled back to the Fantale
area having lost a considerable amount of livestock during attack by the Issa tribe.
Immigration into the area steadily increased during Mengistu's policy of free and
equal access for all and continuing expansionist policies of the Issa and Howiya
Somalia clans. The Ittu have placed considerable pressure on the Kereyu pastoralists
and the Awash National park. Ittu are found living with Kereyu and sometimes on
their own. The majority of the Ittu are Muslims.
3.7.4. Human population
The human population of Kereyu-Fantale district is estimated to be 74 932, which is
about 3 % of the human population in the East Shoa zone of Oromiya. The density
per square km is 64.1. Compared with the density in East Shoa zone of Oromiya, it is
the lowest but it is high when compared with other districts in the rangelands. The
Awash-Fantale district has a human population of 21 442 which is about 12.42% in
69
zone 3 of the Afar Regional National State (eSA 2000).
3.8. RELATIONSHIPS AMONG THE TRIBES AND occurATION
The relationship between the Oromo groups (Ittu and Kereyu) is smooth and they
live in a harmony. Owing to the pressure in the shortage of land, occasionally there is
disagreement on the use of land. The Ittu are generally involved in crop production,
while the Kereyus exercise both systems (crop production and pastoralism). The
Oromo groups (Kereyu and Ittu) are also involved in charcoal and firewood making
as well as off-farm employement (Mudris 1998). The relation between the Oromos
and the Afars is based on hostility. There is no trust between them. The intensity of
the conflict is high when the dry season gets severe. It is at this time that both move
to the border to utilize the available land. Because of the fear among them the
grazing lands bordering the two districts have excellent cover and species
composition, which will be treated in depth in the rangeland evaluation chapter. The
Afars also exercise off-farm employment, charcoal and firewood making and some
cropping activities.
3.9. TERMlINOLOGY
The terminology used in this thesis is in accordance with Trollope et al. (1990),
unless referenced or described.
ANIMAL UNIT (AU): An animal with a mass of 450 kg and which gains 0.5 kg day"
I on forage with a digestible energy percentage of 55%.
AVAILABLE BROWSE: The browse material determined on the basis of maximum
height above the ground to which a specific animal can utilize it (Smit 1996).
BIOMASS: Total amount of living plant material present above and below ground in
a particular area at any given time - kg ha".
70
BROWSE: The sum of the total material of woody species that is potentially edible
to a specific set of animals, and browse is commonly regarded as the current season
growth of both leaves and twigs (Rutherford 1979a).
BROWSE UNIT (BU): Metabolic equivalent of a kudu (100% browser) with a mean
body mass of 140 kg (Dekker 1997).
BUSH ENCROACHMENT: The phenomenon of increasing tree and shrub density
in savanna (Smit et al. 1996).
CANOPY: Cover of leaves and branches formed by the tops or crowns of plants.
CARRYING CAPACITY: Potential of an area to support livestock through grazing
and/or browsing and/or fodder production over an extended number of years without
deterioration to the overall ecosystem (ha" AU).
DECIDUOUS: The seasonal senescens and shedding ofleaves (Smit 1999a).
DElllSCENT: The splitting open of ripe pods to release the seed (Smit 1999a).
GRAZER UNIT: (GZ): A metabolic equivalent of a blue wildebeest (100% grazer),
with a mean body mass of 180 kg (Dekker 1997).
GRAZINGIBROWSING CAPACITY: Productivity of the grazable/browseable
portion of a homogeneous unit of vegetation expressed as the area of land required to
maintain a single animal unit over an extended number of years without deterioration
to vegetation or soil- ha"l AU or AU ha".
PHENOLOGY: Study of time of appearance of characteristics periodic events in the
life cycles of organisms in nature and how these events are influenced by
environmental factors.
71
QUADRAT: A small clearly demarcated plot or sample area of a known size on
which ecological observations and measurements are made.
SHRUB: A perennial woody plant with two or more stems arising from or near
ground level (Smit 1999a).
TRANSECT: An imaginary or real line along which measurements or surveys of
ecological observations are made (Vorster 2000).
A TROPICAL LIVESTOCK UNIT: is the equivalent of one bovine animal of 250 kg
live weight (ILCA 1990).
72
ClHIAl?'f1ElR. 4
PASTOlRALISTS PERCEPTIONS OF lRANGELAND RESOURCE
UTKLIZATION AS RELATED TO LIVESTOCK PRODUCTION
4.1. llNTRODUCTION
Research studies about indigenous rangeland management around the world show
that indigenous people consider resource preservation in their management strategies
(Fortman 1989; Bayer & Grell 1994, 1995). The indigenous pastoral management
system is the basis of the nomads' way of life (Wang & Zheng 1998). However, as
seen in Ethiopia and in many nomadic pastoral areas, rangeland-based life styles and
industries are in difficulty and the rangeland environment is under threat because of
extreme climatic fluctuations, animal diseases, over estimation of the grazing
capacity, land-use changes and the demand from an increasingly important cash-
based economy (Roderick et al. 1998; Scholes 1991; Beruk & Tafesse 2000). In this
segment of Ethiopia, rangeland has further decreased due to volcanic activity
(Figure 4.1). Owing to the wide fluctuations in climatic conditions, overgrazing and
other interacting influences, the household's use of rangeland resources can vary
between years and seasons within a year. A study of the existing rangeland resource
base, their utilization and possible constraints as related to livestock production is an
important first step as this will pave the way for any type of rangeland intervention to
be undertaken in the study area. Therefore, the objectives of the study were to assess
the perceptions of the pastoralists about the existing rangeland resource base,
utilization, constraints and possible solutions. It was also intended to compare the
two pastoral groups in terms of their perceptions about rangeland resources and
related affairs. Finally the perceptions of the pastoralists about the rangelands were
compared to the findings of the scientific way of rangeland evaluation.
73
Figure 4.1. Loss of range land due to volcanic eruption in Kereyu-Fantale district
and the plants grown on the erupted volcanic soil are not grazed by
animals.
4.1. PROCEDURE
4.2.1. Survey design
Secondary information relevant to the study was collected from previous studies,
organizations and other sources. Both informal and formal surveys were carried out.
The informal survey was aimed to confirm or complement the information obtained
from secondary data sources and to get insights from producers and community
members who are directly or indirectly involved in the system. This was undertaken
through interviewing individuals, groups, key informants and personal observations.
The information gathered through the above processes was summarized and used as
a basis to design structured questionnaire to quantify the most important parts of the
74
study and hence to have an overall understanding about the rangeland resource base
of the pastoralists taking households as a unit of analysis. A structured questionnaire
was prepared to elicit information on rangeland resources, perceptions of vegetation
composition, water resources, feed resources, migration, rangeland condition,
livestock population and production constraints. The questionnaire was used to
gather both quantitative and qualitative information at each of the sampled
households. In the questionnaire prepared, there were single and multiple response
questions. Single response questions are those questions where the sampled
household has a single reply and multiple response questions are questions where the
individual household can give more than one answer to a given question. In case of
the latter, the percentage of responses (respondents) will be greater than 100%.
Examples of multiple response questions are marked in Appendix 4.1. To administer
the interview, enumerators were recruited and training was given in collaboration
with the relevant near by organizations. Before conducting the actual survey, the
structured questionnaire was pre-tested by interviewing some households from both
pastoral groups. Thereafter, the necessary adjustments and corrections were made on
the prepared questionnaire. The actual data was collected by both the student (20%
of the households from both pastoral groups) and the trained enumerators under the
close supervision of the student. In the Kereyu-Fantale district (Oromo pastoral
group), there are 20 Kebeles (lowest administrative units) of which two are in the
town. Of the rest, nine were selected randomly after stratifying them into different
production systems. From the identified villages, 90 households were chosen
randomly and interviewed independently.
In the Awash-Fantale district (Mar pastoral group), there are six Kebeles of which,
four were selected randomly for the study. Accordingly, 55 households from the
identified Kebeles were selected randomly and interviewed independently.
4.2. 2. Data Analysis
The data collected was summarised in SPSS (Statistical Package for the Social
Sciences) and analysed using the same programme (SPSS 1996). Descriptive
75
statistics such as means, percentages, range and standard deviations were calculated
to present the results.
4.3. RESULTS
4.3.1. Demograpnies of the pastoralists surveyed and main source of income
The study households had an average total family size of 6.74, with those in Awash-
Fantale district being about 24.6% larger than those in Kereyu-Fantale district (Table
4.1). The respondents in both districts had a similar average age (40.00 ± 11.2 years),
which varied between 22-75 years. All the respondents were males and the majority
(72.6%) were married to a single wife, while the remaining to two (25.3%) and three
(2.1%) wives. The households belong to an Afar and Oromo ethnic groups, of which
the Afars were of the Debene and Waima tribes. The Oromos were of the lttu,
Kereyu-Dulcha and Kereyu-Baso tribes. Of the respondents, 3.3% were from the
Somale ethnic group. The majority (79.3%) of the respondents did not have any kind
of education (Table 4.1).
The main source of income in both pastoral groups was from the sale of animals
(Figure 4.2). The interesting difference between the two pastoral groups lies in the
contribution of the sale of milk and milk by- products to their source of income. To
the Afars, the sale of milk and milk by- product is the second most importance
source of income, while to Oromos it stands last.
76
Table 4.1. Profile of respondents by family size (mean ± SD), ethnic group, tribe and
educational background (Respondents: Oromo = 90 and Afar = 55).
Category Kereyu-Fantale (Oromo) Awash-Fantale (Afar)
Family size (number)
Male 3.26 ± l.8 3.87 ± l.8
Female 2.92 ± l.9 3.96 ± l.8
Total 6.17±3.1 7.69 ± 2.8
Ethnic group (%)
Afar 100.0
Oromo 96.7
Somale 3.3
Tribe of the household
(%)
Ittu 47.8
Kereyu (Dulcha) 28.9
Kereyu (Baso) 20.0
Debene (Afar) 83.0
Waima (Afar) 17.0
Somalia 3.3
Educational
background (%)
Formal 9.9 5.6
Read and write Il.O 14.8
None 79.1 79.6
70
60
.(.J.l~. BIl Oromo
0
'"~0 50 DAfar
0
~40
~
'-
0
~ 30
i':":
0e 20
(I)
0...
10
o
Sale of livestock Sale of milk and milk Sale of crops Employment
by-products
Source of income
Figure 4.2. The main sources of income in the study districts (Respondents: Ororno=
89 and Afar = 53).
77
4.3.2. Livestock ownership in the study districts
With few exceptions, livestock in the traditional pastoral and agropastoral societies
are owned by individuals (perrier 1995) and in nearly all traditional pastoral and
agro-pastoral systems livestock are herded. Accordingly, the mean number of
livestock owned by an Oromo household was lower than that of an Afar household
(Table 4.2). For all animal species, females dominate the population. The majority of
the respondents in Kereyu-Fantale (92.9%) and Awash-Fantale (88.4%) do keep a
mixture of livestock species.
Table 4.2. Mean number of livestock species owned per household in the
study districts (Respondents: Oromo = 90; Afar = 55).
Animal category Kereyu-Fantale (Oromo) Awash- Fantale
(Afar)
Cattle
Cows 4 ± 3.0 10 ± 6.9
Heifers 2 ± 1.6 4 ±3.4
Calves 2 ± 1.7 4 ± 3.6
Bulls and oxen 2 ± 1.7 2 ± 1.3
Camels
Male 1 ± 0.7 3 ± 2.3
Female 4 ± 3.6 12 ± 7.2
Sheep
Male 2±1.4 2 ± 1.3
Female 6 ±4.8 10 ± 5.4
Goat
Male 2 ± 1.6 5 ± 3.5
Female 9± 7.0 21 ±11
78
4.3.3. Pastoralists perceptions to rangeland resources and utilization
4.3.3.1. Rangeland plants
4.3.3.1.1. Poisonous plants
The importance of poisonous plants to the livestock industry of Africa cannot be
over-estimated because about 600 indigenous poisonous plants are known to occur in
Southern Africa alone. In Africa, where livestock is traditionally kept under
extensive conditions on rangeland that is frequently denuded by droughts,
overstocking and uncontrolled fires force the animals to eat poisonous plants, which
they would normally avoid (Naude et al. 1996). Accordingly, in the study districts,
90% of the Oromo respondents and 98% of the Afar respondents replied that, they
knew poisonous plants affecting their livestock production, indicating their
accumulated wealth of plant knowledge over generations. The results indicated that
both herbaceous and woody plants were identified as poisonous to animals (Tables
4.3 and 4.4). The ranking order of these poisonous plants, based upon their
importance as poisonous plants to livestock by the Oromo pastoralists was
Cryptostegia grandiflora (Haqonqol), Tribulis terrestris (Kumeto) and Capparis
fascicularis (Harangama) in decreasing order. Interestingly, the grass species
Cynodon (Aredo) was also identified as a poisonous plant. The Afar pastoralists
ranked the poisonous plants C. grandiflora, Erythrina abyssinica (Dunbeya),
Flueggea virosa (Raberba) and Tephrosia subtriflora (Aro) in decreasing order
(Figure 41.3.).
In the district occupied by the Afars, the poisonous plants are abundant in marshy,
swampy areas and along the riverbanks, while the difference between the two
pastoral groups in the ranking order was mainly related to the availability of a
specific plant in a particular area.
79
Table 4.3. Percentage of respondents indicating common plants as poisonous in the
study districts (Respondents: Oromo = 82 and Afar = 54).
Botanical Local names Percentage of respondents
name
Awash-Fantale Kereyu-Fantale Awash-Fantale Kereyu-Fantale
(Afar) (Oromo) (Afar) (Oromo)
Cryptostegia Halimero Haqonqol 83.3 76.8
grandiflora
Cynodon Aredo - 43.9
dactylon and
Cynodon
plectostachys
Tribulis Kumeto - 74.4
terrestris
Capparis Harangama - 61.0
fascicularis
Erythrina Denubeya 51.9
abyssinica
Flueggea virosa Rebreba 24.1
Tephrosia Aro 13.0
subtriflora
Table 41.4.Poisonous plants affecting the different livestock species as ranked by the
two pastoral groups in Kereyu-Fantale and Awash-Fantale districts (1 =
animal species highly affected by the specific poisonous plant; 4 = animal
species least affected by the specific poisonous plant; Respondents:
Oromo = 82 and Afar = 54).
Type Region and livestock species Region and livestock
of poisonous species
plant
Kereya-Fantale Oromo) Awash-Fantale (Afar)
Cattle Sheep Goat Camel Cattle Sheep Gilat Camel Season
of incidence
Cryptostegia 3 4 2 1 3 4 2 1 Mainly dry
grandiflora season
Cynodon 1 2 2 Wet season
dactylon and
Cynodon
plectostachys
Tribulis 2 1 3 4 Wet season
Terrestris
Erythrina 4 3 2 1 4 3 2 1 Mainly dry
abyssinica season
Flueggea 4 3 2 1 Mainly dry
virosa season
Tephrosia 2 1 1 3 Mainly dry
subtriflora - season
Capparis 3 4 2 1 Mainly dry
fascicularis season
80
Figure 4.3. Example of Cryptostegia grandiflora as a poisonous plant.
According the pastoralists, the different plant species act as poisons to the different
livestock species at different growth stages and different plant parts also act as
poisons to the animals. For example, Cynodon species cause poisoning and can kill
cattle and sheep when they are grazed at the early stage of growth. Unfortunately, the
pastoralists were unable to give a clear reason as to why this poisoning happens at
the early stage. Tribulis terrestris and T subtriflora are leguminous plants that cause
bloating and kill animals particularly when eaten at the early flowering stage.
Cryptostegia grandiflora is the most important climbing type of poisonous plant,
which kills animals when the leaves are eaten and this plant upon consumption,
based upon the information obtained from the pastoralists, will distend the animal
with a large production of saliva. The seeds of Erythrina abyssinica, upon
consumption cause continuous diarrhoea and paralysis leading to the gradual death of
the animal.
According to the information obtained from the pastoralists, Flueggea virosa is a
woody plant that has an effect only on camels in such a way that it paralyses the
81
camel progressing to a gradual death of the animal. However, the Afar pastoralists, in
periods of critical feed shortage, will chop the leaves with the branches and give it to
cattle, sheep and goats. Capparis fascicularis will kill camels and goats, initially by
paralysing the animal and eventually causing the death of the animal. The animals
consume the poisonous plants accidentally when grazing in the field and the
incidence of an animal being poisoned is more common on the side of the Afars than
the Oromos.
Poisoning from the woody plants mainly occur during the dry season, while that
from herbaceous plants during the wet season. Both pastoral groups have developed
traditional methods of treating animals poisoned with different plants (Table 41.5).
Wide ranges of local practices are applied ranging from the provision of water and
milk, to the preparation of soup. They prepare soup from goat's meat indicating their
clear commitment to the welfare of their animals. They are very caring to their
animals and at times, they even carry the calves of the camels to take them for
treatment (Figure 4.41).
With regard to the abundance of poisonous plants, in comparison to the past, 31.0%
of the Oromos replied that there is an increase, while 21.1% reported a decrease,
39.4% said that it is the same as before and 8.5% replied that they do not know. On
the other hand, 81.6% of the Afars replied that there is an increase in the abundance
of poisonous plants, 14.3% indicated that it is the same as before and the remaining
argued that they do not know. Many of the Afar pastoralists could not explain the
reason why poisonous plants have increased, but suggest that the repeated diversions
in the direction of the Bulga river can be one possibility, which brings different
plants from the highlands.
82
'fable 4.5. Percentage of respondents practising different traditional methods of
treating animals poisoned with different poisonous plants (Respondents:
Oromo = 82 and Afar = 54).
Treatmen Cryptostegia Cynodon Tribulis Capparis Erythrina
t grandiflora dactylon and
Cyllodon terrestris
Fascicularis abyssinica
plectostachys
Orom Afar Oromo Afar Oromo Afar Oromo Afar Oromo Afar
0
Provision 24.4 4.0 4.9 30.3
of local
H 'beer (Tella)
Provision 19.5 16.0 18.2
of local
:iiquor
(Areke)
Provision 75.6 35.7 56.0 31.7 69.7 29.0 37.5
of milk
Provision 4.9 32.1 16.0 17.1 6.1 19.4 18.8
of butter
Provision 14.3
oflemon
Provision 3.6
of tobacco
Washing 3.6
the body
of the
animals
Provision 7.3 12.0 31.7 3.0
of
kerosene
Provision 2.4 3.6 20.0 39.0 3.0
of pepper
Provision 26.8 21.4 32.0 14.6 30.3 43.8
of water 't
Provision 17.1 35.7 7.3 27.3
of goat
meat soup
No 7.3 3.6 12.0 14.6 3.0 18.8
control
method
83
Figure 4.4. An Afar carrying his young camel on his back for treatment at a clinic.
4.3.3.1.2. Grasses
Grasses can mean different things to different people, depending on the objective or
attitude towards the plant. Nevertheless, to the pastoralists, different grass species are
the source of feed to their livestock, in addition to their usage for other purposes. In
the study districts, both pastoral group use grasses of different species for livestock
feeding and house roofing, although the type of grass species used for different
activities differ. In the case of the Oromos, 97% and 3% of the respondents use
Cymbopogon commutatus (Senblete) and Chrysopogon plumulosus (Daremo) for
house roofing respectively, while Cenchrus ci/iaris is also used for house roofing
where there are no other alternatives. On the other hand, 38.1%, 23.0%, 10.6% and
28.3 % of the Afar respondents use C. commutatus (Isisu), C. excavatus (melefe),
Enteropgon species (Koref) and Sporobolus ioclados (Hamilto), respectively for
house roofing. Although both pastoral groups use Cymbopogon species for house
roofing, the difference between the two pastoral groups in the usage of other, grass
species for house roofing, particularly with reference to the Afars, lies on the
abundance of the specific grass species within their vicinity.
84
Seventy six percent of the Oromos and 77% of the Afar do not harvest grasses, but
24 % of the Oromos and 23 % of the Afars do harvest grasses from the rangelands,
while in both pastoral groups, harvesting of grasses is undertaken by women and
transported using donkeys or on the back of the women.
The two pastoral groups differed in terms of the main reason for not harvesting
grasses from the rangelands. The Oromos attributed it mainly to the type of
ownership of the land, the Afars on the other hand, ascribed it mainly to the lack of
experience (Figure 4.5). In the communal system, land is owned communally, where
decisions regarding usage of the rangeland resources are made collectively. Thus, the
grasses are considered as a common property. A small proportion of the households
indicated that because of the high number of livestock there is no enough grass to be
harvested.
35
30 29 o It is a common
'Jl
"E 25 properity<U"~0
0 IIILack of experience~ 20
<U
...........
0 15 Ill] Shortage of grasses....
<U
..D
El 10
::l oNumber of livestock is
Z 5 high
0
Oromo Afar
Figure 4.5. Number of respondents giving various reasons for not harvesting grasses
from the rangelands (Respondents: Oromo= 53 and Afar = 31).
All pastoralists indicated that compared to the past, there is a decrease In the
abundance of grasses and legumes. The main reasons given by the respondents for
the decrease in the abundance of grasses and legumes were a decline in the amount
of rainfall and drought, population pressure, increase in livestock numbers,
85
expansion of farming, pests, pressure from weeds, bush encroachment and change in
the channel (direction) of the Bulga river in the case of the Afar.
4.3.3.1.3. Woody plants
Woody plants play a central role in the farming systems of the study districts, where
both pastoral groups use woody plants for a wide range of uses, the primary being as
a source of livestock feed (Table 4.6). Eighty six percent of the Oromo respondents
and 94.4% of the Afar pastoralists use browse as a source of feed for their livestock.
The importance of woody plants is not only limited to the ones listed in Table 4.6,
but their usage extends as a source of medicine and cosmetics. For example, Oromo
women use the bark of Grewia tenax to soften their hair by soaking the bark in water
overnight. The water will have thick oily characteristics. They will roll their hair
either with hand or using sticks by pouring the liquid material. Through their
traditional practices, the pastoralists have developed good skills in the usage of
woody plants as a source of medicine to treat different kinds of diseases. Some of the
woody plants used as a source of medication are: the fruit of Balanites aegyptica and
the roots of Euphorbia species for treating wounds, the fruit of Acacia nilotica for
treating wounds during circumcision of boys, the leaves of Cadaba rotundfolia for
the treatment of influenza and the branches of Salvadora persica for teeth brushing
and the bud or the tip of the leaves used for treating influenza. The list of usage
presented here is not considered comphrensive for the fact that only some individuals
or families within the community know the knowledge of woody plants as a source
of medicine for both humans and animals.
Traditionally the Afars do not cut big trees, as a part of their indigenous knowledge
of natural resource management. If there is a need to do so, the community members
will be informed about the issue and then trees will be cut thereafter. This good
traditional practice is, however, diminishing, either because of poor control in natural
resource conservation, clearing of land for agricultural activities and charcoal
making.
86
Table 4.6. The uses of woody plants in the study districts as ranked by the
pastoralists (1= highest use and 6= least use; Respondents: Oromo = 89
and Afar = 51).
Tree uses Kereyu-Fantale (Oromo) Awash-Fantale (Afar)
House construction 3 3
Firewood/ charcoal 4 4
Browsing 1 1
Fencing 2 2
Wild fruits 5 5
Shade 6 6
Implement preparation 6 6
4.3.3.1.4. Bush encroachment
Bush encroachment is commonly considered a major limiting factor to livestock
production and natural resource conservation because it alters habitat structure and
decreases herbaceous production (0'Connor & Crow 1999). This phenomenon is in
fact common in semi-arid savannas and grasslands of the world (Blackburn &
Tueller 1970; Barnes 1979; Bucher 1987; Harrington et al. 1984; Archer 1989;
Matheson & Ringrose 1994). Accordingly, the majority (78.7%) of the Oromo
respondents and 100% of the Afar pastoralists replied that compared to the past, their
grazing land are more covered with bush and shrubs, which was responsible for a
decline in rangeland condition. In Kereyu-Fantale, there are two scenarios taking
place. Firstly, big trees like A. tortilis are being cleared for firewood, charcoal
making and expansion of agricultural lands and secondly, shrubs and bushes are
increasing in many grazing lands that did not have any kind of woody plants before.
Both pastoral groups ranked a decrease in grass production as the first problem
associated with bush encroachment followed by difficulty in herding.
The pastoralists did not only mention the problems they faced with bush
. encroachment, but equally indicated the benefits of an increase in bushes and shrubs.
87
However, the response from many of the Mar pastoralists (75%) was on the negative
side, i.e., they do not appreciate the benefit obtained from the bushes and shrubs,
while the reply from the side of the Oromo pastoralists was 50:50. The benefits
derived from those who appreciated the benefical side of an increase in bushes were:
feed for livestock, fencing, house construction, firewood, erosion control and shade.
Both pastoral groups indicated that overgrazing and drought are the main causes of
bush encroachment (Table 4.7).
Table 4.7. Reasons for the increase of bushes and shrubs ranked as percentage of the
respondents (1= highest reason; 6= lowest reason; the number in
parenthesis indicate % respondents; Respondents: Oromo = 80 and
Mar = 49).
Reasons Kereyu-Fantale Awash- Fantale
(Oromo) (Afar)
Overgrazing 1 (62.9) 2 (57.1)
Uncontrolled 3 (52.9) 3 (55.1)
livestock and
wildlife movement
Banning of burning 5 (11.5) 5 (22.4)
Drought 2 (54.3) 1 (65.3)
I do not know 4 (24.3) 4 (24.5)
It was argued by the elders that thirty years ago no two trees stood together. They
further elaborated on the reasons for such an increase of woody plants. According to
them, livestock graze everywhere, where forage is available and animals may act as a
dispersal mechanism. Birds were also indicated as possible agents for carrying seeds
from place to place. The other possible explanation given by the elders, specific to
the Mars was transportation of seeds by Floodwater. Big rivers like Awash, Bulga
and others transport different kinds of plant materials during the rainy season. These
seeds will easily germinate and grow quickly along the rivers and from there animals
carry the seeds upon ingestion and deposit it in their dung on grazing sites. The other
possible mechanism indicated by the pastoralists was that, when there is feed
shortage or drought in the adjacent districts, herders bring all their stocks into the
grazing lands occupied by the Mar and the Oromo pastoralists and as a result seeds
from different places might be transported this way.
88
Both pastoral groups indicated that Acacia nubica, A. senegal and A. mellifera are
the main encroaching woody plants in the area. Furthermore, other non-woody plants
were also indicated to increase such as Tribulis terrestris, Parthenium hysterophorus
and Datura species. So far, there are no measures taken by the pastoralists to control
bush encroachment, except discussing the problem among themselves.
4.3.3.1.5. Rangeland
Ten to 20 years' back, forty-nine households from the Oromo respondents and 46
from the Afar pastoralists used the rangelands mainly for grazing (Figure 4.6). The
traditional pastoral herd management system was sustainable and well suited to the
region's natural resource base. Livestock converted rangeland forage into milk, meat
and blood, thereby providing a source of human subsistence. However, with regards
to changes in land use all respondents in both pastoral groups replied that there was
an increase in the size of eropland with a subsequent decrease in the grazing land.
Pastoralists were of the opinion that a shortage of rainfall (drought) associated with
an increase in human and livestock population, were some of the reasons for the
decrease in the size of grazing land and an increase in the size of the cropland. In
addition, all respondents in the Awash-Fantale district, considered bush
encroachment and the repeated change in the direction of Bulga river to have an
immense contribution in decreasing the size of the grassland area. None of the Afar
and only 12% of the households in Kereyu- Fantale district had private grazing land
(0.64 ha on average), and they ascribed this to the shortage of land and to the
traditional communal type of ownership where land is collectively owned.
60
<Jl
"'i:: 49
<U 50 46
"Cl
I:: Ell Oromo
0 40
0.. DAfar
<Jl
e:! 30
to-.
0...
<U 20
"S
Z 100
Cropping Grazing Both
Land use types
Figure 4.6. Purpose of using rangeland by the different pastoral groups in the last
10-20 years (Respondents; Drama = 69; Afar = 47).
89
With regard to the type of land ownership almost all the Afar (47 respondents of the
total 55) and 62 respondents of the total 90 Oromo respondents, wanted in the future
to continue with the communal type of ownership (Figure 4.7). The rationale behind
this choice was reasonably clear: different resources pose different problems and
possibilities of exploitation, accordingly the possibility of increasing the size of the
land to be grazed by the animals of an individual household and the decrease in land
size if owned individually, were the main reasons wanting to continue with the
communal type of ownership (Table 4.8).
70
Cl)
i:: 60
Il.)
'".C:l: 50
0
0.
Cl)
~ 40
III Oromo
....... 30 lil Afar0....
-Ils.) 20
i 10 3 2
0 ..:':':'
Private Communal Both
Types of ownership
Figure 4.7. Type of ownership preferred by the pastoralists for managing their
rangelands in the future.
Table 4.8. Percentage of respondents by type of reasons for the choice of communal
type of ownership in the grazing lands (Respondents: Oromo = 62 and
Afar= 47)
Reasons offered Kereyu-Fantale (Oromo) Awash-Fantale
(Afar)
The size of land will 53.7 -
decline if privately
owned
Not possible to use 5.6 -
migration
Possible to graze as 37.0 52.7
desired
Strengthens relations 3.7 18.4
among Q_eople
For good protection of 10.5
animals
Tradition 18.4
90
The future use of the land, the livelihood and lifestyle of the family depends partly
on the type of ownership of the land. To this effect, the Oromo pastoralists,
preferring individual type of ownership, attached two reasons for their choice. On the
one hand it was intended to block the usage of the rangeland by immigrants, as this
was one of the major problems in the study area. It was also meant to block free
access to land for members of the community. This dilemma in the type of land
ownership led the Oromo pastoralists to undertake repeated discussions amongst
themselves ending with no clear direction. Sustainable grazing management or land
resource use is a key issue of concern in most arid and semi-arid regions of the
world.
The sustainability of arid and semi-arid rangelands requires constant adaptation to
change, not only utilising the opportunities, but also using resources at a sustainable
rate so that they remain available year after year. Pastoralists, in the study district
used to follow their practical knowledge of the landscapes and resource assessments
to determine grazing at landscape level. However, many complexities are involved in
assessing the sustainable use of rangeland in the study districts, particularly with
reference to the Oromo groups. Subsequently, sub-division of the grazing land into
different units for efficient utilisation of the rangeland and which also allow for
recovery thereafter did not exist in the study districts. The main reasons attributed
were the shortage of land for proper allocation, tradition, the lack of common
agreement, lack of knowledge, increased livestock number and the expansion of crop
farming.
The Afars traditionally subdivide their grazing land into what is locally known as
"Alta" and "Kalo". Alta is a mountainous, rocky area with a high potential for
livestock production and is mainly used in the wet season, whereas Kalo areas are
swampy, marshy and usually located along riverbanks and mainly used in the dry
season. However, this traditional practise of rangeland management is diminishing,
as one of the elders indicated that his brother has stayed continuously at Aha for 10
years. In addition, in the past, when grass was abundant, both pastoral groups used to
rest and burn their grazing land as measures of rangeland improvement. However,
91
due to the shortage of land and the lack of appropriate technology in rangeland
improvement practices, there is currently no rangeland improvement practice
applied.
4.3.3.1.6. Perceptions towards rangeland condition
Since vegetation IS the foundation for rangeland use, developing rangeland
management strategies and plans requires information about the vegetation ecology
and an understanding of the rangeland ecosystem processes. One possible way of
studying the condition of the rangelands is through interviewing the pastoralists who
are knowledgeable of their rangeland ecosystem. This in turn can be integrated with
the scientific way of rangeland evaluation for a better understanding of the rangeland
ecosystem and development of possible intervention mechanisms. Accordingly, 99%
of the Oromo pastoralists and 73% of the Afars indicated that the condition of their
rangeland, based on their own subjective judgement, is poor. Twenty six percent of
the Afars and 1% of the Oromo pastoralists rated the grazing land as moderate in
condition. Many interacting reasons, with profound implications for the future of the
rangelands and pastoral production systems, were given for this poor state of the
rangelands. The most important is overgrazing followed by drought (Table 4.9).
According to the Oromo pastoralists the other major problems were increase in
human population and immigrants coming from other districts creating pressure on
the rangelands. The Afars further identified bush encroachment and the repeated
change in the direction (channel) of the Bulga river as their major problems (Table
4.9).
Periodic assessment of the condition of the rangelands is part of the traditional
natural resource management practices on which the welfare of their animals is
based. This assessment, which is mainly based on the availability of grass, water,
suitability to the different livestock species and security to the herders, can be
unconditional or be carried out on an individual or group basis. In general, greetings
in these pastoral groups mean holding discussions about water, rainfall, rangeland
92
Table 4.9. Reasons contributing to the poor rangeland condition, as ranked by the
pastoralists (1= Most important reason; 11= Least important reason;
Respondents: Oromo= 87 and Afar= 51).
Constraints contributing to Kereyu- FantaIe (Oromo) Awash -FantaIe
poor Rangeland condition (Mar)
Drought (Shortage of rainfall) 2 2
Increase in human population 3 7
decreasing the amount of
grazing land
Overgrazing 1 1
Immigration 4 -
Bush encroachment 7 3
Little support from the 6 7
government in rangeland
management
Lack of knowledge 5 5
Poor soil status 8 8
Expansion of crop cultivation 9 -
Poisonous plants 10 6
Expansion of towns 11 9
Repeated change in the 4
direction of the Bulga river
condition and welfare of their animals. Given the above problems related to the
rangelands the study also assessed the pastoralist 's perceptions of the possible ways
on how to improve the condition of the rangelands. To this effect, 46% of the Afar
respondents did not suggest any solution (Table 4.10), while the rest of the
respondents suggested a wide range of solutions such as the adoption of new
agricultural technologies, stopping migration, resting of the grazing land, stock
reduction, sharing of the grazing lands and others.
It was also an objective of this study to identify the major grazing areas for rangeland
assessment. Accordingly, the results of the household survey indicated that the most
frequently grazed lands by the animals of the Oromo pastoralists were those near the
villages, Fantale, Harole, Bulga riverside, Haroresa, Kara, Jirmee and Fati and Ladi.
In the case of the Afars grazing lands near the villages, Fantale, Bulga riverside, near
the hot spring, Samayu meda and Degage are the most important grazing areas.
93
Table 4.10. Pastoralists suggestions on how to improve the rangelands as percentage
of the respondents (Respondents: Oromo= 70 and Afar= 43).
Suggested solution Kereyu- Fantale (Oromo) Awash-Fantale (Afar)
I do not know 45.5
Adoption of agricultural l.7 12.4
technologies
Stop migration 13.6
Availability of 5.10 12.1
alternative feed
resources
Resting of the grazing 13.6 15.2
lands
Reduce number of 6.8 9.1
livestock
Requires setting rules on 1l.9
the use of the rangelands
Share of the grazing land 20.3
Rotational grazing (Dry 27.1 3.0
and wet season)
Use of irrigation l.70 6.1
Stopping deforestation 3.0
4.3.3.2. Water resources
Tracking water of consumable quality for livestock is one of the major occupations
for pastoralists and one of the key determinants of pastoral movement and migration.
Accordingly, water in the study area is available from 6 possible sources, which
include boreholes, wells, rivers, rock water catchments, irrigation canals and ponds.
The perceptions and realities of water availability and quality (rated subjectively)
were quantified in the study districts. The results showed that the majority (85.2%) of
the sampled households in Afar replied that there was no problem related to the
quantity of water, while 65.4% indicated that the water was not clean. On the other
hand, for the Oromos, both quantity (65%) and quality (62.5%) of water is a major
problem.
The influence of water shortages both in terms of quality and quantity on the
pastoralists, particularly on the side of the Orornos has been significant in these
marginal environments. The distribution and type of water facilities can influence the
frequency with which animals are watered. In general, the further a producer lived
94
from the water source, the more likely he will be forced to practise alternative day
watering of his animals, which is quite common within the Oromo pastoral groups.
Accordingly, nearly 50% of the Oromo households water cattle, sheep and goats in
the dry season every other day and nearly all of them travel between 13-18 hours a
day to the watering place in a round trip. This can have a substantial demand on the
labour required to water the animals with loss of energy both to the herder as well as
the animals.
The critical months of water shortage as reported by both pastoral groups was from
November to June. In both pastoral groups, the alternative water sources when there
is critical shortage are buying water from the pumps (6.6-13.3 % of the respondents)
and to move to distant water sources (86.7-93.4% of the respondents). The most
important sources of water for animals in Awash-Fantale (Afar) in the dry season
were rivers and hot springs, while in Kereyu-Fantale (Oromo) they were boreholes,
different rivers and irrigation canals. In Kereyu-Fantale, during the wet season, the
main sources of water were ponds, surface water and irrigation canals, while in
Awash-Fantale it comprised of ponds and rivers (Figure 4.8). In Kereyu-Fantale, the
distant water sources include lake Beseka and different river sources, while' in
Awash-Fantale it consists mainly of rivers. All of the Ororno respondents and 71%
of the Afar pastoralists replied that the sources of water for humans and animals are
the same (Figure 4.9). The water collected on the surface is only available when
there is rain and it is also dirty (Figure 4.10), while the water from the irrigation
canals (Figure 4.11) carry wastes from the factories or farms, which held a health
hazard for humans and animals alike.
95
60 I- 100Dry season 0 Wet season I 90 lO Wet season • Dry season I
Cl>
ê..:, 50 Ororno <Il 80 r-r-
"0 ë Afar
c:: 40 -8 700 c::0
C.c.,, ~0.. 60l>
<!:: 30 ....... 50.0., 0 40
1>1) ~
Ol
ë.., 20 ë8
g.., .
30
....,
c, 20 r-r-
p... 10
h n I I 100 0 '-- L,.c la.. I11III
~~<.. ~~ f:>~
~,,<:i, ~o
~~ ~~
~o ;..o..~
~$
Types of \.\Ilter source
Figure 4.8. Percentage of Afar and Oromo respondents utilising different water
sources by season (Respondents Oromo= 88 and Afar= 49).
Figure 4.9. Livestock and people utilising the same sources of water at a borehole
(Oromo).
96
Figure 4.10. An Oromo woman fetching water from the surface of a shallow pond.
Figure 4.11. Calves watered from an irrigation canal (Mar).
97
According to the OpInIOn of the Oromo pastoralists, the main problem regarding
water resources is the inequity in their distribution. Due to such an inadequate
distribution of water sites at times both humans and livestock have suffered greatly
and some livestock even died. In some grazing lands there is abundant grazable
forage material but there is a shortage of water, while the opposite is also true. In the
case of the Afars there is also an imbalance in the distribution of watering points. In
both districts there are unutilised grazing lands in the dry season due to water
shortages and the pastoralists make use of such grazing lands by grazing the animals
for two days and watering them on the third day. Alternatively they use the grazing
land only during the wet season. Therefore, proper distribution of watering sites and
forage quantity is of paramount importance for the proper utilization of the rangeland
resources. Both pastoral groups appreciate the efforts that were undertaken by both
government and non-government organizations in improving the water supply.
However, further efforts must be made to improve the quantity and quality (to the
Oromos) and the quality of water to the Afars.
4.3. 3.2. Natural minerals
The information obtained from the group discussions with the respective elders and
the response of the sampled households indicated that there are 3 types of natural
minerals, which are utilised by animals. The Oromo groups locally call them as
Haya, Bole and Bojjii. The Afar generally know them as Haya and classify them as
black, red and white based on the colour of the soil from which the minerals are
derived.
Raya
Haya is a kind of soil mineral in dry, wet or dust form (powdered and as a wet
muddy lump). It is obtained from the ground by digging the soil on the areas where
this mineral is found. It has a grey or cement colour with a very fine-textured
materials. This mineral is suitable for all kinds of animals and it tastes salty or like
sour sugar.
98
Bole
Bole is a kind of soil mineral, which has a different texture from that of Haya, It
consists of a mixture of large gravel like materials. Bole is found scattered in spots
among other soils. Animals get the mineral by going to the spot and licking the spot
where the mineral is found. It is more salty than Haya and licked more by cattle and
goats than by sheep and camels.
Bojjii
Bojjii is a mineral salt having a dull whitish colour and is found as very fine salt
grains. It is mainly available near the banks of salty water and is baked on the ground
after the salty water has evaporated. According to the respondents and the elders it
has the highest salt concentration. It is well taken by cattle, while it is harmful for
sheep and goats. Sheep and goats take this mineral indirectly i.e., by eating the salt
tolerant grasses. Camels also consume this mineral in the diluted form, which is
locally known as Hora (Figure 4.12).
Figure 4.12. Saline water under preparation for camels (Oromo).
99
The respondent households and the elders indicated that if their animals do not get
the above-mentioned minerals in time their productive and reproductive performance
will be affected (decrease in milk production, do not increase in weight, will have
poor physical condition and they will be weak). Without these minerals, the animals
will be more susceptible to different kinds of diseases.
In both pastoral groups, Haya is the most frequently used natural mineral. Almost all
the respondent households indicated that the minerals are offered or consumed by the
animals mainly during the wet season and the main reasons for such usage were the
abundance of water and grasses during the wet season. These salty natural minerals
induce thirst in the animals and only during the wet season have the animals enough
water to drink after consuming the mineral source. During the dry season the animals
will obtain the elements from plant sources. With regard to the availability to
animals, 39.2% of the Oromo respondents indicated that the animals will lick these
mineral resources directly In the grazing fields, while 58.1% of the households
transport the mineral soil to home and feed them and 2.7% use both methods.
Similarly, 85.7% of the Mars practise feeding the animals directly in the field, while
14.70% feed their animals by bringing the mineral soil to their home.
4.3.4. Feed resources
The feed resources that are available in Kereyu-Fantale district are natural pasture,
woody plants, crop residues (residues of maize stover, teff straw, remaining of
tomato production), weeds and sugarcane tops. In some villages, they also harvest
grasses from the different plantations. The grazing lands are grazed continuously
throughout the year. Crop residues are fed mainly from October to January, while
sugar cane tops are fed during the dry season from November to June. Despite the
production oflarge amounts of sugar cane by-products in the vicinity, only 26.7% of
the Oromo respondents use sugarcane by-products (sugar cane tops) as a feed
resource. The molasses produced is mainly exported and the remainder is used by
enterprises or organizations outside of the study districts. The lack of transport,
financial shortage, accessibility and the lack of knowledge are the major problems in
not using sugar cane by-products by the pastoralists for livestock feed. Migration in
100
response to feed shortages is common throughout the year, but mainly from
November to June. The feed resources that are available in Awash-Fantale are
natural pasture and browse. The use of crop residue is limited to villages that have
started cropping very recently, like Sabure and Bolyita. Migration in response to feed
shortage is similar to that of the Oromo pastoralists.
All the sampled households in both pastoral groups reported that there is a critical
feed shortage during the dry season (November to June). The first measure taken by
both pastoral groups to solve feed shortage is migration. Both pastoral groups rarely
sell animals as the first measure to solve feed shortages (Figure 4.13). Buying feed
for coping feed shortage is also practised by those households close to the towns but
not by the pastoralists far out in the rangelands.
Fifty eight percent of the Oromos and 28 % of the Afar provide supplements to their
animals during the dry season: There is some variation between the two pastoral
groups as to which feed they supplement their animals (Table 4.11). This was due to
the variation in the availability of a specific feed in a given area. The Oromo
respondents involved in crop production make use of crop residues as a supplement
to their animals.
80
r:n
ïv::
70 o Oromo
'"s0:: 60
0 III Afar
0..
tr: 50
~
'0- 40
evo 30
.<. Il
vs
.:..:
g 20
v
0... 10
0
Sale of Buy feed Migration Use reserve Lopping
animals feed browse trees
Types of first measures taken to solve feed shortage
Figure 4.13. First measures taken by the different pastoral groups (percentage of
respondents) to solve feed shortages; Respondents: Oromo= 86 and
Afar=52).
101
Table 4.11. Type of supplement offered during the dry season by the different
pastoral groups (1= Most used; 5= Least used; Respondents: Oromo= 52
and Afar= 15).
Types of supplement Kereyu- Fantale Awash-Fantale
(Oromo) (Afar)
Sugarcane by-products 3
Crop residues 1
Browse 2 2
Grasses 4 1
Wheat milling by- 5
products
The main animal categories given the supplements are calves, small ruminants,
milking cows and oxen. The main reasons of offering the supplements according to
their purpose they are given is milk production, to alleviate weakness in calves, for
ploughing oxen and to strengthen sick animals.
In response to a question asked whether they conserve feed for the dry season or not,
the households in both pastoral groups had a 50:50% reply. In the case of the Afars,
they conserve standing hay and browse, while the Oromo groups conserve mainly
crop residues and browse (Table 4.12). The difference between the two pastoral
groups lies in the type of production system in which they are. In general, the Afars
are more pastoral than the Oromos.
102
Table 4.12. Type offeed conserved during the dry season in the study districts
ranked as percentage of the respondents (1= the highest; 4= Lowest;
Respondents: Oromo= 47 and Mars= 28).
Type of feed conserved Kereyu-Fantale (Oromo) Awash-Fantale (Afar)
Standing hay 1
Crop residues 1
Browse 2 2
Cut hay 3
Sugar cane residues 4
Both pastoral groups replied that there are browse species that provide year round
browsing. Some of these are Balanites aegyptica, Ziziphus mauritania, Acacia
tortilis, A. nilotica, Grewia bicolor, Salvadora persica and Capparis fascicularis.
Some of them are found along river or in marshy and swampy areas. On the other
hand, some browsed species are deciduous (Figures 4.14 and 4.15).
4.3.5. Migration
Migration (mobility) is an inherent strategy of the pastoralists to optimise production
of a heterogeneous landscape under a precarious climate.The fact that grazing
resources are found in different places at differe~t times affects the herder's strategy.
For example, pastoralists tend to prioritize mobility (whether in the form of
nomadism or seasonal transhumance) and an opportunistic approach to resource
management (Hussein et a11999;Ndikumana et al. 2001). These strategies ensured
the persistency of pastoralism over centuries, leading to perceptions of its benign
nature as practiced in semi-arid environments (Dyson-Hudson 1980). Accordingly,
75% of the Oromo pastoralists and 98.1% of the Mar respondents, move their
103
Figure 4.14. Condition of the browse during the dry season with only Grewia
species retaining some yellowing, senescing leaves.
Figure 4.15. Condition of the browse in the dry season with only Cadaba rotundfolia
retaining leaves.
104
livestock from place to place in search of feed and water. Three mam factors
determine the route of migration namely, distance to grazing, availability of feed for
livestock and the security of the route. In both pastoral groups, the duration of stay of
camels and the other livestock species (cattle, sheep and goats) in migration differs.
The camels migrate year round, while the mean duration of stay in migration for
cattle, sheep and goats was about 4 months.
It was important to identify the most frequent places of migration for rangeland
evaluation and for the study of the implication of this migration on the rangeland
ecology. The Oromo pastoralists indicated the most frequent places of migration in
the study districts were Bulga riverside, Harole, Kara, Halerne, Fantale, Churcher,
Gababa and Aleka. In the case of the Afars, Samayu, Madala, Fantale, Bulga river
side, near the hot spring (top of bud) were some of the most important places of
migration. The Afars also move their animals outside of the study district (mainly in
the neighbouring Afar districts). In the case of the camels, 90% of the Oromo
pastoralists move their animals to other districts inhabited by other Oromo tribes by
travelling up to 250 km away from their residence. Both pastoral groups indicated
that compared to the past, the intensity of migration has increased because of the
shortage of rainfall and drought, which resulted in a decrease in the amount of forage
produced for the animals.
This migration, which is indicated by Toupet (1975) as highly adaptive and based on
a profound symbiosis between herders and environmental conditions of risk and
uncertainty, was not carried out without the pastoralists facing many problems. The
major problems faced by the pastoralists, as ranked by them in a decreasing order
were: (1) feed and water shortage, (2) death of animals, (3) increase in the incidence
of diseases and (4) problems from wild animals. The elders also discussed other
problems during group discussions. At times the temporary houses at the migratory
places will be destroyed by wind. In another circumstance, stubborn camels on which
they transport their goods will leave away the women transporting the materials.
Furthermore, the elders and the community members discussed difficulties in
.handling babies, management of lambing, kidding or calving and thirst and hunger of
105
the herders during migration. This information mainly obtained from the group
discussion coincided with the perception of the individual household response.
Accordingly, 90% of the Orornos and about 60% of the Afar pastoralists replied that
migration is a bad practice (Figure 4.16). A small proportion of the Afar pastoralists
indicated its importance as a measure to decrease cattle diseases. Many diseases
affecting all classes of livestock species arise during migration (Table 4.13), whose
importance varies from one class oflivestock to another.
100 o It is a bad
90
en practice
E
Q) 80
'"s0 70 III It gives rest tocen..... 60 the grazing landQ)
(;,...,
0 50
Q)
40 El It is important in00
E'" shortage of
Q) 30
u... grazing land
Q) 20
Po. o It is good to
10 decrease cattle
0 diseases
Oromo Afar
Figure 4.16. Perceptions ofOromo and Afar pastoralists towards migration
(Respondents: Oromo= 85 and Afar= 47).
106
Table 4.13. Common diseases observed during migration by livestock species
(percentage of the respondents; Respondents: Oromo= 75 and Afar= 42).
Type Kereyu-Fantale (Oromo) Awash-Fantale (Afar)
of diseases
Cattle Sheep Goat Camel Cattle Sheep Goat Camel
Rinder pest 44.1 68.0 43.3 2.2
Contagious 50.0 62.1 53.3 35.7 56.1 39.3
Bovine
(capparine)
Pleuro
Pneumonia
Pastrolisis 2.9 6.8 6.7 51.1 57.1 46.3 75.0
Anthrax 35.3 42.2 21.4 12.2
Black leg 38.2 4.4 4.9
Actino- 4.4
bacillius
Foot 23.5 4.4 4.9
and mouth
Deep 1.5 2.3 23.3 2.6
wounds on
foot pad
Bloat 20.0 4.4 3.3 1.3 3.6
Liverfluke 13.6 24.1 6.7 4.4 2.4
Diarrhoea 11.4 10.3 3.6
Coenuruses 15.9 31.0 6.7
(Plant
poisoning)
Bottle jaw 2.3 3.4
oedema
Cough 6.9 10.0 2.2 2.4
Skin disease 3.4
Forest fly 7.1
4.3.6. Rangeland use conflict
The word conflict has been used to describe a wide range of interactions between
herders and farmers over natural resources, interactions that are qualitatively
different from each other and clearly of different degrees of severity. Thus, the
umbrella term conflict has been used to cover tension between resource users,
straightforward arguments between individuals, disputes between individuals or
groups, or with the state, legal proceedings between resource users, political action to
evict certain resource users, theft, raiding of livestock, beatings, killing of humans or
107
livestock, and large-scale violence between groups involving multiple killings
(Hussein et al. 1999). The usage of this term under the context of this study was to
explore land use conflict in the study districts, which reflects ongoing competition
over access to scarce land resource between different pastoral groups. Conflict has
been conspicuous for over 30 years and is intensifying. To this effect, information
referring to conflicts on rangeland resources in this part of the country is quite
sensitive. Nevertheless, 42% and 46.3% of the Ororno and Mar pastoralists,
respectively reported that there is a conflict in resource use among the pastoralists
and the main conflict is between the Mar and Ororno pastoralists. There was a
variation in opinion to a question asked if the intensity of the conflict over rangeland
resource use increased over the last 3-4 years (Figure 4.17). Many of the pastoralists
in both districts replied that the intensity has increased in the last 3-4 years. The main
reason for the increase of conflict by both pastoral groups was livestock feed
shortage (86.7-90.5% of the respondents in both pastoral groups), where the intensity
of the conflict increase in the dry season when feed shortage is critical. The other
reason given by both pastoral groups was a lack of clear boundary demarcation
between the two neighbouring pastoral districts. Some households, on the other hand,
reported that there was a decrease in the intensity of the conflict over the last 3-4
years and have attributed this to a change in attitude of pastoralists due to education,
agreement among the pastoralists and protection by the government.
en ~~--------------------------------~
i:: 80
-~g 70 o Ororno
8.60
en IIIl Afar
~ 50
""o"'~ 40
~ 30
~ 20
~ 10
c, 0 +----'--
Increased Decreased Do not know
Figure 4.17. Pastoralist's perceptions to the intensity of conflict in the last 3-4 years.
(Respondents: Oromo= 85 and Mar=51).
108
4.3.7. Resource utilisation from conservation area
Both pastoral groups are directly or indirectly dependent on the Awash national park
of Ethiopia for their livestock feeding. Particularly in the dry season when there is
critical feed shortage women usually harvest grass from the park (Figure 4.18) or
herders will graze their animals in the park. Twenty percent of the Oromos and
56.6% of the Afars replied that their stock graze in the park during the dry season.
This response however, seems low from what is actually happening in the study
districts. Similarly, 73.3% of the Oromos and 94.2% of the Afars replied that the
game and their livestock use the same source of water.
Figure 4.18. Grass harvested from Awash National Park by an Oromo woman and
transported with a donkey.
4.4. DISCUSSION
4.4.1. Demography of the pastoralists surveyed
The average family size in the study districts was 6.74 and this was comparable with
the results of Tibabu (1997) and Mudris (1998). The result also indicated that the
109
level of education of the respondent households was low in both pastoral groups.
This has many implications. In one way, it shows that the status of education in the
study districts must be improved. The low level of education can also be a hindrance
for the transfer of technology.
Both pastoral groups in the study districts depend primarily on the sale of livestock
as a source of income and this is in agreement with a report of Perrier (1995). The
second source of income for the Afars was the sale of milk and milk by-products,
while for the Oromos it was the sale of crops. This is a very interesting difference
because these two pastoralist groups are living in neighbouring district where one
would expect a similar pattern. This reflects in part the difference in the type of
production system they are following. The Afar pastoralists in general are more
involved in pastoralism than the Oromo respondents. This difference might also be
associated with the culture of the two pastoral groups.
4.4.2 Livestock ownership
The results of the study indicated that females dominate the livestock population for
all species. This is in agreement with the reports of others (Coppock 1994; Tibabu
1997; Mudis 1998; Roderiek et al. 1999; Ndikumana et al. 2001) who indicated that
a female dominated herd structure was used to offset long calving intervals and thus
stabilise milk production. Perrier (1995) argued that western ranchers focus strongly
on the production of livestock for terminal products, especially meat. Most
traditional pastoral and agropastoral groups, on the other hand, place at least as
strong emphasis on harvesting a continuous flow of non-terminal products, such as
milk, fibre, manure, traction and transport. The traditional goal of almost all
communal land livestock owners is an unlimited increase in the number of animals
owned. Numbers and not productivity have been the major objective (Bembridge &
Tapson 1993). Many pastoral and agropastoral households keep more than one
species of livestock, which is also practiced by the pastoralists in the study districts.
Herd diversity appears to be a strategy that is particularly useful in arid areas, where
advantage can be taken of the various adaptations of the different livestock species.
Studies have shown that mixed stocking with two or more species of different
110
feeding habits make more effective use of vegetation and are often more profitable.
Moreover, different livestock species are valued for different reasons (Behnke et al.
1993; Scoones 1995b).
The mean number of sheep and goats owned per household was lower than that
mentioned in previous reports (Tibabu 1997). Although previous data indicating
ownership per household in the case of Mars was not segregated into sheep and
goats, it was reported that the mean ownership of sheep and goats per household to
be 47. The result of this study indicated the mean number owned per household for
sheep and goats in Awash-Fantale district to be 38. This decline in the number of
small ruminant mainly comes from sheep and this could possibly be associated with
disease problems, which was supported by the elders in the study districts. Disease as
a major constraint to small ruminant production over the years has been noted by
Coppock (1994) in the Borana rangeland of Ethiopia and by Wilson (1988) in Africa.
4.4.3. Poisonous plants
Plant poisoning is a problem that can have an economic impact on the livestock
owner when animals succumb and reports of animals being poisoned by a variety of
plants is documented in different countries (Garland 2000). Devastating outbreaks of
poisoning have been reported under conditions of droughts, overstocking and
uncontrolled fires in Africa. For example, during 1926 and 1927 about 600 sheep
died of plant-induced photosensitivity in the Northern Cape Province and during
1929 and 1930 over one million were killed by Geigeria species in Griqualand west
(Naude et al. 1996).
It was indicated by the pastoralists that both woody and other herbaceous plants at
certain times cause poisoning to livestock. Although resource and time did not permit
chemical analysis on the poisonous properities of the mentioned plants, the incidence
of being poisoned by these plants is indicated in the literature (Irvine 1961;
Verdcourt et al. 1965; Drummond & Moll 1977) from Kenya, Uganda and Southern
Africa. Verdcourt et al. (1965) for example, indicated that 0.05% of a quaternary
111
ammonium compound was found in Capparis fascicularis. The genus Tephrosia,
containspoison (Verdcourt et al. 1965) a very few plants whose foliage appears to be
somewhat toxic, e.g., Tephrosia taub is poisonous to rabbits. According to an old
record of a German officer in Tanganyika, T radicans, causes diarrhea in cattle,
though it is a good cover and liked by cattle. The principle toxin is tephrosin, which
also has insecticidal properties. Cryptostegia grandiflora is also reported to be
poisonous to livestock in Mexico and Australia where it was introduced (Irvine
1961) and according to Irvine (1961), Erythrina abyssinica was found to contain
traces of alkaloids (0.05- 0.10%). Erythrina abyssinica contains a curare-like poison,
which if injected into the blood stream produces anaesthesia, paralysis and even
death due to respiratory failure (Drummond & Moll 1977).
In East Africa, no species of grass is considered poisonous. However, several have
been known to develop toxic properties under certain conditions. Species of the
grasses Aristida, Cymbopogon, Themeda triandra, Cynodon dactylon and Cynodon
plectostachyus have all shown to be capable of building up quantities of hydrocyanic
acid in their tissues. This abnormal production of hydrocyanic acid is brought about
under the influence of drought, frost, trampling, mowing or wilting by the action of
enzymes on the cyanophoretic glycosides normally present in the plants (Verdcourt
et al. 1965).
The season in which poisonous plants of woody origin affect animals was indicated
to be mainly during the dry season, while that of the other plants during the wet
season. The reason for this is related to the abundance of these plants in marshy,
swampy areas and along the river. During the dry season, animals graze in these
mentioned areas and get these plants accidentally when they are grazing on other
plants. This is particularly true in the case of the Mars. For the other plants, the wet
season is related to their abundance in that specific season. The incidence of
poisonous plants is more common on the side of the Mars than on the side of the
Oromos. This suggests that their distribution, ways of dissemination, economic
effects and the effectiveness of the traditional control methods be studied in more
detail.
112
In general, plant poisoning will continue to be a problem and there are numerous
reports to this effect in the veterinary literature (Garland 2000). It is not uncommon
to find poisonous plants growing on sown and natural pastures and public lands
(Jarries et al. 1992).
4.4.4. Uses of plants and Bush encroachment
The results of the study indicated that both pastoral groups use rangeland plants such
as grasses and woody plants for different purposes. The uses of woody plants by the
pastoralists for different purposes have been described by many authors (Skerman
1977; Le Houerou 1980; Walker 1980; Woodward & Reed 1989; Scholes & Walker
1993; Breman & Kessler 1995; Diress et al. 1998; Sharon 2000; Smit 2002). The
primary use of woody plants in the study district was as a source of feed for
livestock, which is in agreement with the report of Diress et al. (1998) who reported
that in the northern tip of Mar the pastoralists primarily use woody plants for
livestock feeding. Both pastoral groups have an excellent knowledge of native
vegetation, a common fact among the rural people in Africa that has spurred interest
in ethno botany (Morgan 1981; Stiles & Kassam 1984; Marx & Wiegand 1987;
Mathias-Mundy & McCorkle 1989). Accordingly, plants used as traditional
medicines for people or animals may have a role to play m developing more
sustainable health practices in situations where imported drugs are expensive or
unavailable (Tafesse 1990). Further research is required to investigate the usage of
different woody plants species. Forage value of plants should not be the sole criteria
for management objectives and the community should be taught to conserve woody
plants that are disappearing.
Increase in bushes and shrubs in the grazing lands was one of the problems indicated
by the pastoralists, which is in agreement with the general problem of the rangelands
in Ethiopia (Coppock 1994; Alemayehu 1998; Beruk & Tafesse 2000; EARO 2001)
and other semi-arid savannas and grasslands of the world (Buffington & Herbel
1965; Blackburn & Tueller 1970; Barnes 1979; Bucher 1987; Harrington et al. 1984;
Archer 1989; Matheson & Ringrose 1994). The causes of bush encroachment are not
113
absolutely clear to both the pastoralists and the rangeland ecologists. However,
varied and complex reasons are associated with bush encroachment. In light of the
existing knowledge and the debate about the possible causes and mechanisms of
bush encroachment, the possible reasons indicated by the pastoralists fall within the
scientific context. The decrease in grass production due to increase in the density of
woody plants is well known (Abel 1997; Smit 2002). Smit (2002), based on his
extensive review of the bush encroachment problem in Southern Africa indicated that
both natural process and secondary determinants could result in increases in the
density of woody plants. The details of the bush encroachment problems will be
discussed in Chapter 5.
4.4.5. Rangeland and perceptions towards rangeland condition
The fact that the communal type of ownership within the grazing lands is preferred
by almost all the Mars and many of the Oromos confirms the fact that, in a pastoral
production system, in contrast to the individually owned livestock, land resource is
frequently owned and managed by groups (Perrier 1995). Pierrier (1995) explained
that the socio-economic conditions of the pastoral system tends to favour individual
ownership of valuable productive resources that can be controlled, such as livestock,
and communal ownership of less valuable and difficult to control resources such as
rangelands. The difference between the Mars and the Oromos with regard to choice
of land ownership is an indication of the situations in which the two pastoral groups
live and the type of production system in which they are operating. The Mars in
general are more pastoral than the Oromos. Tsui et al. (1991) argued that with
population increase and the expansion of the agricultural frontier, land use and
tenurial arrangements will change more towards privatized rangeland areas.
The numerous problems faced by the pastoralists in the study districts with regard to
rangeland resource, be it from problems imposed externally such as the allocation of
rangeland to non-pastoral use, immigration or internally (e.g., population growth) is a
common characteristics among nomadic pastoralists living in a semi-arid area
(Unruh 1990; Thurow & West 1996). The external forces, particularly the allocation
114
of land to non-pastoral use have displaced and up setted the ecological viability of
their economies in Ethiopia (Desta 1993; Ali 1993; Tibabu 1997; Mudris, 1998;
EARO 2001) and elsewhere (e.g., Harbenson 1990; Unruh 1990; Behnke 1994).
Behnke et al (1993); Kipuri (1995); Scoones (1995b); Mokwunye (1996) and David
et al. (2000) argued that pastoral people are suffering considerable erosion of their
lifestyles in many arid parts of Africa through both external and internal problems.
If one traces back the root cause of land allenitation, it can be attributed to the
attitude towards the pastoralists and their mode of production. International
development agencies and African governments have devoted considerable effort to
the suppression of the pastoral techniques of land and livestock management on the
presumption that pastoralism was inherently irrational, ignorant, unproductive and
ecologically destructive (Hardin 1968; Picardi & Siefert 1976; Livingstone 1977;
Sandford 1983). However, recent studies by anthropologists, ecologists and
economists indicated that the indigenous systems of pasture protection and
sustainable mode of exploitation were compatible (Homewoods & Rodgers 1989;
Galaty & Johnson 1990; Fratkin 1997). In contrast to the previous generations of
researchers, these people have stressed the immense body of indigenous knowledge
of the environmental processes (Unruh 1995; Sillitoe 1998). This study supports the
latter assumption as was evident from people's response to issue of their grazing
lands and their excellent knowledge of plants. Most of the problems the pastoralists
face in the study area were created externally without leaving alternatives to the
pastoralists such as land allenitation. In general, the pastoralists living in the study
area are dealing with grazing systems, which are characterised by complexity, high
variability and uncertainty. Feasible management options in these unstable
environments are not simply attenuated or less precise versions of management in
more temperate settings. Furthermore, this system as it stands now cannot be moved
sustainably and alternatives like modernizing livestock production and improving
social welfare needs attention.
Previous studies (Tibabu 1997; Mudris 1998; Sharon 2000) and different overviews
(Desta 1993; Ali 1993.;EARO 2001) have indicated that the study area is a resource
115
use conflict area. Land is occupied by different development enterprises,
conservation areas, crop producers and the pastoralists. Furthermore, land is required
for the expansion of farming by people living in the town and by workers in the
different plantations. Just considering the Oromo group alone, because of the
difference in the type of production systems they are following (crop production and
pastoralism), rangeland can mean one thing to those involved in crop production and
mean another thing to those who are exclusively deriving their livelihood from
livestock production. Therefore, the coming conflict in land should not be expected
from the different pastoral groups alone, but also from pastoralists of the same ethnic
group and this was quite evident when collecting data in the field. Numerous studies
have confirmed that with the increase in human and livestock population and
accompanied by the repeated drought, more pressure will be exerted on the land
(Baxter 1994; Ward et al. 1998; Freudenberger & Wood 2000). To this effect, many
places in the study districts have already reached a threshold point (Schloeder &
Jacobs 1993), where the rangelands .cannot be upgraded without assistance. Although
it is quite unjustified to put all the blame on the animals, research done elsewhere
have indicated that overgrazing can lead to a decline in the quality and productivity
of the rangelands (Lazenby & Swain 1969; Coupland 1979; Cossins & Upton 1985;
O'Connor 1985).
Hussein et al. (1999) indicated that an important characteristic of the semi-arid
regions, which affects the livelihood strategies of both herders and farmers, is a short
rainy season (3-4 months), whether this is mono-modal (west Africa) and bimodal
(East Africa), and the unreliability of rainfall (inter-annual fluctuations, fluctuations
in seasonal and spatial distribution). In addition, they have suffered recurrent,
prolonged and extensive droughts during which the rains sometimes have failed
completely. Drought was one of the problems indicated by the community as a factor
increasing more pressure on the land. Since drought is a gradual phenomenon, it is
difficult to clearly demarcate the beginning and end of drought. Wilhite & Glantz
(1985) defines drought as (a) meteorologically (b) agriculturally (c) hydrologically
and (d) socio-economically.
116
In meteorological terms, Pratt et al. (1997) suggest that drought can occur when
rainfall is below half the long-term average or when rainfall in two successive years
falls 75 % below average. According to Coppock (1994) drought is defined as "when
two or more consecutive dry years occur in which the length of the growing period
(LGP) is less than 75% of the mean, i.e., a drought is driven by several consecutive
rainy seasons in which deficient rainfall has a detrimental effect on the production
system.
The Society of Range Management (1974) defines drought as the prolonged dry
weather where the precipitation is less than three-quarters of the average rainfall. The
major criticism forwarded in using meteorological criteria in defining drought rests
on the limitations of so called normal precipitation of an area. Drought with respect
to agriculture is the stress that causes plants to wilt or die and it lowers production.
This is not only a function of the amount and distribution of rainfall, but also a
function of other determinants such as temperature, soil characteristics and
management of the land. Vegetation cover and vigour in a given area can provide a
valuable indicator of meteorological drought. Degraded areas are very much prone to
physiological stress and drought is likely to recur. This explains in part why
pastoralists living in poor condition or degraded areas are affected by recurrent
droughts. The people with degraded rangelands are creating their own droughts and
increase the intensity or frequency of it (Snyman 1998).
Hydrological drought is largery concerned with underground and irrigation water
supply. Snyman (1999a) indicated that catastrophic droughts caused by a succession
of low rainfall years are common on the African content and at these times forage
resources dwindle to very low levels. Examination of the rainfall data in the study
area, even in the last five years, indicated that the rainfall was below average by
about 4-23%. The details of the problem of drought are discussed in Chapter 5.
In conclusion, the future of the rangeland in this study area is dependent in careful
planning of land, considering the different components of the ecosystem in such a
. way that one is complementary to the other.
117
4.4.6. Nahiral minerals
Optimum livestock performance is possible only if the forage contains the necessary
mineral elements. Essential mineral required for normal functioning and productivity
such as Sodium (Na), Phosphorous (P) and Copper (Cu) are often insufficient
amounts in natural forage (Kabaija & Little 1987). According to Oba (1998), in the
lowlands, a primary constituent of improved range management technique is mineral
supplements and the pastoralists practice is quite related to the improvement of the
productive and reproductive performance of their animals. Furthermore, Simonsen &
Mitiku (1998) in their study at the northern tip of Afar indicated usage of minerals
for livestock but from a different source (from hot springs). This study was
conducted at the southern tip of Afar, a region with 29 districts and the studies
generally indicate that pastoralists have developed traditional methods of combating
mineral deficiency, though they cannot put it in a scientific way. Mohammed et al.
(1989) studied soil minerals by collecting soil samples from the Rift valley of
Ethiopia, which encompasses the study districts as well. Their results showed that
these mineral soils are rich in sodium (Na) but generally poor in other elements. The
red soil has relatively better amount of phosphorous (P), which improves the Plevel
in the other mineral soils when mixed with them.
4.4.7. Water resources
The result of this study showed that both quantity and quality of water was a major
problem to the Oromo groups, while the Afars express more on the quality problem.
Within the arid rangelands, water is clearly the decisive resource affecting
productivity (Snyman 1989,1998; Solomon et al. 1991; Falkenmark & Suprapto
1992; Falkenmark & Rockstrom 1993; Smith 1992). Steen (1994) argued that water
shortage is so great that it strongly restricts the potential for producing plant and
animal biomass even with the best-known techniques. The competition for water will
also increase as the human population continues to grow and countries become
increasingly industrialised. In the arid and semi-arid areas, surface water is scarce
and most of these water sources are recharged by rainfall. Other water sources are
118
dependent on underground reservoirs whose supply are unknown and are often
affected by insufficient recharge. Water sources, therefore, reflect the climate and
thus the number and proximity of the water sources will change with climate. Quality
of water is also affected by climatic factors. Problems related to water, particularly to
the Oromo groups, is a reflection of denial of access to water sources for long
periods due to the establishment of the irrigation scheme and the national park (Lane
et al. 1993; Sharon 2000). Therefore, improving the supply of water both in quantity
and quality should deserve due attention to improve the welfare of the pastoralists.
4.4.8. Feed resources
In extensive rangeland systems, livestock production is highly dependent on the
availability of natural grazing, the quantity and quality of which are primarily
determined by the amount and distribution of rainfall, given the temperature regime,
soil type and topography of a particular rangeland site. In East Africa, rainfall
fluctuates widely from year to year (Solomon et al. 1991). Accordingly, livestock in
the rangeland areas, experience large seasonal fluctuations in feed availability and
quality, which is a normal phenomenon in the arid and semi-arid areas (Snyman
1998). In the study area, there is no available data on livestock weight variations
amongst the traditionally managed stock. However, from the general situation in
Ethiopia and the pastoralist' s perceptions in the area, there is a loss in live-weight
and a decrease in milk production throughout the dry season and during the drought.
The stress on cattle due to shortage and low quality of forage results in low
productivity characterized by low calving rates (50-70%), high death rates for calves
before weaning (25-35%), a slow rate to mature size, and a low commercial offtake
(Sullivan & Farris 1976). The critical shortage of féed reported in this study is in
agreement with the situation prevailing across the rangelands in Ethiopia (Coppock
1994; Alemayehu 1998; Mudris 1998; Oba 1998; Beruk & Tafess 2000; EARO
2001) and the shortage in the study districts is mainly a reflection of keeping large
stock numbers in a relatively small area. Furthermore, in a tropical environment,
availability and quality of forage remain high for a short period of time because of
119
the rapid rate of senescence and decay.
The Afars had limited choices of feed resources in companson to the Oromo
pastoralists. Scoones (1995b) argued that mixed farming systems usually increase the
diversity of feed available to animals compared with access to rangeland resources
alone. Pastoralists in this segment of the country follow different mechanisms to
cope with feed shortage like migration, lopping browse trees, buying feed, using
reserve feed and the last was the sale of animals, which are the general practices in
these pastoral systems (Solomon et al. 1991; Smith 1992; Behnke et al. 1993;
Coppock 1994; Scoones 1995b). The reason why the sale of animals in response to
feed shortage was the last option can be explained by the fact that the marginal
nature of the pastoral environments has imposed certain constraints to livestock
production. Hogg (1997) and UNDP (1997) argued that livestock are bred for their
resilienceto drought and disease rather than productivity. Livestock are both the
backbone of their economies and a cultural value in their own right.
In the study area, there are a large amount of by-products produced from the sugar
factory. However, the result in the study indicated that only 27% of the respondents
use the by-products from the factory for livestock feeding. These are people close to
the factory and those involved in crop production and partly employed in the factory.
Pastoralists far out in the rangelands and which do not have alternatives (like the use
of crop residues) do not have access to this by-product. In this part of the country, the
rangeland is degrading rapidly. Therefore, it is essential to facilitate the use of this
by-product for those pastoralists without any alternative by organizing the people in
the form of cooperatives.
4.4.9. Migration
Scoones (1994) & Swallow (1994) argued that flexible movement over rangelands
means that a great variety of grass and tree associations can be exploited, making
good use of the varied phenology, production dynamics and forage quality of the
different sources. Migration in response to resource availability by different pastoral
120
groups is well documented in Ethiopia (Coppock 1994; Tibabu 1997; Alemayehu
1998; Mudris 1998; Oba 1998; Sharon 2000) and elsewhere in the Africa (Smith
1992; Behnke et al. 1993; Scoones 1995a).
The central focuses of the issue of migration in this study rely on three major points,
namely, (1) what does this migration mean to range ecology? (2) expansion of bush
encroachment and (3) disease incidences and the perception of the pastoralists about
migration. This migration in smaller areas creates pressure on the rangeland
resources leading to high grazing intensity. High grazing pressure, in turn, reduces
the growth rate and reproductive potential of individual plants and in doing so
influences the competitive relationships among the different species. Overgrazing of
grasses is identified as the main cause of increased woody plant density in the eastern
areas of Botswana (Van Vegten 1983). Furthermore, when animals move from place
to place in search of feed and water, they can serve as agents in the dispersal of seeds
of different species increasing the chance of bush encroachment.
Diseases could also be transmitted from place to place and from domestic to wild
animals and visa versa where the significance of this problem is discussed by
Grootenhuis (1995), considering the wildlife of Kenya. Macpherson (1995) also
discussed the importance of transhumance on the epidemiology of animal diseases,
which include the relative importance of factors like risk of coming into contact with
geographically limited or seasonally abundant pathogens and also the opportunity for
the interaction of domestic and wild animals. Solutions to address this situation
remain unclear but unless they are investigated the importance of many parasitic
infections go unrecognised (De Beer et al. 1991). The perception of the pastoralists
indicating that migration is a bad practice opens more room for looking at ways and
means of modernizing their life style and the production system they are following.
4.4.10. Rangeland resource use conflict
In semi-arid Africa, a number of analysts see tensions, competition and violent
conflict over natural resources as omnipresent in these regions (e.g., Mathieu 1995a,
121
1995b). Accordingly, the result of this study is in agreement with the report of others
in Ethiopia (Tibabu 1997; Alemayehu 1998; Coppock 1994; Mudris 1998; Oba 1998;
Sharon 2000; Beruk and Tafesse 2001) and in other pastoral areas in Africa (Smith
1992; Behnke et al. 1993; Scoones 1995b). Hussein et al. (1999) argued that
increasing conflict is mainly due to two factors: (1) changing patterns of resource use
and increasing competition for resources and (2) the break down of traditional
mechanisms governing resource management and conflict resolution. Conflicts could
arise from many problems; however, in the study area the main reason was feed
shortage. Therefore a break-through in solving the nutritional constraints oflivestock
in the study area can solve a major portion of the problem in the study area.
Pastoralists perceptions of the intensity of the conflict varied, but many indicated that
the intensity has increased. In light of the growing pressure on rangeland resources in
arid rangelands, compounded by population growth and settlement by immigrants
(Hussein et al. 1999; World Bank 2001) it is likely that the intensity of conflict will
increase unless proper interventions are made.
4.4.11. Resource utilisation in conservation area
Like the situation that exists in Eastern and Southern Africa (e.g., Lane & Swift
1989; Coe & Goudie 1990; Schulz & Skonhoft 1996; Bourn & Blench 1999), the
pastoralists living in the study districts compete with wildlife for livestock feeding.
The result of the study indicated that the pastoralists either harvest material or graze
their animals in conservation areas particularly during the dry season. In light of the
existing situation in the study area, where efforts to improve the nutritional
constraints of livestock is not quite visible, the dependency of the pastoralists on the
conservation area for livestock production will increase, which will gradually lead
ecological deterioration of the rangelands. Ultimately, the rangeland resource base
will reach to a- point where it cannot support the different components of the
ecosystem. Therefore, mechanisms to utilise all possible alternative feed resources
must be planned, if the further degradation of the rangeland ecology is to be
prevented.
122
4.5. CONCLUSIONS ANDRECOMMENDATIONS
The following conclusions and possible recommendations are made from this study:
.:. These two pastoral groups live in neighbouring districts. Nevertheless,
one of the interesting differences between them lies on their source of
income. While the Mars second source of income was the sale of milk
and milk by-products, the Oromos usually do not sell these products .
•:. The problem of poisonous plants and bush encroachment is more
prevalent on the side of the Mars than that of the Oromos whereas, water
is one of the major problems to the Oromo respondents .
•:. Livestock feed shortage is the major constraint in the study area. This
problem will continue to be the major cause of conflict between the
different pastoral groups unless proper interventions like the use of the
by-products from the sugar plantation and increasing the productivity of
the rangelands are made .
•:. Despite the difference in ethnicity, it is essential that the management of
the two districts work close together as they share common problems .
•:. The development of incentives like giving prices for those actively
involved in natural resource conservation and other mechanisms
(involving the elders and village leaders) to encourage the community to
get involved in conservation activities should be considered .
•:. Equipping the agricultural offices with trained personnel in rangeland and
wildlife sciences needs due consideration. None of the agricultural offices
have even one person trained in rangeland management.
.:. With the current approach of the communal grazing systems, sustainable
utilization of the rangeland ecosystem is not possible. Therefore,
alternatives must be planned like better education and employment
opportunity .
•:. The findings in this study clearly justify the need of a more in depth study
on the rangeland ecosystem, which is presented in the next chapters.
123
RANGELAND EVALUATION IN KEREYU-FANTALE AND
AWASH-FANTALE DISTRICTS
5.1. lINTRODUCTION
Poor nutrition caused by inadequate amounts and low quality of feed is one of the
major causes of low livestock productivity in tropical areas' (Kaitho 1997).
Convectional feeds such as grains and oilseed cakes are not produced in sufficient
amounts to meet the requirements of man and livestock in the tropical areas. In these
areas, native or natural pastures make up the bulk of the feed. In contrast to the
highlands, in which exotic forages are more easily established (Jahnke 1982),
sustainable forage interventions in the rangelands depend to a greater degree on
improved use of promising indigenous plants (Coppock & Reed 1992). Forage plants
usually consumed by domestic animals and game consist mainly of grasses and
woody plants. Therefore, assessment of the condition of the vegetation utilised by
grazing and browsing herbivores is fundamental for the sustainable utilisation of the
grazing ecosystem.
Trollope et al. (1990a) defined rangeland condition as the state of health of the
rangeland in terms of its ecological status, resistance to soil erosion and potential for
producing forage for sustained optimum livestock production. Furthermore,
rangeland condition is a function of all plant forms (trees, grasses and shrubs) that
occur in it (Danckwerts 1982; Friedel 1987). Rangeland condition cannot, therefore,
be simply indexed according to its usefulness for a single priority land use. As with
grassland, the composition and structure of each of the other components vary, which
adds an extra and complicating dimension to rangeland assessment. In addition, the
rangeland is frequently used by pastoralists who own different animal types
(browsers and grazers). Assessment techniques need to consider the different
vegetation components for the proper utilization of the available rangeland resources.
Since animal production is directly related to rangeland condition, rangeland
124
degradation will result in a lower income (Danclcwerts & Tainton 1996).
A large number of shrub and tree species have been documented as useful livestock
fodders. These plants form an important component of the diet, especially for a
variety of the herbivores and also have been used traditionally as a source of fodder
for domestic livestock in Asia, Africa, and the Pacific (Skerman 1977; NAS 1979; Le
Houerou 1980; Smit 2002). Owing to the importance of both components (grasses
and woody vegetations) as a source of feed, it was deemed necessary to assess both
components by using different techniques for each of them. The accuracy of
determining rangeland condition and trend will depend on the assessor's ability to
measure changes as well as the correct interpretation thereof (Van der Westhuizen et
al. 1999). Adequate knowledge of the particular veld type, as well as of determinants
such as soil type, topography and climate are therefore essential. In addition to the
assessment of the woody vegetation in terms of density by species, a preliminary
investigation was also undertaken on browse (leaf) production using the techniques
developed by Smit (1989a, b, 1994, 1996).
Until recent times, research on rangeland dynamics has historically focused on the
effects of various management practices on forage production and animal response,
little attention given to the impact of grazing on the nutrient dynamics of the soil.
Therefore, the objectives of this study were to determine the condition of the
rangelands by using herbaceous and woody measures and some soil parameters. In
the end, the intention was to come up with possible recommendations to minimize
the further degradation of the rangeland ecosystem.
5.2. PROCEDURES
5.2.1. Site selection
In the identification of major rangelands commonly grazed by the animals of
pastoralists for the assessment, initially important documents and maps pertinent to
the study area regarding vegetation were reviewed and examined. This was followed
125
by a discussion with the respective agricultural development officers and the elders
of the two pastoral groups about the nature and location of the major rangelands that
are commonly grazed by the animals of the pastoralists. A repeated reconnaissance
survey was made with respective elders and experts from the agricultural
development offices throughout the study districts to observe the nature and
condition of the rangelands. The final decision was also supported by results from
Chapter 4. Therefore, in the selection of rangelands sites for the study, the
importance of each of the major rangelands as perceived by the pastoralists and the
respective agricultural development officers was taken as a major criteria and their
representation of the major rangelands areas in the study districts proper was also
considered. However, one rangeland area in Awash-fantale district was included in
the rangelands sites identified for the study, because of the pastoralists fear of the
expansion of Prosopis juliflora to their rangelands from the neighbouring districts.
Accordingly, 10 rangelands sites from Kereyu-Fantale and Il rangelands sites from
Awash- Fantale were identified for the assessment (Table 5.1).
5.2.2. lField layout
Two different field layouts were used for the study. In rangelands sites without
woody vegetation, three sampling sites (the minimum number of replicates) were
identified through repeated ground surveys after thoroughly observing the nature of
the vegetation and slope. Adequate care was also taken in such a way that each
identified sampling site be representative of the rangeland site from which the
sampling site was identified. For rangeland sites where both components (grass and
woody layer) contribute as a source of livestock feed, a sampling block of land, 3 km
x 1 km (300 ha), representative of the vegefation under consideration, was
demarcated. In all of the rangeland sites, a 25-50 m strip of land was left aside in
order to avoid edge effects. The demarcated area was further sub-divided into three
equal plots for the purpose of stratification (1 km x 1 km). In each of the sub-divided
plots, a belt transect of 50 x 4 m was laid out randomly. The total number of belt
transects per rangeland site was, therefore, three. At each of the belt transects, the
four corners of the plot were marked using wooden pegs and this was undertaken
126
before the beginning of the rainy season. The above layout was done for both of the
study districts.
In three rangeland sites that were either close to a river or lake, sampling was
undertaken not closer than 400-600 m from the river or lake, depending on the
situation.
5.2.3. Data collection
From the rangeland sites of the two study districts, grass species composition, basal
cover, estimated soil erosion and natural pasture yield data were collected for the
herbaceous layer, while woody species composition, density, browse production and
phenological data were collected for the woody layer. Soil samples were also taken
to undertake physical and chemical analyses for some parameters. The techniques
and/or methods followed in collecting the different parameters are given below.
5.2.3.1. Floristic composition of the herbaceous layer
The species composition of the herbaceous layer was determined, based on the
frequency of occurrence, using a wheel point apparatus (Tidmarsh & Havenga 1955)
where the nearest plants were recorded. For the rangeland sites without the woody
vegetation, at each of the sample site, two plots of 100 x 15 m were marked in such a
way that they give a total of 1 000 point observations per site. In rangeland sites with
woody vegetation, 300 point observations were recorded at each of the belt transect
and the readings were undertaken at 3 m interval by revolving the wheel point.
Although, the use of the wheel point method in rangeland condition assessment has
not been practiced in Ethiopia, studies conducted by Hardy & Walker (1991) using
the Richards function (Richards 1959), have shown that a sample size of 300 point
observations are adequate for detailed scientific studies.
127
Table 5.1. Rangeland sites identified for the study in Awash-Fantale and Kereyu-
Fantale districts.
Awash- Altitude Vegetation Kereyu- Altitude Vegetation
Fantale (m. a. s.l) components Fantale (m.a.s.l) component
Studied studied
Lahali 932 Herbaceous and Aleka 1050 Herbaceous layer
woody layers
Degage 1028 Herbaceous and Harolel 1036 Herbaceous and
woody layers woody layers
Madala 909 Herbaceous layer Harole 993 Herbaceous layer
Madalal 909 Herbaceous and Chopi 1137 Herbaceous and
woody layers woody layers
Top of bud 811 Herbaceous layer Mogassa 1046 Herbaceous and
woody layers
Samayu 999 Herbaceous and Kara 1384 Herbaceous and
woody layers woody layers
Top of bud I 809 Herbaceous and Kolbayu 1063 Herbaceous and
woody layer woody layers
Bulga 778 Herbaceous and Tulu Dimitu 1474 Herbaceous and
riverside woody layers woody layers
(BRS)l
Dellakara 868 Herbaceous and Fantale' 1654 Herbaceous and
woody layers Near to the woody layers
Cretor
Awash- 811 Woody layer Beseka 990 Herbaceous layer
Dulcha
junction
Asobabad 865 Herbaceous layer
IThe marked rangeland sites fell on either side of the two pastoral groups.
At each point observation, the nearest herbaceous plant, within a radius of 300 mm
was recorded by species for grasses which included both annuals and perennials.
Non-grass herbaceous species were combined as forbs. If no herbaceous species of
the given criteria occurred within the given radius of the point, it was recorded as
"bare ground". Bare ground was treated as if it was a plant species and gave an
128
indication of plant density (Mentis 1984), which is also an important additional
parameter for recording real changes in rangeland condition (Danckwerts & Teague
1989).
Identification of the species was undertaken at two levels. Plants that can be
identified easily in the field were identified using field guides (Irvine 1961;
Verdcourt et al. 1965; CADU 1974; Ibrahim & Kabuye 1987; Reinhard & Admasu
1994; Van Oudtshoorn 1999) and experienced field technicians' knowledge from
Adami Tulu research center was also used. For those plant species not identified in
the field, a herbarium sample of each species was pressed, labelled and sent to the
National Herbarium of Addis Ababa University for identification. The vegetation
survey was undertaken from 15 August 2001 to 15 September 2001, at the time when
most plants were in their flowering stage.
The identified grass species were classified into three groups based upon desirability.
The desirability ratings were based on their long-term reaction to grazing (ecological
groups) and palatability. The ecological status (decreaser and increaser species) as
defined by Foran et al. (1978) was also taken into consideration. Accordingly, they
were divided into highly desirable, desirable and less desirable species. Highly
desirable grass species include species that are decreasers and perennials with a high
palatability based upon the pastoralists perceptions, while the desirable grass species
are those that increase in abundance with moderate over-utilisation (Increaser Ha),
perennials and which were average or high in terms of their palatability. The less
desirable species include those species of grasses that increase in abundance with
severe or extremely severe over utilisation (Increaser Ilb and Ilc). This group
includes both perennial and annual species that were less palatable as perceived by
the pastoralists. The classification of the grass species into decreaser, lIa, lIb and lIc
followed the method described by Vorster (1982) with some modification to suit the
local conditions. However, Increaser I species were not considered in the study, the
reason was indicated by Tainton (1981), in the more arid region, the Increaser I
group is not represented since climatic limitations, notably rainfall, prevent
succession from proceeding beyond a certain stage.
129
5.2.3.2. Basal cover
The same apparatus (wheel point) used in determining species composition was used
for basal cover assessment and was determined from the proportion of strikes as
described by Tidmarsh & Havenga (1955) and Foran et al. (1978).
5.2.3.3. Grass dry matter yield
The standing crop of the natural pasture was used to estimate the available yield as a
measure of annual productivity. At each of the sample sites, the herbaceous
vegetation (grasses & forbs) was harvested at stubble height when the herbaceous
plants were at 50% flowering stage. This was done by clipping the vegetation in 5
randomly placed quadrates (1 x lm each) per identified sampling site. Separation
was made between grasses and forbs, but the scale of the sampling, did not permit
separation into the different grass· species. At the end of the fieldwork, the samples
were oven-dried at 105 oe for 24 hours and weighed.
5.2.3.4. Soil erosion
At each of the sampling site, the status of the soil surface was assessed visually at
every fifth point of the wheel point during the vegetation survey. The assessment of
soil erosion was based on the amount of pedestals (higher parts of soils, which can be
held together by plant roots, with eroded soil around the tuft) and in severe cases, the
presence of pavements (terraces of flat soil, normally without basal cover, with a line
of tufts between pavements according to the method described by Baars et al. (1997).
The values given were 5,4,3,2,1 and 0 for no soil movement, slight sand mulch,
slope-sided pedestals, steep-sided pedestals, pavements and gullies, respectively.
Of the rangeland sites identified in Awash-Fantale district, in one of them, it was not
possible to study the herbaceous layer, because it was difficult to cross the Bulga
river during the time when plants could be identified. As a result, no survey in the
herbaceous layer was undertaken.
130
5.2.3.5. Woody vegetation sampling
All rooted live woody plants in each of 50 x 4 m (200 m2) belt transects, regardless
of being single-stemmed or multi-stemmed, were counted and used for an estimate of
woody vegetation density per hectare. Furthermore, the spatial canopy of all rooted
live woody plants encountered in the belt transects, were measured at peak biomass.
The measurement consisted of the following: (i) maximum height, (ii) height where
the maximum canopy diameter occurs, (iii) height of first leaves or potential leaf-
bearing stems, (iv) maximum canopy diameter, and (v) base diameter of the foliage
at the height of the first leaves. The spatial canopy volume of each woody plant was
calculated using the BECVOL model (Smit, 1989a,b, 1996).
In order to measure the heights and diameters of the woody vegetation, two
calibrated aluminium poles of 2 and 4 m were used. Dimensions of those woody
plants too tall to measure with the poles were taken using a dimension meter (Smit
1989 c). Furthermore, each woody plant was allocated a leaf carriage score, in order
to determine leaf phenology : where, 0= no leaves, 1= 1-10% of full leaf carriage, 2=
11-40 % of the leaf carriage, 3= 41-70% of the full leaf carriage, 4 = 71-100% of full
leaf carriage (Smit 1994; Smit 2001). The leaf phenological scoring was undertaken
for two consecutive months during the wet seasons (third and fourth week in August
and September, 2001) and for two consecutive months during the dry season (third
and fourth week in December 2001 and January 2002). For the purpose of leaf
phenological observations during the dry season, 12 representative plants per species,
were randomly selected in each plot but, in the case of less abundant species, the
available individuals were selected. The selection procedure for the woody
vegetation was based on a proportional sampling of all representative sizes of the
woody vegetation in each plot.
5.2.3.6. Soil sampling and chemical analysis
Ten soil samples were taken randomly in each of the sample sites to a depth of 150
mm with an auger. Each set of 10 samples was bulked to account for short-range
131
spatial variability, thoroughly mixed and one sub sample taken, labelled and kept in a
plastic bag for analysis. The analyses included pH (H20), electrical conductivity
(EC), percent total nitrogen (N), percent organic carbon (OC), texture, available
phosphorous (P) and available potassium (K). Initially the soil samples were air-
dried then sieved to pass a 2 mm sieve. The pH of the soil was determined in a 1:2.5
ratio of the respective suspensions (Thun et al. .1955), while electrical conductivity
was determined in a 1:2.5 soil water suspension (Chopra & Kanwar 1976). Percent
OC was determined according to the Walkley & Black method (1934), total N using
the Kjeldahl procedure (Jacks on 1958) and available P using the method of Bray &
Kurtz (1945). The different particle sizes were determined using the Bouyoucous
hydrometer method (Day 1965) and available K determined using the Morghan
extracting solution.
Of the rangeland sites, identified for the study in Awash-Fantale, soil samples were
not taken in two of them. In one case the soil was too rocky and in the other, there
was a logistical problem in obtaining the sample.
5.2.4. :Data analysis
5.2.4.1. Grass species composition and other related parameters
The frequency of occurrence of each species, including that of the bare ground, was
expressed as a percentage of the total number of points, while the proportion of the
different grass species according to their desirability (highly desirable, desirable and
less desirable) and life form (annual and perennial) was also calculated using
percentages. In the ordination of the different grass species, a detrended
correspondence analysis (DCA) in the DECORANA package (Hill 1979), was used,
unfortunately failed to reveal a meaningful community gradient in both districts
(Appendix 5.1 and 5.2). Basal cover was calculated from the proportions of strikes
as described by Foran et al. (1978) and Mentis (1981). For data on grass DM yield,
estimated soil erosion values and soil parameters, mean and standard deviations were
calculated for each of the studied rangeland sites.
132
Estimation of the grazing capacity was made from the yield of the grass layer, using
the formula proposed by Moore et al. (1985) and again described by Moore &
Odendaal (1987) and Moore (1989) as follows:
Y> d -ê- [DMx fl
r
Where, y = grazing capacity (ha AU-I)
d number of days in a year (365)
DM = total grass DM yield (kg ha")
f = utilization factor
r daily grass DM required per LSU (2.2% of body mass = 10 kg
dal)
The grazing capacity was calculated using both hectare per Large Stock Unit (ha
LSu-I) and hectare per Tropical Livestock Unit (ha TLu-I). In the calculation of
grazing capacity based upon TLU, the assumption taken was that, an animal will
consume 2.5 % of its body weight (Boudet and Riviere 1968; Minson & McDonald
1987), thus each TLU will consume 6.25 kg of forage dry matter daily. In both LSU
and TLU, the utilization factor used is 0.35. Theoretically, the utilization factor
expressed as a decimal value represents that part of the available material that can be
consumed. Actual consumption is limited by grazing preferences of the animals and
losses due to trampling and environmental factors. The percentage of available dry
matter that the animals will actually consume is determined by factors like,
palatability of the plant material and the species of the animal (bulk feeder or
concentrate feeder). However, even when the animals will be able to consume a high
percentage of the available dry matter, their intake should be limited to pre-
determined level to avoid overgrazing. The utilization factor may thus vary from 0.2
(20%) to 0.5 (50%) with an average of 0.35 (35%) that is commonly used (Smit
2002).
The condition of each of the rangeland sites, based upon species composition, was
calculated by multiplying the percentage frequency of each of the species with the
value attributed to each species based upon desirability and this was summed for
each rangeland and expressed as a percentage. The experience and knowledge that
was developed in South Africa particularly Vorster (1982), Tainton (1999) and Van
133
Oudtshoorm (1999) were used with some modifications to suit local conditions.
Grass species of highly desirable, desirable and less desirable qualities were
multiplied with a value of 10, 7 and 4, respectively, while bare ground and forbs
were allocated a value of 1 as they were undesirable. The maximum score is 1000,
i.e., 100% highly desirable species and the minimum is 100, i.e., 100% undesirable
species.
5.2.4.2. Woody layer
Leaf dry mass and leaf volume estimates were calculated using a modified version of
the quantative description technique of Smit (1989a) and Smit (l989b) as described
by Smit 1994. This technique provides estimates of the leaf dry mass and the leaf
volume at peak biomass, based on the relationship between the tree's spatial canopy
volume and its leaf dry mass and leaf volume. This technique was compiled into the
BECVOL-model (Biomass Estimates from Canopy Volume) (Smit 1994). It
incorporates regression equations, developed from harvested trees, which relate
spatial volume (independent variable) to leaf volume (dependent variable): in y = -
3.880 + 0.708x, r=0.941, p<O.OOI, and leaf volume (dependent variable): In y = -
2.933 + 0.697x, r = 0.948, p<O.OOl. Spatial tree canopy volume (x) was transformed
to its normal logarithmic value, while y represents the estimated leaf dry mass (g)
and leaf volume (cm') respectively. The number of ETTE ha" was subsequently
calculated from the leaf volume estimates (1 ETTE = mean leaf volume of a single-
stemmed tree with a height of l.5 m = 500 crrr' leaf volume (Smit 1989a). The
BECVOL-model, in addition to the regression equation for specific tree, incorporates
a number of "general" regression equations (e.g., for microphyllous and broad-
leaved tree species), which (Smit 1989a, b) facilitated the calculations for those
species that did not exist on the list of the harvested tree species. In addition to the
total leaf DM ha", stratified estimates of the leaf DM ha" below l.5 m, 2.0 m and
5.0 m respectively, were also calculated, with the BECVOL- model (Smit 1994). The
height of l.5 m represents the mean browsing height of goats (Aucamp 1976) and
impala (Aepyceros melampus) (Dayton 1978), while 2.0 m and 5.0 m represents the
mean browsing height of kudu (Tragelaphus stepsiceros) (Wentzel 1990) and giraffe
134
(Giraffa camelopardalisï (Skinner & Smithers 1990), respectively. These browsing
heights are mean heights and not maximum browsing heights, which were only used
to draw comparisons. It is known that large individuals are able to reach higher than
those mean heights, e.g., 2.5 m and 5.5 m for kudu and giraffe, respectively (Dayton
1978), while breaking off branches may enable some browsers to utilize browse at
higher strata (Styles 1993).
Estimates of the browsing capacity for the most important woody plants based on
leaf DM were made from the leaf DM estimate, using the formula proposed by Smit
(l999b).
Y>DMI.5) X P x f
r
Where, y = browsing capacity (ha BUl)
DM (1.5) = totalleafDM yield ha-l up to a height of 1.5 m
p = phenology (0= no leaves, 1.0 = peak biomass)
r = daily browse DM required per BV (3.5 kg dal)
BU Browser unit defined as the metabolic equivalent of an average
kudu with a mass of 140 kg (Dekker 1997)
f utilisation factor
From the stratified BECVOL estimates of leaf dry mass below 1.5 m and leaf dry
mass greater than 1.5 m up to a height of 5 m, the approximate browsing capacity
was calculated for the year. The available dry mass of the two strata was calculated
by multiplying the total dry mass with a utilization factor and phenology of the
different woody plant species in different months. The leaf phenology for December
and January (critical dry months) and August and September (wet months) was
obtained by collecting data in the field, while for the rest of the months it was
estimated based upon the long term experience of the pastoralists. The rainfall
pattern of the study area was also taken into account. The leaf phenology and
palatability for each of the specific rangeland sites were obtained by taking the
135
average of the species in the respective rangelands. The allocated leaf phenological
values ranged from 0.2 (lowest leaf yield) to 1 (maximum leaf yield).
5.2.4.3. Statistical analysis
In order to establish relationships among the variables studied (Grass yield, basal
cover, bare ground, estimated soil erosion values, ETTE ha", soil parameters and
altitude), particularly with reference to their relation with the grass yield, a
correlation matrix was undertaken using SAS (Statistical Analysis System) (1987).
Based on the results obtained in the correlation matrix and taking into account the
normal restrictions with regard to the inclusion of correlated variables in multiple
regression functions, a series of stepwise multiple regression analyses were
conducted with grass DM yield as the dependent variable and the other measured
environmental variables as independent variables using SAS (1987).
5.3. RESULTS
5.3.1. Awash-Fantale district (Afar)
5.3.1.1. Herbaceous layer
5.3.1.1.1. Grass species composition
A total of 32 grass species were identified in the study district, with Chrysopogon
plumulosus, Dactyloctenium aegypticum, Tetrapogon cenchriformis, Sporobolus
natalensis, Cymbopogon commutatus, Paspalm dilatatum, Sporobolus spicatus and
Sporobolus ioclados, being the most abundant species (Table 5.2). Based upon their
life form, 7l.9% were perennials and the remaining (28.1%) annuals, whereas, based
upon desirability, 18.8%, 28.1% and 53.1% comprised highly desirable, desirable
and less desirable grass species, respectively. Of the 10rangeland sites studied in
Awash-Fantale district, in 5 of them C. plumulosus was the dominant grass species
ranging in abundance from 3.7% to 86.3%. At the Lahlai site, the most dominant
136
grass was T cenchriformis (44.8%), while Madala 1 was dominated by D.
aegypticum (30.0%). The two dominant grass species at Asobabad were S. spicatus
(41.8%) and S. ioclados (29.1%), while S. natalensis (13.4%) and D. aegypticum
(12.0%) were abundant at Samayu, and 36.0% of the grass layer along the Bulga
riverside was dominated by P. dilatatum (Figure 5.1). The non-grasses, combined as
forbs, comprised 10.5%, despite the variation in the abundance of forbs across the
rangeland sites, which varied between 0.6% (Top of bud) and 32.2% (Samayu).
5.3.1.1.2. Bare ground
The percent bare ground (mean ±SD), over the rangeland sites within the Awash-
Fantale district, varied from 1.8 to 10.8 with a mean of 5.5, which was highest at
Asobabad and lowest at Top of bud (Table 5.3).
5.3.1.1.3. Basal cover
The percent basal cover (mean ± SD) for the rangeland sites was highest at Top of
bud (4.8 ±0.4), and lowest at Samayu (2.4 ±0.6), while the overall mean basal cover
across the rangeland sites studied was 3.4%. Of the studied rangeland sites, 30% had
a basal cover less than 3%, whereas, 50% of them had a cover value of less than four
percent. Only 20% of the studied rangelands had a cover value greater than 4%,
which in general implied that most of the rangelands have a very low basal cover
(Table 5.3).
5.3.1.1.4. Grass dry matter yield!
The dry matter yield of the grass layer (mean ± SD) was lowest at Samayu (256.6 kg
ha") and highest 'at Degage (857.0 kg ha") with a mean yield of 437.1 kg ha-1
137
Table 5.2. Desirability, life form, and frequency (%) of the different grass species in
Awash-Fantale district.
Species Desirability Life form Frequency (%)
Chrysopogon plumulosus Highly desirable Perennial 29.73
Dactyloctenium aegypticum Less desirable Annual 8.05
Tetrapogon cenchriformis Highly desirable Perennial 6.90
Sporobolus natalensis Less desirable Perennial 5.85
Cymbopogon commutatus Desirable Perennial 4.59
Paspalum dilatatum Less desirable Perennial 4.47
Sorobolus spicatus Less desirable Perennial 4.17
Sorobolus spicatus Less desirable Perennial 4.17
Sorobolus ioclados Desirable Perennial 2.92
Paspalum glumaceum Desirable Perennial 2.59
Setaria verticellata Less desirable Annual 2.18
Coelachyrum poiflorum Less desirable Annual l.73
Cynodon dactylon Less desirable Perennial 1.45
Trag_usberteronianus Less desirable Annual l.37
Eragrostis cilianensis Less desirable Annual l.36
Pennisetum stramineum Desirable Perennial l.32
Cymbopogon excavatus Highly desirable Perennial 1.30
Cenchrus ciliaris Highly desirable Perennial 0.90
Cenchrussetigerus Highly desirable Perennial 0.86
Heteropgon contortus Desirable Perennial 0.70
Panicum coloratum Highly desirable Perennial 0.52
Urochloa panicoides Less desirable Annual 0.29
Lintonia nutans Desirable Perennial 0.25
Chloris roxburghiana Desirable Perennial 0.20
Hyparrhenia hirta Desirable Perennial 0.11
Aristida adscensionis Less desirable Annual 0.06
Aristida adoensis Less desiable Perennial 0.05
Sporobolusfestivus Less desirable Perennial 0.02
Digitaria ternata Less desirable Annual 0.02
Eragrostis racemosa Desirable Perennial 0.01
Chloris pycnothrix Less desirable Perennial 0.01
Sorghum purpureosericeum Less desrabie Annual 0.01
Enneapogon schimperanus Less desirable Perennial 0.01
138
Asobabad Cl S. ioclados Dellakara ~ C. plumulosus
o S. spicatus IIID D. aegypticum
Cl S. verticella ta
8.38% IEl S. natalensis
El T. cenchriform is
41.77%
1.51%
12.42%
Degage Cl C. plu m ulosu s Samayu II1II S. natalensis
DC. cam m utatus Cl T. berteronianu s
Cl S. natalensis 3.74% El S. verticellata
ID. DD. aegyp ticu m
12.00% l1li C. plumulosus
Top of bud DC. plum ulosus Bulga riverside s C. dactylon
Cl S. natalensis Cl P. dilatatum
8.32%
5.19% ClP. glu m aceu m
36.15 22.84%
86.89%
Lahali Iii! C. plumulosus Madalal I!!I C. plumulosus
Cl C. commutatus IBiI C. p oiflorum
12.88% Cl H. contonus Ell C. com m utatu s
III S. natalensis Cl C. dactylon
Cl T. cenchriform is Cl D. aegypticum
12.17%
44.
6.85%
4.29%
o C. plumulosus Top of bud I Cl C. plumulosus
El C. poiflorum IBC. com m u tatu s
4.5~7% IiIII P. stramineum 10.24% Cl D. aegyticu m
10.83% III T. berteronianus ClP. glum aceum
5.29% iM~· EI S. natalensis
58.58%
Figure 5.1. The most dominant grass species in each of the rangeland sites in
Awash-Fantale district.
139
Table 5.3. Mean and standard deviation of bare ground (%), basal cover (%), grass
yield (kg ha-i), estimated soil erosion values and grazing capacity (ha
LSUi and ha TLUi) for the rangeland sites studied in Awash-Fantale
district (N= 10; HL = Herbaceous layer and WL = Woody layer).
Parameter Rangeland
Rangeland Lahali Dega- Bulga Madala- Top of Dell a- Sama Top of Madala Asob-
sites ge River 1 budl Kara -yu bud Abad
Vegatation Ill- Ill- Ill- Ill- Ill- Ill- Ill- Ill- Ill-
layer & & & & & &
studied WL WL WL WL WL WL
Bare 4.3± 3.4 ± 8.1 ± 2.9± 4.6± 4.1± 10.2± 1.8± 5.0± 10.8±
ground (0/0) 1.44 0.63 1.6 0.9 2.2 0.5 4.4 l.l 0.5 3.1
Basal cover 2.9± 4.6 ± 2.3 ± 3.9± 3.4± 3.3± 2.4± 4.8± 3.7± 3.0±
(0/0) 1.2 2.0 0.8 0.6 0.3 0.8 0.6 0.4 0.7 0.0
GrassDM 360.2± 857 292.8 328.8± 325.0± 359.8± 256.6± 558.8± 488.1±1 543.8±
yield 20.2 ±44.6 ± 35.8 96.9 90.3 43.2 69.5 90.9 41.8 45.8
(Kg ha")
Estimated Rocky 3.00± 4± 3.00± 3± 4± 4± 3± 3± 2±
soil erosion 1 0.0 0.6 0.0 0.0 0.6 0.6 0.6 0.6
Grazing 25.3± 10.7± 31.2 ± 27.5± 28.1± 25.4± 35.6± l6.3± l8.7± l6.8±
capacity 1.4 0.6 4.1 9.6 8.6 3.0 10.7 2.8 6.3 1.4
(hal LSU')
Grazing 15.8± 6.7 ± 19.5 ± 17.4± 17.6± 15.9± 22.2± 10.2± 11.7± 10.5±
capacity 0.9 0.4 2.6 6.0 5.4 1.9 6.4 1.7 3.9 0.9
(ha TLU')
for the studied rangeland sites. The lowest and the highest grass DM yields were
obtained in rangeland sites with a woody layer (Table 5.3). Unfortunately, the grass
DM yield was not studied on a seasonal basis.
5.3.1.1.5. Estimated soil erosion
The estimated soil erosion values, across the rangeland sites, varied between 2 and 4
(Table 5.3), while ll.l %, 55.6% and 33.3% of the studied rangeland sites had a
value of 2, 3 and 4, respectively. This indicates that the condition of the soil surface
varied between steep sided pedestals to slight sand mulch, which clearly shows the
loss of soil from the surface.
140
5.3.1.1.6. Grazing capacity
The grazing capacity varied from 10.7 to 35.6 (ha LSUI) with a mean value of 23.6
(ha LSUI), whereas, when the grazing capacity expressed in ha TLUI, the values
varied between 6.6 and 22.2 with an overall mean value of 14.7 (Table 5.3). The
grazing capacity was the lowest at Samayu, where the yield of the grasses (256.6
kg/ha) and the basal cover (2.4) of the site was the lowest. On the other hand, the
grazing capacity of Degage was better than other sites, where the yield of the grass
was higher than the others.
5.3.1.1. 7. Rangeland condition
The condition of the rangeland sites, based on desirability (grass species
composition, forbs and bare ground) was highest at top of bud (90.52%), and the
lowest at Samayu (30.52%) (Figure 5.2), while the mean across the rangeland sites
was 62.79%.
~.......:i? ~;;§> ~" '" ~..~ ~ {Jl'"{f "_'" ~
. b.~ ~~c ~ N
~'b' 0'C' l "_~'"~~ »~. o: '" S"<::)~ ~ ê§J ~'b'
"..o~ ~~ ~<,i
~~
Rangeland sites
Figure 5.2. Rangeland condition of the rangeland sites in the Awash-Fantale district
based upon desirability of species.
141
5.3.1.2. Woody vegetation
5.3.1.2.1. Woody vegetation composition, density, ETTE and palatability of
woody plants
In the rangelands studied, 25 woody species were recorded in the transects (Figure
5.3). There were also other woody species that were not recorded within the
transects, such as Dobera glabra, which will increase the number of woody species
in the area. Based upon abundance (%), the dominant woody plants in the rangeland
sites studied were species of Acacia (A. senegal and A. nubica), species of Acalypha
(A. fruticosa and A. indica) and species of solanum (s. marginatum and S. incanum)
and Vernonia eineraseens (Figure 5.3). The Acacia species (A. senegal, A. robusta,
A. mellifera, A. nubica, A. tortilis, A. nilotica and A. seya!) generally comprised 43%
of the total woody vegetation density, which implies that the area is dominated by
Acacia species. Of the species of Acacia, A. tortilis (0.31 %) and A. nilotica (0.57%)
were the least abundant (Figure 5.3).
When the woody vegetation data is expressed in terms of ETTE (%) three woody
species were dominant, namely, A. senegal, A. nubica and P. juliflora (Figure 5.4),
which were followed by A. nilotica, C. rotundfolia, Acalypha species and A.
mellifera.
The woody vegetation density was highest at Top of bud (3 733 plants ha-I) and
lowest in rangeland dominated by P. juliflora (467 plants ha-I). However, when the
woody vegetation density data was expressed in terms of ETTE ha", the values were
higher in rangeland sites dominated by P. juliflora (10 722 ETTE ha") than at top of
the bud (10 130 ETTE ha"). Of the 8 rangeland sites, in three of them, the highest
ETTE ha" value was contributed by A. senegal (Dellakara= 4 633; Lahlai= 3 617
and Top ofbudl= 7 999 ETTE ha"), while two rangeland sites were dominated by A.
nubica (Madalal= 6 602; Samayu= 6 984 ETTE ha"). In the remaining three-
rangeland sites, A. robusta (1 658 ETTE ha-I), A. nilotica (4 296 ETTE ha-I) and P.
juliflora (10 722 ETTE ha-I) contributed the highest ETTE values ha-I. This result
142
indicated that description of woody vegetation in terms of density alone might not be
a good indicator of the competitive behaviour of woody plants on the herbaceous
layer. Therefore, from the viewpoint of competitive behaviour, it is better to express
the values in ETTE ha" rather than density. The general conclusion is that the area is
bush encroached particularly with A. senegal, A. nubica and P. juliflora.
25.00
20.00
;? Awash- Fantale~
'" 15.00(.)
~'s" 10.00
.J:J
<t: 5.00
0.00
1 2 3 4 5 6 7 8 9 10 II 12 13 14 IS 16 17 18 19 20 21 22 23 24 25
Species number
Figure 5.3. Abundance (%) of woody vegetation in Awash-Fantale district (based upon woody
plants ha"). Key to the species: 1 = Acacia senegal; 2= Grewia bicolor; 3 = Vernonia
cinerascens; 4= Grewia tenax; 5= Acacia robusta; 6= Grewia villosa; 7= Balanites
aegyptica; 8= Rytigynia neglecta; 9= Acacia fruticosa & Acacia indica; 10= Acacia
mellifera; 11= Solanum marginatum & Solanum incanum;12= Acacia nubiea; 13=
Acacia torti/is; 14= Cordia sinensis; 15= Withania somnifera; 16= Cadaba
rotundfolia; 17= Salvadora persica, 18= Causerina equisetifolia, 19= Capparis
fascicularis; 20 = Ficus gnaphalocarpa; 21= Acacia nilotica; 22= Prosopis julifloria;
23= Cryptostegia grandiflora; 24= Erythrina abyssinica; 25= Acacia seyal.
35.0
30.0
,-.. 25.0
"...$._._, 20.0
u.1
f-< 15.0
~
10.0
5.0
0.0 II I I Iii _ II III lOO- III II1II _...~--""-
1 2 3 4 5 6 7 8 9 10111213 1415161718192021 22232425
Species
Figure 5.4. Species composition of woody vegetation based on ETTE (%) in Awash-Fantale
district. Key to the species: 1 = A. senegal; 2= G. bicolor; 3 = V. cinerascens; 4= G.
tenax; 5= A. robusta; 6= G. villosa, 7= B. aegyptica; 8= R. neglecta; 9= A. jrutieosa &
A. indica; 10= A. mellifera; 11= S. marginatum & S. incanum;12=A. nubiea; 13= A.
tortilis; 14= C. sinensis; 15= W somnifera; 16= C. rotundfolia; 17= S. persica; 18= C.
equisetfolia; 19= C. facscicularis; 20 = F. gnaphalocarpa; 21= A. nilotica; 22= P.
juliflora; 23= C. grandiflora ; 24= E. abyssinica; 25= A. seyal.
143
The pastoralists classified each of the woody species on the basis of palatability, i.e.,
highly palatable, intermediate and unpalatable. Accordingly, 51.9%, 14.8% and
33.3% were highly palatable, intermediate and unpalatable, respectively (Table 5.4).
Table 5.4. Palatability of woody plants as rated by the pastoralists in Awash-Fantale
district.
Highly palatable Intermediate Unpalatable
Acacia senegal Grewia bicolor Solanum incanum
Vernonia eineraseens Acacia robusta Withania somnifera
Grewia tenax Acacia nubica Acalypha indica
Ficus gnaphalocarpa Cadaba rotundfolia Erythrina abyssinica
Acacia mellifera Casuarina equisetfolia
Acacia tortilis Cryptostegia
grandiflora
Grewia villosa Acalypha fruticosa
Cordia sinensis Solanum marginatum
Balanites aegyptica Prosopis juliflora
Rytigynia neglecta
Salvadora persica
Acacia nilotica
Capparis fascicularis
Acacia seyal
5.3.1.2.2. Browse production
The leaf DM for the rangeland sites and for each individual woody plant that
contributed more than 1% to the total leaf DM are presented in Figures 5.5 and 5.6.
In addition, the results of the phenological observations for these species, where 1
means the highest leaf DM and 0.2 the lowest leaf DM for each of the species in the
respective rangeland site for the month of August and September (wet months) and
December and January (dry months) are presented below.
144
Lahlai
The total leaf DM for the Lahlai site was 1 056 kg ha" (Figure. 5.5), dominated by
A. senegal (842 kg ha"l, Figure 5.6) and four additional species of woody plants
contributed more than 1% to the total leaf DM, whereas the total leaf DM below 1.5
m and greater than 1.5 - 5.0 m was 340 kg ha·l and 716 kg ha"l, respectively. The
leaf phenology for the 5 species during August and September was 1.0 meaning that
the trees were full of leaves, while, A. senegal, G. villosa, R. neglecta had a value of
0.4 (low leaf abundance) during December and January. Balanites aegyptica and G.
tenax had a value ofO.6 and 0.2 (least in December and January), respectively.
o 0.0-1.5 m Ill> 1.5-5.0 m 0 >5.0 m
Lahlai Degage Madalal Top of Samayu Della kara Awash- BRS
budl Dulcha
lRangeland sites
Figure 5.5. Browse production stratified into height strata across the rangland sites
in Awash-Fantale district.
Degage
Degage had the lowest totalleafDM (489 kg ha"l) with a leafDM below 1.5 m and
greater than l.5 up to 5 m Of 155 kg ha" and 334 kg ha", respectively (Figure 5.5.).
On a leaf DM basis, A. robusta contributed the most (387 kg ha", Figure 5.6) and
four other individual woody plants contributed more than 1% to the total leaf DM.
The phenology of the leaves for G. bicolor, G. tenax and A. robusta was 1 in August
and September, while it was 0.6 in December and January. Acacia senegal had a leaf
phenological value of 1 during the wet months and 0.4 in the dry months whereas, V
eineraseens had a phenological value of 1.0 (wet months) and 0.2 in December and
January.
145
1000 .,-------------j 500.------------_,
800 Lahlai -...,400 Degage
...I::
0>1.5- 5m ~300 D>1.5-5m
~ 400 Il!!! 0.0-1.5 m ~ 200 II 0.0-1.5 m
~ 200 ~ 100
.....:l o .....:l O~--~--~,_-=~-=~._
As Gv Ba Rn Ot As Gb Vc Ot Ar
Species species
",",2000 .,------------ 2000 .,------------
] 1500 Madalal
eo .a 1500 Top of bud Ii eo1000 m >5m i 1000
Cl 0> 1.5 - 5 m D>1.5-5m
~ 500 IiiII 0.0-1.5 m
Cl
.~ 500 IlII 0.0-1.5 m
.....:l .....:l
Am As Anu Ot As SJl Ot Anu A.if
Species Species
2000 .,-----------, 1200 ,......-----------,
Samayu 1000 C>5m
dl 1500 Bulga
...I:: ~'"eo 800
0>1.5-5 m
...::<: Iiiil 0.0-1.5 m
C 1000 ElI>5m '--' 600
0>1.5-5 m
~ 400
~ 500 II 0.0-1.5 m 200
.....:l O-f-l-.L...r-J!IEl..,..IlI!!L.""'---r-"-':ói...r~....I::::::L"""",L..,--I'~
As An Cs Ws Am Ea Cr Aif ~ Ani Ce Cf As Cg
Species Species
1200 3000
"'"' r--,1000 Delia kara '7 2500 m>5m
'" ct!...I:: 0>1.5-5 m Junctioneo 800 i~2000 .. 0.0-1.5 mC 600 D>1.5-5m 1500
~ 400 II 0.0-1.5 m Cl 1000
't;l 't;l
Q)
.....:l 200 I D Q).....:l 5000 = Iii! IiiiI II1II g 0
As At GvAn Cs Am Vc Cr Ani Pj Cg
Species Species
Figure 5.6. Estimates of the leaf dry mass (kg ha") at peak biomass, with subdivision into
height strata, of woody plants in Awash-Fantale district. Key to the species:
As= senegal; Gv= G. villosa; Ba= B. aegyptica; Rn= R. neglecta; Gt= G. tenax;
Ar= A. robusta; Am= A. mellifera; Anu (An) = A. nubica; Sol= Solanum; Cs=
C. sinensis; Ws= W. somnifera; Ea= E. abyssinica; Cr= C. rotundfolia; Aif= A.
fruticosa and A. indica; Sp= S. persica; Ani= A. nilotica; Ce= C. equisetfolia;
Cf= C. fascicularis; Cg= C. grandiflora; At= A. tortilis; Vc= V cineraseens;
Pj= P. juliflora.
146
Madalal
The total leaf DM was 1 781 Kg ha" (Figure 5.5.) with 789 kg ha" and 986 kg ha"
occurring below l.5 m, and above l.5 m up to 5 m respectively, only 6 kg ha" of the
leaf DM occurs above 5 m. Acacia nubica contributed the highest to the total leaf
DM (1 529 Kg ha, Figure 5.6), while three additional species contributed more than
1% to the total leaf DM. All the species had a phenological value of 1 during August
and September, while all except Acacia tortilis had a value of 0.2 in December and
January. Acacia tortilis had a phenological value of 0.6 during December and
January.
Top of bud 1
Top of bud 1 ranked third in terms of total leaf DM (2 364 Kg ha-I, Figure 5.5) and
the leafDM below l.5 m was 660 kg ha", while that greater than l.5 up to Sm was 1
704 kg ha". Acacis senegal contributed most to the total leaf DM (1 882 kg ha",
Figure 5.6). Four additional species contributed more than 1% to the total leaf DM
and all species had a leaf phenological value of one in August and September. Acacia
senegal, so/anum species, G. villosa had a leaf phenological value of 0.4, while G.
tenax, A. mellifera, A. nubica, A. fruticosa had a value of 0.2 during December and
January. The leaf phenology of Balanites aegyptica was 0.6 for the similar time
period
Samayu
The total leaf DM was 2 343 kg ha" and the Leaf DM below l.5 m and greater than
1.5 up to 5 m was 909 kg ha" and 1 432 kg ha-I respectively (Figure 5.5), while that
greater than 5 m was 2 kg ha". Acacia nubica contributed most to the total leaf DM
(Figure 5.6.), whereas four additional species also contributed more than 1% to the
total leaf DM. All the species had the highest leaf phenological value of 1 during
August and September, while in December and January, A. senegal, A. nubica, C.
sinensis and A. mellifera had a value ofO.4. Withania somnifera and G. tenax·had the
147
lowest leaf phenological value of 0.2, while A. forti/is had a value of 0.6 for the
similar time period
Bulga riverside (BRS)
Bulga riverside had the highest number of woody plant species and poisonous woody
plants such as E. abyssinica and C. grandiflora, although, their contribution to total
leaf DM was only 196 kg ha". The highest leaf DM was contributed by A. nilotica (1
054 kg ha", Figure 5.6) and the total leaf DM was the second largest among the
rangeland sites (Figure 5.5). The total leaf DM found below l.5 m and greater than
l.5 up to 5 m was 762 kg ha"1 and 1 602 kg ha", respectively (Figure 5.6). The
Bulga riverside, also, had the highest leaf DM above 5 m (1 117 kg ha") and nine
woody species contributed more than 1% to the total leaf DM. Furthermore, this
rangeland also had the highest number of evergreen woody plants such as C.
rotundfolia, S. persica, C. equistifolia and C. grandiflora.
Della kara
The totalleafDM was 1 648 kg ha" (Figure 5.5), with a leafDM below l.5 m (436
Kg ha") and greater than l.5 up to 5.0 m of 1 212 Kg ha". On a leafDM basis, this
rangeland site was dominated by A. senegal (1 080 kg ha", Figure 5.6) and 7 other
individual species contributed more than 1% to the totalleafDM.
Awash-Dulcha (Junction)
The highest leaf DM was recorded in this rangeland site (2 833 Kg ha", Figure 5.5)
and the leaf DM below l.5 m and greater than l.5 up to 5 m was 722 Kg ha" and 1
881 kg ha"I, respectively. Prosopis juliflora contributed the highest to the total leaf
DM (2 373 kg ha"I, Figure 5.6) and there were only three woody species in this
group, which contributed more than 1% to the totalleafDM (Figure 5.6).
148
5.3.1.2.3. Browsing capacity (BC)
In the study districts, the pastoralists own camels and goats that are browsers. The
browsing height for goats is usually up to 1.5 m, while camels can reach up to 3.5 m.
The leaf phenology for the months of August, September, December and January
were obtained from field measurement, while for the rest of the months, it was
estimated based upon the experience of the pastoralists and the rainfall pattern of the
area that has a bimodal nature. In most of the rangeland sites, the woody plants were
deciduous, indicating that the browsing capacity (BC) varies from season to season
in accordance with leaf flushing and leaf senescense. For ease of presentation, the
year was divided into two broad categories, wet months (WMS) i.e., July, August,
September, March, April and May and dry months (DMS) October, November,
December, January, February and June. Accordingly, the results of the approximate
BC for each of the rangeland sites and for those species that contributed most to the
totalleafDM is presented below.
Degage
The BC, expressed as ha BUl, for the height below 1.5 m, during the wet and dry
months, was 17.14 and 25.80 respectively, while for the height greater than 1.5 up to
5 m was 7.97 to 12.00 for the wet and dry months, respectively (Figure 5.7). In both
seasons and for both heights, the BC of A. robusta was higher than for others species
(38.33ha BUl wet months for the height below 1.5m; 52.34 ha BUl for the height
30.0
25.0 I D 0.0-1.5 m WMS r:::l 0.0-1.5 m DM S A wash-Fan taleop 20.0 0>15-50 m WMS III > 1 5-5 m DM SIII...'_c":, 15.0
u 10.0
III
5.0 I
I~""0.0
Rangeland sites
Figure 5.7. Browsing capacity (ha BUl) across the rangeland sites in Awash-
Fantale district.
149
00.0- 1.5 m WMS III0.0- 1.5 m DMS 00.0-1.5 m WMS I!Ii 0.0-1.5 m DMS
0>1.5- 5.0 m WMS Cl >1.5-5.0 m DMS 0>1.5-5.0 m WMS El >1.5-5.0 m DMS
1
500 ,------=----------, 800 ~---------~
Degage '";'::> 600 Della kara
~ 400
8' 200
o:l O-+="--,.LJB..Ii:l.,l-IIII-"ly::tlLU.,.J..IIIII:a,..<:IIII_..,UI1I..U..j
;> ::l Vl E ....
As Gb Gt Ar ti ~ U <t: U
Species Species
00.0-1.5 m WMS !lIII 0.0-1.5 m DMS I 00.0-1.5 m WMS III 0.0-1.5 m DMS
I 0>1.5-5.0 m WMS IJ >1.5-5.0 m DMS 0>1.5-5.0 m WMS [J >1.5-5.0 m DMS
1000.,.------------,
r-.
~ :~~I ~ j I b 800 Madala I600~ ;~~_- I1l.rlU Lahlai ,~,Jl Jl . CO..O._c:l:, 400UCO 2000
As Gv Ba Vc Rn Gt Am As Sol Anu Gv
Species Species
DO.0-l.5mWMS 1!!I0.0-l.5mDMS 00.0-1.5 m WMS III 0.0-1.5 m DMS
0>1.5-5.0 m WMS Cl >1.5-5.0 m DMS 0>1.5-5 m WMS rn >1.5-5 m DMS
400 .-----------------------~
Samayu
As At Anu Ws Am As Sol Gt Am Anu Aif
Species Species
00.0-1.5 m WMS III 0.0-1.5 m DMS 00.0-1.5 m WMS IiII 0.0-1.5 m DMS
D>1.5-5.0m WMS m>1.5-5.0mDMS 0>1.5-5.0 m DMS r;:] >1.5-5.0 m DM
200.0 -,--------------------
r- 800 ,-----------------------------~
-; 150.0 Awash Dulcha -;""6' 00
:::::> Bulga riverside
';; 100.0 ';; 400
.g.c:: ..c::50.0 U200
CO
0.0
As Pj Cg ·Cr Aif Sp Ce Cf As Gt Fn As Cg
Species Species
Figure 5.8. Browing capacity (ha BU·I) for important species based upon their
contribution to browse production in Awash-Fantale district. Key to the species:
As= A. senegal; Gb= G. bicolor; Gt= G. tenax; Ar= A. robusta; Gv= G. villosa;
Anu = A. nubica; Cs = C. sinensis; Arn= A. mellifera; Vc= Veineraseens;
Cr- C. rotundfolia; Ba= B. aegyptica; Rn= R. neglecta; Sol= Solanum; At = A.
tortilis; Ws = W somnifera; Aif= A. frutieosa and A. indica; Pj= P. juliflora; Cg=
C. grandiflora; Sp= S. persica; Ce= C. equistefolia; Cf= C. fascicularis; Fn= F
gnaphaloearpa.
150
below 1.5m dry months; 11.67 ha BUl for the height greater than 1.5 up to 5 m wet
months; 15.94 ha BU-I for the height greater than 1.5 up to 5 m, dry months), which
was a direct reflection of the amount of leaf DM of the plant. This implied that less
land is required to sustain a browser unit on A. robusta than on the other woody
species (Figure 5.8).
Lahlai
The BC (ha BUl) at Lahlai was lower for the height below 1.5m than the height
greater than 1.5 up to 5 m, which indicated that the leaf DM up to 1.5m was lower
than the leaf DM for the height greater than 1.5 up to 5 m (Figure 5.7). Across the
year and for the heights of 0 -1.5 m and greater than 1.5 up to 5 m, A. senegal made
the biggest contribution to the BC, making it the most important source of browse for
the area (Figure 5.8).
Madalal
The BC (ha BUl) over the year, for the height below 1.5 m, varied between 3.17 to
5.19 ha BUl, while that of the height greater than 1.5 up to 5 m varied between 2.01
ha BUl (Wet months) to 3.29 ha BU-I (Dry months) (Figure 5.7). Acacia nubica
contributed the most to the browsing capacity across the year and for both heights
(5.0 ha BUl for the height below 1.5 m wet months; 9.8 dry months and 4.1 ha BUl
in wet months for the height greater than 1.5 up to 5 m; 8.1 ha BU-I for the height
greater than 1.5 m in the dry months), which was a direct reflection of the high leaf
DM (Figure 5.8).
151
Top of Bud I
The BC (ha BUl) across the rangeland site and during the wet and dry season for the
height below 1.5 m was 4.37 ha BUl and 7.54 ha BUl respectively, while that of the
height greater than 1.5 up to 5 m for the wet and dry months was 1.69 ha BU-l and
2.92 ha BUl, respectively (Figure 5.7). Acacia senegal contributed the most to BC,
i.e., 9.77 (wet months) to 15.68 (dry months) ha BUl for the height below 1.5 m and
1.40 (wet months) to 2.25 (dry months) for the height greater than 1.5 m up to the
height of Sm (Figure 5.8). This result indicated that, most of the leaf DM was found
above the height of 1.5 m favoring camel browsing rather than goats.
Samayu
The BC in the rangeland site and for the wet and dry seasons for the height below 1.5
m was 3.17 and 5.19 ha BUl respectively, while for the height greater than 1.5 up to
5 m for the wet and dry months was 2.01 and 3.29 ha BUl, respectively (Figure
5.7). The BC (ha BUl) was the highest for A. nubica (Figure 5.8, 5.71 to 9.16 ha
BUl for the height below 1.5 m and 3.33 to 5.35 ha BUl for the height above 1.5 up
to 5 m).
Awash-Dulcha junction
In this rangeland site, the BC below 1.5 m was 4.56 ha BUl for the wet months,
while for the dry months it was 4.71 ha BUl. The BC (ha BUl) for the height from
greater 1.5 up to 5 m was 1.70 (wet months) and for the dry months, it was 1.76 ha
BUl. Season did not have a significant influence on the BC, as this area has the
evergreen woody plants P. juliflora (dominant) and C. grandiflora. The browsing
capacity for P. juliflora in the wet and dry months for the height below 1.5 m was
4.78 and 4.86 ha BUl respectively, while for the height greater than 1.5 up to 5 m in
the wet and dry months was 1.83 and 1.86 ha BUl, respectively (Figure 5.8).
152
Bulga riverside (lBRS)
The BC along the Bulga riverside for the height below 1.5 m was 3.83 ha BUl (wet
months) and 5.09 ha BUl (dry months), while that of the height greater than 1.5 up
to 5 m was 3.14 ha BUl (wet months) and 4.17 ha BUl (dry months) (Figure 5.7).
Cadaba rotundfolia and A. senegal had the highest BC (ha BUl) over the year
(Figure 5.8) and E. abyssinica, C. rotundfolia, S. persica, E. equistefolia and C.
grandiflora were evergreen species which maintained their leaves throughout the
year (non-deciduous).
Della kara
The BC was higher for A. senegal during the wet and dry seasons for both heights
(17.01 ha BUl and 27.31 ha BUl during the wet and dry months for the height
below 1.5 m, respectively) and (2.45 ha BU-I and 3.92 ha BUl during the wet and
dry months for the height greater than 1.5 up to 5 m, respectively) (Figure 5.8). In
this rangeland site, the BC for the height below 1.5 m for the wet and dry months
was 6.24 and 9.08 ha BU-I respectively, while for the height greater than 1.5 m up to
5 m was 2.25 (wet months) & 3.53 ha BUl (dry months).
5.3.1.3. Soil Parameters
The soil of the studied rangeland sites had a pH values ranging between 8.2 and 10.4
with a mean value of 8.62. The pH was highest at Asobabad and lowest along the
Bulga riverside (Table 5.5). The electrical conductance (EC) (ds m") varied from
0.096 to 3.69 with a mean value of 0.573. The value was highest at Asobabad and
lowest at Top of bud.
The percentage sand was the highest at Asobabad and the lowest along Samayu with
a mean value of37.70% indicating that the proportion of the sand is higher than that
of the other components (Table 5.5). Across the rangeland sites, the proportion of
silt (mean value of 36.44%) was higher than the clay (25.88%). The percentage total
153
nitrogen (N) was the lowest at Asobabad, while in the rest of the sites, the percent
nitrogen ranged from 0.103 to 0.184. The percentage organic carbon across the
rangeland sites varied from 0.187 to l.214 with a mean value ofO.92. Similar to the
% N, it was the lowest at Asobabad, while the ratio of carbon to nitrogen (eN ratio)
varied from 6.00 to 10.00. The available phosphorus (ppm) ranged from 2 to 1l.38
and the available potassium (ppm) varied from 201 to 421 with a mean value of
33l.56.
Table 5.5. Physical and chemical characteristics of soils of the rangeland sites in Awash-Fantale
district.
Range Soil texture Other soil parameters
Land
site
Sand Silt Clay PH: EC N OC CN Available Availabl
(%) (%) (%) (ds (%) (%) ratio P (ppm) e
m") K (ppm)
Degage 35± 40± 25± 8.3± 0.l68± 0.147 1.17± 8± 4.7± 409±
3.l 0.6 3 O.l 0.02 ±O.04 0.03 0.04 0.6 58.0
Madala 34± 4l± 25± 8.32± 0.l56± 0.l74 0.962 6± 4.47± 343±
8.02 7.02 2.08 0.02 0.01 ±O.06 ±O.05 1.7 1.7 40.5
Madala 37± 37± 26± 8.36± 0.l57± 0.l84 1.105 7± 6.l0± 383±
1 5.0 3.0 3.0 0.04 0.01 ±O.04 ±O.l 1.5 3.4 82.1
Top of 45± 37± l8± 8.37± 0.096± 0.112 0.886 8± 4.30± 276±
bud 3.06 5.8 6.0 0.3 0.0 ±O.02 ±O.2 0.6 1.9 98.1
Asobab 64± 23± B± 10.4± 3.69± 0.03± 0.187 7± 7.07± 342±
ad 1.5 1.00 1.15 0.21 0.10 0.01 ±O.OI 0.50 4.10 40.55
Samay 2l± 30± 49± 8.73± 0.265± 0.116 1.071 10± 3.6± 257±
u 1.5 4 5.l 0.06 0.03 ±O-.Ol ±O.l 0.5 0.8 66
Top of 44± 40± 16± 8.5± 0.096± 0.103 0.772 8± 4± 201±
budl 2 0.0 2 0.02 0.0 ±O.04 ±O.06 0.5 0.2 25.5
Bulga 25± 50± 25± 8.21± 0.29± 0.l19 0.90± 8± 11.38± 352±
river 0.7 1.7 5.5 0.1 O.l ±O.03 0.3 0.6 12.1 47.04
side
Della 34± 30± 36± 8.4± 0.243± 0.l35 1.214 9± 2± 42l±
kara 6.11 12.06 6.0 O.l 0.05 ±O.05 ±O.42 0.0 0.2 121
154
5.3.1.4. Correlation matrix among the studied variables
Among all variable correlated, the results of relationship between grass yield, basal
cover and ETTE with other variables was presented in this section. The summary of
the correlation matrix is shown in Appendix 5.3.
Among all correlated variables, grass yield was non-significantly (P>0.05) related
with bare ground, estimated soil erosion and significantly (P<O.00 1) negatively
correlated with ETTE (P < 0.001). The latter correlation between the yield of grass
and ETTE was very strong (r= -0.65) which impies that there was a decrease in the
DM yield of grasses with an increase in ETTE ha" value. There was a strong positive
correlation (P < 0.01) between the DM yield of grass and basal cover, while basal
cover was negatively correlated with bare ground (P<O.Ol) and ETTE ha-1 (P<0.05).
The ETTE ha" values were negatively correlated with sand content (P<O.Ol) and eN
ratio (P<0.05).
5.1.3.5. Multiple regression between grass :DM yield and other parameters
The rate at which rangeland is stocked is probably the single most important
management factor affecting animal performance and the profitability of a livestock
system (O'Reagain & Turner 1992). Stocking tate is one of the variables under direct
control of the manager, while the chosen stocking rate largely determines the degree
of interaction between the grazing animal and the vegetation, which, in turn,
determines production per animal and animal production per hectare (Morris et al.
1999). However, in a pastoral community land is owned communally, where
decisions regarding rangeland resources are made collectively. In such a system, it is
difficult to establish the stocking rate for the studied rangeland sites since
information on animal numbers was not available. Nevertheless, in light of the
results presented in the correlation matrix and taking the normal restrictions with
regard to the inclusion of correlated variables in multiple regression functions into
account, a series of stepwise multiple regression analyses were conducted with grass
DM yield as the dependent variable in order to establish which factor or factors,
155
among the measured environmental variables, other than the stocking rate, can be
used as an indicator of the variability in grass DM yield. In agreement with the result
of the correlation matrix, the stepwise regression analyses showed that ETTE ha-!
and basal cover were the most important factors affecting grass yield. Forty two
percent of the variation in the grass yield was explained by ETTE ha-! and the
remaining Il % by basal cover (Table 5.6). The result suggested that other
parameters like stocking rate needs to be included for a more complete explanation.
Table 5.6. Standard coefficient, rvalues, R2 increment (%), R2 values (%) and
significance of multiple regression with dry matter yield as a dependent
factor (N= Nitrogen; OC= Organic carbon; CN ratio = Carbon nitrogen
ratio; P= Phosphorous; K= Potassium; NS= Nonsignificant; *= P<0.05;
**= P<O.Ol)
Variables Standard r value R2 increment Rl signifi
coefficient (from (%) values -cance
correlation (%)
matrix)
ETTE -0.4742 -0.65 17.48 42.15 **
Basal cover 0.3709 0.59 10.69 52.84 *
Bare ground -0.25 0.05 NS
(%)
Estimated -0.31 0.11 NS
Soil erosion
PH 0.13 0.22 NS
Electrical 0.16 1.03 NS
conductivity
% sand 0.32 0.04 NS
% silt -0.02 0.50 NS
% clay -0.38 0.13 NS
% total N 0.01 0.00 NS
%OC -0.05 0.57 NS
CN ratio -0.16 0.99 NS
Available P -0.14 2.20 NS
Available K 0.06 0.93 NS
Altitude 0.33 1.54 NS
156
5.3.2. Kereyu-Fantale district (Oromo)
5.3.2.1. Herbaceous layer
5.3.2.1.1. Grass species composition
The most dominant grass species across the rangeland sites was C. plumulosus,
which was followed by S. natalensis. S. spicatus, C. dactylon, H. contortus, C.
commutatus, D. aegypticum and C. poiflorum, while the rest of the species were less
abundant (Table 5.7). Based upon desirability, 23.68%, 18.42% and 57.9% were
highly desirable, desirable and less desirable, respectively whereas, when expressed
in terms of life form, 34.21% were annuals and 65.79 % were perennials. Forbs,
generally comprised 4.14% of the composition and varied in their abundance from as
low as 0.32% (Chopi) to as high as 14.53% (Mogassa) across the rangeland cites.
In five of the rangeland sites, the most dominant grass species was C. plumulosus
(Figure 5.9), while at Harole1, C. commutatus (18.10%) and D. aegyptcium
(15.90%) were the most abundant. Kara was dominated by S. natalensis (31.10%),
while Mogassa by D. aegypticum (17.02%) and C. dactylon (16.78%). Grasslands
along Lake Beseka were dominated by S. spicatus. In addition to C. plumulosus, at
Fantale near the crater, S. natalensis (21.27%) and T. triandra (8.51%) were found in
higher abundance than the other species.
5.3.2.1.2. Bare ground
The percentage bare ground (Mean ± SD), as determined by the point method, varied
between 0.33 to 9.56 with a mean value of 5.02 (Table 5.8), which clearly indicated
loss of soil from the surface. In six of the rangeland sites, the percent bare ground
was greater than 5%, while it was less than 5 in the others and it was lowest at
Kolbayu (0.33%) and highest at Mogassa (9.56%).
157
'fable 5.7. Desirability, life form and frequency (%) of the different grass species in
the Kereyu-Fantale district.
Species Desirability Life form Frequency (%1
Chrysopogon plumulosus Highly desirable Perennial 30.37
Sporobolus natalensis Less desirable Perennial 12.10
Sporobolus spicatus Less desirable Perennial 8.37
Cynodon dactylon Less desirable Perennial 5.85
Heteropgon contortus Desirable Perennial 5.51
Cjlmbopogon commutatus Desirable Perennial 5.34
Dactyloctenium Less desirable Annual 5.06
aegypticum
Coelachyrum poiflorum Less desirable Annual 3.73
Tetrapogon cenchriformis Highly desirable Perennial 2.40
Paspalm glumaceum Desirable Perennial 1.84
Setaria verticellata Less desirable Annual 1.69
Cenchrus ciliaris Highly desirable Perennial 1.46
Themeda triandra Highly desirable Perennial 0.90
Chloris roxburghiana Desirable Perennial 0.81
Eleusine indica Less desirable Annual 0.55
Cymbopogon excavatus Highly desirable Perennial 0.50
Pasapalm dilataum Less desirable Perennial 0.48
Panicum coloratum Highly desirable Perennial 0.47
Cenchrus setigrus Highly desirable Perennial 0.47
Panicum snowdenii Less desirable Annual 0.37
Urochloa panicoides Less desirable Annual 0.33
Digitaria milanjiana Highly desirable Perennial 0.31
Tragus berteronianus Less desirable Annual 0.30
Digitaria ternata Less desirable Annual 0.30
Eragrostis cilianensis Less desirable Annual 0.28
Aristida adoensis Less desiable Perennial 0.22
Sporobolus festivus Less desirable Perennial 0.20
Pennisetum mezianum Less desirable Perennial 0.15
Lintonia nutans Desirable Perennial 0.11
Chloris gayana Highly desirable Perennial 0.10
Eragrostis racemosa Desirable Perennial 0.08
Bothriochloa radicans Less desirable Perennial 0.06
Microchloa kunhtii Less desirable Annual 0.04
Hyparrhenia hirta Desirable Perennial 0.03
Sorghum Less desrabie Annual 0.02
purpureosericeum
Aristida adscensionis Less desirable Annual 0.02
Enneapogon schimperanus Less desirable Perennial 0.01
Setaria ustilata Less desirable Annual 0.01
158
Chopi lil T. cenchriformis Mount fantale ~ T. triandra
D S. natalensis liJ S. natalensis
IlD. aegypticum lEIE. indica
D C. plumulosus
D C. dacty/on
5.04% 8.51% D H. contortus
Iii! C. plumulosus
Harolel EI S. natalensis Il:] S. natalensis
El D. aegyptiu m Harole [3 C. poiflorum
DC. commutatus D C. plumulosus
El C. ciliaris 0.145
Il:! C. excavatus
~007
0.667
Kara DS. nata/ensis Kolbayu DS. nata/ensis
~ S. verticellata DC. Poiflorum
8.90% SH. contortus DC. P/umu/osus
Cl C. dactylon El C. Roxburghiana
lEIC. poiflorum D P. cotoratum
31.10%
9.20%
Til:! s. natalensis Tuludimitu D C. commutatus
0.069 ID C. plumulosus 8 C. plumulosus
(T\ f!lI H. contortus
\__)
0.86
38.0%
Lake beseka rEI C.dactylon I Mogassa a S.nata/ensis
I El P.g/umaceumD S. spicatus D D.aegyptium
El C.dacty/on
5.71%
IIIll A.aspara
83.81%
Figure 5.9. The most dominant grass species in each of the rangeland sites in
Kereyu-Fantale,
159
5.3.2.1.3. Basal cover
The result of the percentage basal cover (mean ± SD) followed a pattern opposite to
that of the percentage bare ground (Table 5.8). It was the highest on the rangeland
site where the percentage bare ground was the lowest (Kolbayu) and lowest in the
site where the percentage bare ground was the highest (Mogassa). Of the rangeland
sites, 50%, 30% and 20% had a cover value ranging between 2-3, 3-4 and 4-5
respectively, which implied that in most of the rangeland sites the cover of the
vegetation was low.
5.3.2.1.4. Grass DM yield
The yield of the grass layer (mean ± SD) ranged between 168.52 kg ha-1 to 807.69 kg
ha" with an overall mean of 42l.92 kg ha" (Table 5.8). Similar to that of the result
in Awash-Fantale district, both the highest (Kolbayu) and the lowest yields
(Mogassa) were recorded in rangelands with woody vegetation.
5.3.2.1.5. Estimated soil erosion
The estimated soil erosion values (mean ± SD) ranged between 2 to 4 indicating that
there was a loss of soil due to erosion (Table 5.8). Of the ten rangeland sites in the
district, 10%, 60% and 30% of the rangelands had values of 2,3 and 4 respectively,
implying 10% of the studied rangeland sites had steep sided pedestals, 60% slope
sided pedestals and 30% slight sand mulch.
5.3.2.1.6. Grazing capacity
The grazing capacity (mean ± SD) based upon ha LSU-1, declined most at Mogassa
(54.15) and was better at Kolbayu (11.15) than at the other rangeland sites, which
implies that more land is required to sustain a LSU at Mogassa than at Kolbayu, i.e.,
a direct reflection of their grass production (Table 5.8). The pattern was similar for
TLU (declined most at Mogassa = 33.84 ha TLU-1 and best at Kolbayu = 7.06 ha
160
TLU1). In general, the mean grazing capacity for the studied rangelands based upon
LSD and TLD were 24.83 ha LSU1 and 14.5 ha TLUI, respectively.
Table 5.8. Mean and standard deviation of bare ground (%), basal cover (%), grass
DM yield (kg ha") and estimated soil erosion values for rangeland sites in
Kereyu-Fantale district (HL= Herbaceous layer; WL= Woody layer).
Rangelands
Parameters A1eka Harol Beseka Chopi Mogassa Kara Kolba Tulu Fantale Harol
e -yu dimtu e-l
Vegatation fIT, fIT, fIT, fIT, fIT, fIT, fIT, fIT, RB fIT,
layer studied & & & & & &
WL WL WL WL WL WL
Bareground 5.8± 5± 8.2± 9.4± 9.6 5.9± 0.3± OA± 3A± 2.1±
(%) 0.7 0.7 1.6 0.3 ±l.l 1.1 0.2 0.0 2.1 1.9
Basal cover 2.6± 2.7± 2.7± 2.5± 2.4 3± 4.7± 4.6± 3.5± 3.5±
(%) 0.7 0.4 0.8 0.5 ±G.4 0.4 1.5 0.5 0 0.1
Grass yield 391.9± 402± 350± 284.3± 168.5± 353.5± 807.7± 467.7± 616± 377.6±
(kg ha") 102.9 38.9 76.7 22.1 76 38.7 67.5 26.6 20.4 82.9
Estimated soil 2± 3± 3± 3± 4± 3± 4± 3± 4± 3±
erosion 0.0 0.5 0.6 0.6 0.6 0.6 0.0 0.0 0.6 1.0
Grazing 23.3± 22.7± 26.07± 32.1± 54.2± 25.8± 1.2± 17.1± 11.8± 24.2±
capacity 6.6 2.3 6.8 2.6 40.0 3 1.0 0.8 0.3 4.8
(ha LSU-I)
Grazing 13.6± 13± 14.7± 19.6± 33.8± 16.1± 6.97± 10.6± 7A± 14.7±
capacity (ha 4.2 1.4 4.2 1.6 25.0 1.9 0.6 0.5 0.2 3.1
TLUI)
5.3.2.1. 7. Rangeland condition
The condition of the rangeland sites based upon the-results of the point survey, varied
between 37.5% (Lake Beseka) to 90.3% (Aleka), while the mean rangeland condition
of all sites was 63.42%. The conditions of the rangeland sites was above the mean in
50% of them and lower than the mean in 50% of the rangeland sites.
161
100 ~;lU.j
~,90 80.347928
~ 80 . 75.3472.29
lO:
.2 70
.<:::
-g 60 53.66 52.08 50.8
8 50 42.56
-g 37.540
t<l
~eo 30
lO: 20
& 10
o
Figure. 5.10. Rangeland condition of the rangeland sites in Kereyu-Fantale district
based upon desirability of species.
5.3.2.2. Woody layer
5.3.2.2.1. Woody vegetation composition, density, lETTE and palatability values
A total of 34 woody species were identified in the transects laid out for the sites
(Figure 5.11). Similar to that of the Mars, two species of Acalypha were recorded,
i.e., A. jruticosa and A. indica. The Solanum species consisted of S. incanum and S.
marginatum. There were also other species that did not fall in the transects, but were
found in the field, which actually increase the number of woody species in the area.
The most abundant (%) woody species in decreasing order were A. nubica, A.
senegal, V cineraseens, Acalypha species, G. villosa, G. bicolor and G. tenax.
Important tree species such as A. nilotica, B. aegyptica and A. tonilis were less
abundant (Figure 5.11).
The density of woody vegetation was highest at Mogassa (4 900 plants ha·l) and
lowest at Tulu dimitu (600 plants ha-I). When the tree density data was expressed in
terms of ETTE ha-I, the value ranged between 967 ETTE hii-I (Tulu dimitu) to 14
178 ETTE ha-I at Mogassa. When the data was expressed in terms ofETTE (%), six
162
woody plants were most important. These in a decreasing order are A. senegal, A.
nubiea, A. tortilis, G. bieolor, G. villosa and A. mellifera (Figure 5.12).
12345678910111213141516171819202122232425262728293031323334
Species
Figure 5.11. The density of woody plants expressed in abundance (%) in Kereyu-
Fantale district. Key to the species: 1= R. neg/ecta; 2= A. mellifera; 3= G.
tenax; 4= G. villosa; 5= So/anum species; 6= A. nilotica; 7= B. aegyptica; 8= A.
senegal; 9= V einerasens. 10= C. sinensis; 11= C. fascicularis; 12= S. perisca; 13= A.
fruticosa, 14= D. cinerea; 15= L. trifolia, 16= A. tortilis; 17= G. bicolor; 18= A.
nubica; 19= C. ova/is; 20= P .oranatus, 21= Fgnaphalocarpa; 22= C. rotundfolia;
23= B. aethiopum; 24= A. robusta; 25= R. natalensis. 26= B. oleoides; 27= R.
commaanis; 28= P. africanum; 29= D. afromontana; 30= T. brownii; 31= C. africana;
32= A. seyal; 33= 0. rochetina; 34= A. xanthoph/oea.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1920 21 22 2324 25 2627 28 29 30 31 3233 34
Species
Figure 5.12. Woody vegetation expressed in terms of % ETTE in Kereyu-Fantale
district. 1= R. neglecta; 2= A. mellifera; 3= G. tenax; 4= G. villosa; 5= So/anum
species; 6= A. nilotica; 7= B. aegyptica; 8= A. senegal; 9= V cinerasens; 10= C.
sinensis, 11= C. fascicularis; 12= S. persica; 13= A. fruticosa; 14= D. cinerea; 15= L.
trifolia, 16= A. tortilis; 17= G. bicolor; 18=A. nubica; 19= C. ova/is; 20= P. oranatus;
21= F gnaphalocarpa; 22= C. rotundfolia; 23= B. aethiopium; 24= A. robusta; 25=
R. natalensis; 26= B. oleoides; 27= R. commaanis; 28= P. africanum; 29= D.
afromontana; 30= T. brownii; 31= C. africana; 32= A. seyal; 33= 0. rochetina; 34=
A. xanthoph/oea.
163
The ETTE ha" value was the highest at Mogassa (14 178 ETTE ha") and lowest at
Tulu dimitu (976 ETTE ha"). In two of the rangeland sites, A. senegal (Chopi = 6
214 ETTE ha"; Mogassa = 7 414 ETTE ha") contributed the highest ETTE ha-I
value, while A. tortilis had the highest in two of the rangelands (Kolbayu= 847 ETTE
ha-I; Tulu dimitu= 335 ETTE ha-I). At Fantale, Olinia rochetina (684 ETTE ha")
had the highest ETTE ha", whereas, at Kara Grewia bicolor (642 ETTE ha") and A.
nubica at Harolel (6604 ETTE ha").
Of the identified woody species, 51.4 % were highly palatable, 18.9% intermediate
and 29.7% unpalatable based upon the perception of the pastoralists (Table 5.9)
Table 5.9. Palatability of woody plants as rated by the pastoralists in Kereyu-Fantale
district.
Highly palatable Intermediate UIllpalatabie
Acacia Senegal Grewia bicolor Solanum incanum
Vernonia eineraseens Pygeum africanum Withania somnifera
Grewia tenax Acacia robusta Acalypha fruticosa
Ficus gnaphalocarpa Acacia nubica Plectranthus oranatus
Acacia mellifera Cadaba rotundifolia Rhus natalensis
Acacia torti/is Commiphora africana Olinia rochetina
Grewia villosa Barbeya oleoides
Cordia sinensis Ricinus communis
Balanites aegyptica Terminalia brownii
Rytigynia neglecta Dracaena afromontana
Salvadora persica Acalypha indica
Acacia ni/otica
Capparis fascicularis
Acacia seyal
Acacia xanthoploea
Dichrostachys cinerea
Lanthana trifolia
Borassus aethiopum
Cordia ovalis
164
5.3.2.2.2. Browse production
Mogassa
The totalleafDM was the highest (3 311 kg ha"l) at Mogassa, with a leafDM below
1.5 m and greater than 1.5 up to 5 m of 1044 and 2 009 kg ha" respectively (Figure
5.13). Of the total leaf DM, 258 kg ha"1was found above 5 m and on a leaf DM
basis, Mogassa was dominated by A. senegal (1746 kg ha"l) and seven other
individual woody species contributed more than 1% to the total leaf DM (Figure
5.14). All species had a leaf phenological value of 1 (maximium leaf abundance)
during August and September, whereas, during December and January L . trifolia, V.
cinerascens, R. neglecta, A. fruticosa, A. nubica, and G. villosa had the lowest
phenology of 0.2. Grewia tenax and A. senegal had a leaf phenology of 0.4, while B.
aethiopum, B. aegyptica, G. bicolor andA. tortilis had a value ofO.6.
Kolbayu
The totalleafDM was 546 kg ha" (Figure 5.13), dominated by A. tortilis (201 kg ha"
I, Figure 5.14). Of the totalleafDM, 199,345 and 2 kg ha" (Figure 5.13) was found
below 1.5m, greater than 1.5 up to 5 m and greater than 5 m respectively, while 9
additional woody plants contributed more than 1% to the total leaf DM (Figure
5.14). In August and September all the species had the highest leaf phenology of 1, 1
1111 0.0-1.5 m !ill >1.5-5.0 m 0 >5.0 m I
3500
~ 3000
J::::
-Cl 2500~ 2000~
C-l 1500ct! 1000
Q)
_I 500
0
~....
~o
CJ
Figure 5.13. Total and stratified leafDM (kg ha"l) of the rangeland sites in Kereyu-
Fantale district.
165
2000 2000 r-------------------------~
"...,
-t=.s Chopi Harolel1500 11500
00
C.
~ 1000 i
00
1000
0....
t.s 500 8 500
II) II)
....J ...'."J
0 o
As o- Fg Ot Am Gb An Cs Ba o- Cr
Species Species
200
"...,
[13>5.0 m Fantale Kolbayu 0 >1.5-5.0 m-t=.s ISO :::: 0>1.5-5.0 m Bil 0.0-1.5 m
..0..:0 II 0.0-1.5 m
~ lOO ~~~~~
8
t.s 50
II)
....J
0
Az Rn Or Bo Pa Dc Tb Ca Ase Da As It At Gb Gv Ot Aif An Ba Po
Species Species
2000.-------------------------~ 80.---------------------------~1
Tulu dinitu 0 >1.5- 5.0 m
~ 1500 ~ro60 III 0.0- 1.5 m
s s:i o >5m ~1000 [13>1.5m-5m ~40
o III < 1.5 m o2
'!ii 500
Q) I (ij 20Q)....J ....J
o
Vc Aif Gt Anu Gv Gb At As As At Gb Ba Aif
Species Species
200 ,-------------------------------------------,
Kara
~ 150
..0..:0
~100
o
~ 50
II)
....J
Ani Ba ~ ~ ~ cr ~ M ~ h M Gb
Species
Figure 5.14. Estimates of the leaf OM (kg ha") at peak biomass, with the subdivision into height strata, of woody
plants contributing more than one percent to the total leaf DM in each of the rangeland sites. Key to
the species: As= A. senegal; Gv= G. villosa; Fg= F. gnaphalocarpa; Gt= G. tenax, Am= A.
mellifera; Gb= G. bicolor; Anu= Ainubica, Cs= C. sinensis, Cr= C. rotundfolia; Ba= B. aegyptica;
m= R. neglecta; It= L. trifolia; Ax= A. xanthophloea, Rn= R. neglecta; Or= 0. rochetina; Bo= B.
oleoides; Pa= P. africanum; De= D. cinerea; Tb= T. brownii; Ca= C. africana; Ase= A. seyal; Da=
D. afromontana; At= A. tortilis; Aif= A. indica and A. fruticosa; Po= P. oranatus, Vc= V
cinerascens, Ani= A. nilotica; Cf= Ci fascioularis; Ce= C. equistefolia; Af= A.fruticosa;
166
while in December and January the species had a variable leaf phenology. Acacia
senegal and G. villosa had a value of 0.4 whereas, L. trifolia, species of Acalypha
and V eineraseens had the lowest phenological value (0.2). With regard to the other
species, G. bicolor, G. tenax, D. cinerea, Solanum species, B. aegyptica, P. oranatus
and C. ovalis had a value of 0.6. Acacia tortilis had a better leaf phenology of 0.8 in
the dry season.
Fantale
The total leaf DM was 638 kg ha-I, with a leaf DM below l. 5 m, greater than l. 5 up
to 5 m and greater than 5 m of 192, 379 and 67 kg ha-I, respectively (Figure 5.13)
and ten individual woody species contributed more than 1% to the total leaf DM
(Figure 5.14). Olinia rochetina (152 kg ha"), A. seyal (117 kg ha-I) and R. natalensis
(114 kg ha") contributed most to the total leaf DM (Figure 5.14). The leaf
phenology in August and September was 1 for all species and the species in this
rangeland had better phenological values during December and January than the
species in the other rangeland sites. Rhus natalensis, 0. rochetina, B. oleoides, R.
communis, P. africanum and D. afromontana had a phenological value ranging
between 0.8 and l. The leaf phenology of G. bicolor, A. robusta, A. xanthophlea, D.
cinerea, C. africana and A. senegal was 0.4. Terminalia brownii and A. tortilis had a
value ofO.6.
Kara
The total leaf DM was 658 kg ha" (Figure 5.13) with the highest leaf DM
contributed by G. bicolor (143 kg ha") and followed by A. tonilis (130 kg ha")
(Figure 5.14). Of the total leaf DM, 386 kg ha-I was found below l.5 m, while the
remaining 272 kg ha-I above l.5 m up to the height of 5 m. Twelve individual woody
plants contributed more than 1% to the total leaf DM (Figure 5.14). The leaf
phenology of R. neglecta, G. villasa, Solanum species, A. nilotica, B. aegyptica, C.
rotundfolia and D. cinerea was 0.4 in December and January, while that of G. tenax,
S. perisca, A. tortilis and G. bicolor had a value of 0.6. Acacia mellifera, A. senegal,
167
V cinerascens, C. sinensis, species of Acalypha, L. trifolia had the lowest leaf
phenology (0.2) during December and January. During the wet months (August and
September) all species had a leaf phenology of 1 (at peak biomass).
Chopi
The total leaf DM was 2 692 kg ha", with a leaf DM below l.5 m and greater than
l.5 up to 5.0 m of 1 039 kg ha" and 1 653 kg ha-I respectively (Figure 5.13). The
highest leafDM was contributed by A. senegal (1 461 kg ha-I, Figure 5.14) and five
additional woody plants contributed more than 1% to the total leaf DM. The leaf
phenology, for all species in August and September was 1, while it was lowest (0.2)
for A. senegal, G. villosa, V cinerascens, G. tenax and A. mellifera. Cadaba
rotundfolia had the highest (0.8) and Solanum and G. bicolor (0.4). Ficus
gnaphalocarpa had a value ofO.6 during the dry months.
Tulu dimitu
The lowest leafDM was recorded at Tulu dirnitu (194 kg ha-\ with 122 kg ha" and
72 kg ha" was found below l.5 m and greater than l.5 up to 5.0 m, respectively
(Figure 5.13). Of the total leaf DM, A. tortilis and G. bicolor contributed 71 and 39
kg ha-I, respectively (Figure 5.14) and all the individual woody species in the
rangeland site contributed more than 1% to the totalleafDM (Figure 5.14). The leaf
phenology during August and September was 1, while during December and January
varied among the different species. Acacia senegal and D. cinerea had a value of 0.4,
while that of A. tortilis, G. bicolor and B. aegyptica (0.6) and Acalypha species
having the lowest (0.2).
Harolel
The total leaf DM was 1 999 kg ha", with a leaf DM below l. 5 m and greater than
l.5 up to the height of 5 m of 1 022 kg ha" and 977 kg ha", respectively (Figure
5.13). The highest leaf DM was contributed by A. nubica (1 555 kg ha-I) and four
168
other individual woody species contributed more than 1% to the total leaf DM
(Figure 5.14). All the species had a phenological value of 1 during August and
September, whereas, in December and January; A. tortilis, C. sinensis, and B.
aegyptica had a value of 0.6. Acacia nubica, A. mellifera and V. eineraseens had the
lowest leaf phenology (0.2), whereas, C. rotundfolia had the highest (0.8).
5.3.2.2.3. Browsing capacity
Harole1
The BC for the rangeland site, across wet and dry months (WMS and DMS) and for
I
both heights varied between 2.36 and 3.80 ha BUl (Figure 5.15), while the BC for
A. nubica for the height below l.5 m for the wet and dry months was 5.43 and 10.59
ha BU-l (Figure 5.16), respectively. The BC for A.nubica, for the height greater than
l.5 up to 5 m was 3.81 ha BUl (wet months) and 7.43 ha BUl (dry months) and
contributed most to the BC (Figure 5.16).
00.0-1.5 m WMS ~ 0.0-1.5 mDMS
l1li>1.5-5.0m WMS 0>1.5-5 m DMS
70.0 -,----------------------,
,-..60.0
-;':::> 50.0 Kereyu-Fantale
CO 40.0
c<I
...c 30.0
u20.0
CO 10.0
0.0 ...f-C1:illIIIII."Y~.o..,-="""""'Y-
Figure 5.15. Approximate browsing capacity (ha BUl) across the rangeland sites
studied in Kereyu-Fantale district.
169
00.0-1.5 m WM III0.0-1.5 m DMS 00.0-1.5 m WMS !ill 0.0-1.5 m DMS
0>1.5-5 m WMS 1:1>1.5- 5.0 m DMS 0>1.5-5.0 m WMS [J >1.5-5.0 m DMS
200 ,------------~ 2500.0 ..,---------------,
"7::> 150 Chopi _~ 2000.0
CO ':::.
Fantale
ca 100
CO1500.0
..c
CJ 6 1000.0U
CO 50 CO 500.0
0 o .0 ..jI..IIIL..U.,l::III:EIL,UII-,.....w,LIII.-,dllll..llpillJtipHl,J..IiI<o,LBWj
As Gv Fg Gt Am Gb Ax Rn Or Bo Pa De Tb Ca As Da
Species Species
00.0-1.5 m WMS III 0.0-1.5 m DMS 00.0-1.5 m WMS Ill! 0.0-1.5 m DMS I
I 0>1.5-5 m WMS [J >1.5-5.0 DMS 0>1.5-5 m WMS El >1.5-5 m DMS
"7:~::. 68000 1 Kara J I
~ ::::. ,. ,. " ,,- ,.. ,_,J1,.d!, ,_
Rn Gt As Cf Dc At M ~u ~ & ~ ~ ~ G ~
Species Species
CJ 0.0-1.5 m WMS III 0.0-1.5 m DMS 00.0-1.5 m WMS II 0.0-1.5 m DMS
CJ >1.5- 5.0 m WMS [J >1.5-5.0 m DMS 0>1.5-5 m WMS El >1.5-5.0 m DM
1000
800 Mogassa
"7:::.
C.O, 600
.c 400
U
CO 200
As At Dc ~ Ba Ai[
0
Aif Gt An u Gv ~ At As
Species Species
00.0-1.5 m WMS III0.0-1.5 m DMS
o >1.5-5.0m WMS 1:1>1.5-5.0 m
3000 ,-------------------
~ 2500
"7:::. 2000 Kolbayu
~ 1500
~ 1000
CO 500
Lt At ~ Gv Gt Af An u Vc Sol Ba Co Po
Species
Figure 5.16. Browsing capacity (ha BU-I) for the mos! important individual woody plants by season
and height strata in Kereyu-Fantale district. Key to the species: As= A. senegal: Gv= G.
villosa; Fg= F. gnaphalocarpa; Gt= G. tenax; Am= A. mellifera; Gb= G. bicolor; Ax=
A. xanthophloea; Rn= R. neglecta; Or= 0. rochetiana; Bo= B. oleo ides; Pa= P.
africanum; Tb= T. brownii, Ca= e. africana, Da= D. afromontana, Cf= Cifascicularis;
De= D. cinerea; At= A. tortilis; Anu= A. nubica, Cs= e. sinensis, Ba= B. aegyptica; Cr=
e. rotundfolia; Vc= V cinerascens; Aif= A. fruticosa and A. indica; Lt= L. trifolia; Co=
e. ovalis, Po= P. oranatus.
170
Chopi
The BC at Chopi, in the wet and dry months, for the height below 1.5 m was 2.50
and 4.16 ha BU-l respectively, while it was 1.57 ha BUl (wet months) and 2.62 ha
BUl (dry months) for the height greater than 1.5 up to 5 m (Figure 5.15). The BC
for A. senegal was the highest (4.56 and 8.87 ha BUl for the wet and dry months for
the height below l.5 m respectively; 2.44 and 4.76 ha BUl for the wet and dry
months for the height greater than 1.5 up to 5 m respectively) (Figure 5.16).
Mogassa
The BC at Mogassa was 2.49 and 4.14 ha BUl for the height below 1.5 m for wet
and dry months respectively, while it was 1.29 (wet months) and 2.15 ha BUl (dry
months) for the height greater than 1.5 up to 5 m (Figure 5.15). Acacia senegal
followed by A. nubica had the highest BC among the individual woody species
(Figure 5.16).
][(ara
Compared to the above ones, Kara had a lower BC of 6.77 (wet months) and 11.26
ha BUl (dry months) for the height below 1.5 m and for the height greater than 1.5
up to 5 m, the BC was 9.62 (wet months) and 16.00 ha BUl (dry months) (Figure
5.15). Similarly, all the species had low BC (Figure 5.16) implying that their browse
production was low.
Kolbayu
The BC in this rangeland site was slightly higher for the heights greater than 1.5 up
to 5 m than the height below 1.5 m, which indicated that a large proportion of the
leaf DM was found above l.5 m (Figure 5.15). Similarly, the BC for the height
below l.5 m was 14.30 ha BUl (wet months) and 21.45 ha BUl (dry months), while
that for the height greater than l.5 up to 5 m was 8.27 (wet months) and 12.41 (dry
171
months) (Figure 5.15). At the height below 1.5 m, G. villosa and G. tenax had better
browsing capacity than other species; on the other hand, for the height above 1.5 m,
A. tortilis had a better browsing capacity than the other species (Figure 5.16).
Fantale
This rangeland had low browsing capacity which was reflected in the number of
hectare needed per browser unit (18.94 and 25.17 for the height below 1.5 m during
the wet and dry months respectively), while the corresponding values for the height
greater than 1.5 up to 5 m was 9.56 ha BUl (wet months) and 12.71 ha BUl (dry
months) (Figure 5.15). Relatively, R. natalensis had better BC than the others
(Figure 5.16) and the reason for the low BC was associated in part to the low
palatability of 6 of the species found in this rangeland.
Tuludimitu
The BC at Tuludimitu was the lowest among the rangeland sites, which was a
reflection of low browse production (Figure 5.15). The BC across the year, for the
height below 1.5 m (22.89 ha BUl and 34.90 ha BUl for the wet and dry months
respectively) was better than the height greater than 1.5 up to the height of 5 m
(38.47 ha BU-l and 58.66 ha BUl for wet and dry months respectively). Relatively,
A. tortilis had better browsing capacity than the others (89.26 ha BUl for wet months
and 14.83 ha BUl for the dry months year for the height below 1.5 m, while for the
height above 1.5 up to 5 m, the corresponding values were 51.57 ha BU-l and 70.39
ha BU-l for the wet and dry months respectively) (Figure 5.16).
5.3.2.3. Soil parameters
The pH of the soils in general varied from 7.9 to 8.95 and the soil along lake Beseka
was slighter higher in pH than the other rangeland sites. The electrical conductance
(EC) in (ds m") of the soil surface was highest in soil at lake Beseka than the soils in
other rangeland sites (Table 5.10). The EC in the other rangeland sites ranged from'
0.140toO.180.
172
The proportions of the components of the soil surface texture (% sand, % silt and %
clay) varied from one rangeland site to the other (Table 5.10) and the proportions of
sand, silt, and clay in rangeland sites were 40%, 35.7% and 24.3%, respectively.
The total nitrogen (%) was lowest in the rangeland site close to Lake Beseka (0.06%)
and the value in the other rangeland sites ranged between 0.12 to 0.25 indicating that
there was a variation in the % N from one rangeland site to the other. Similar to the
total nitrogen, the % organic carbon was lowest near to lake Beseka (0.48%), while
in the others it ranged from 1.14 to 2.16%. The CN ratio varied from 7 (Harole) to 13
(Fantale) and the available P (ppm) was between l.8 to 4.93, while that of available
K from 266 to 867 ppm indicating that there was wide variation in available K
(Table 5.10).
5.3.2.4. Correlation matrix among variables studied and multiple regression of
grass yield in relation to other parameters
The correlation among the variables studied, i.e., grass yield, basal cover, bare
ground, estimated soil erosion value, ETTE, pH, electrical conductivity, sand, silt,
clay, % total nitrogen, % organic carbon, eN ratio, available phosphorous, available
potassium and altitude is shown in Appendix 5.4. Of these, the relationships that are
considered most important were the relationship of grass yield, basal cover, ETTE
ha" with the other variables measured. Accordingly, grass yield was positively
correlated with basal cover (P<0.001), negatively correlated with bare ground (P<
0.001), ETTE ha" (P<O.OI) and pH (P<0.01), positively correlated with the CN ratio
(P<O.OI), available K (P<O.OI) and altitude (P<O.OI). Basal cover was negatively
correlated with bare ground (P<0.001) and positively correlated with available K
(P<O.OI).
173
Table 5.10. Physical and chemical characteristics of soils of the rangeland sites in
Kereyu-Fantale district.
Range Soil texture Other soil parameters
land site
'}~
Sand Silt Clay pH EC N OC CN Available Available
(%) (%) (%) (ds m") (%) (%) ratio Phosphorous Potassium
(ppm) (ppm)
Aleka 52 ± 33 ± 15 ± 8.2± 0.16± 0.17± 1.23± 9± 2.6 ± 409 ±
6 5 1 0.19 0.01 0.06 0.2 0.6 0.1 38
Harole 32 ± 35± 33± 8.4± 0.17± 0.16± 1.14± 7 3.5± 326±
7.1 2 4.7 0.1 0.01 0.02 0.1 ± 0.7 36.7
0.0
Harolel 45± 28.0± 27± 8.5± 0.14± 0.19± 1.47± 9± 4.l± 266.5±
17.6 12.5 12.5 3.0 0.1 0.1 0.5 3 1.49 105
Chopi 32± 38± 30± 8.4± 0.17± 0.25± 1.9± 8± 4.3± 268±
2 0 2 0.01 0.02 0.01 0.03 0.0 1.3 44
Mogassa 39± 40± 2l± 8.5± 0.15± 0.13± l.l± 8± 4.93± 280±
16.4 12.5 5.03 0.3 0.1 0.02 0.35 1.2 3.2 60.6
Kara 23± 48± 29± 8.2 0.2 0.25 2.l6± 9± 3.3± 43l±
3 4 1 ±O.l ±O.Ol ±O.Ol 0.5 0.0 1.1 111
Kolbayu 27± 44± 29± 8.2± 0.2± 0.2± 1.7± 10± 4.8± 867±
3 2 1 0.2 0.02 0.02 0.1 0.5 0.6 69
Tulu 53± 26± 2l± 8.3± 0.2± 0.2± l.5± 10± 1.93± 312.2±
Dimitu 0.0 1 1 0.15 0.01 0.01 0.07 0.01 0.2 20.5
Fantale 4l± 40± 19± 8± O.l± O.l± 1.6± l3± 1.8± 323±
1 2 1 0.05 0.08 0.02 0.1 1 0.6 9.5
eseka 50± 29± 2l± 9± 3.6± O.l± 0.5± 8± 3.1± 426.5±
5.5 2.5 3 0.6 0.03 0.01 0.2 l.0 0.1 123.5
A stepwise multiple regression analyses was undertaken to have an in sight into the
possible factors that give an explanation to the variation in grass yield (Table 5.11).
Accordingly, bare ground, ETTE ha", eN ratio and available K were selected.
174
Table 5.11. Standard coefficient, r values, R2 increment (%), R2 values (%) and
significance with dry matter yield as a dependent factor (Multiple
regression) .
Variables Standard r value ru ru Significanc
coefficient (lFrom increment values e
correlation (%) (%)
matrix)
Bare -0.3227 -0.69 7.21 48.20 P< 0.001
ground
(%)
eN ratio 0.5106 0.68 23.81 73.27 P< 0.001
Available 0.3187 0.59 8.45 83.10 P< 0.001
K
ETTE ha-l -0.2335 -0.50 4.47 88.0 P <0.01
5.4. DISCUSSION
5.4.1. Grass species composition
In general, the two study districts were dominated by C. plumulosus, different
species of sporobolus, C. dactylon, H. contortus, C. commutatus, D. aegypticum, P.
dilatatum, C. poiflorum and T. cenchriformis. The species identified in this study
were, in general, similar to those reported by Schloeder & Jacobs (1993) from their
study of the nearby Awash national park and the findings of this study generally
concurs with the concept that, the potential natural community of a site, is dominated
by one or a few species, which are best adapted to the specific combination of
environmental factors of that site (RISe 1983). Furthermore, the dominance of C.
plumulosus in 50% of the rangeland sites in both districts indicates the ability of this
grass species to resist severe overgrazing. The abundance of S. spicatus and S.
175
ioclados at Lake Beseka and Asobabad was related to the adaptability of these grass
species to saline soils (CADU 1974; Ibrahim & Kabuyu 1987).
In both districts, the percentage of the highly desirable (18.75% to 23.08%) species
was less than the desirable (17.95% to 28.13%) and the less desirable (53.12% to
58.97%), while the grass layer was mainly dominated by C. plumulosus.
Notwithstanding the limitations of such a classification of grass species, which is
subjective, it gave an indication of the nature of the grass layer, both from an
ecological (response to' grazing) and palatability point of view. The highly desirable
grass species are those group of grass species which include decreasers (species
which dominate in good condition rangeland) and which were rated as highly
palatable species by the pastoralists of the study districts. Grass species such as C.
ciliaris (Ayana & Baars 2000; Amsalu 2000), C. gayana (Amsalu 2000), D.
milianjiana (Ayana & Baars 2000; Amsalu 2000), P. coloratum (Ayana & Baars
2000) were grouped as decreasers, which is in agreement with the resuIt of this
study. Themeda triandra and the species mentioned above are also classified as
decreasers in South Africa (e.g., Tainton et al. 1976; Van Rooyen et al. 1991; Van
Oudtshoorn 1999) and East Africa (Barker el al. 1989).
The desirable grass species consisted of those grass species, which are high or
intermediate in their palatability and were Increaser Ha and perennial grasses. The
Increaser Ha are groups of grasses, which increase with moderate over utilization
(Vorster 1982). The grass species that were grouped are C. commutatus, E.
racemosa, H. contortus, P. glumaceum, H. hirta, P. stramineum, S. loc/ados, C.
roxburghiana and L. nutans. The likely reason for such variation is associated with
the fact that species may possibly differ in their reactions to grazing intensity
(degradation). Janse van Rensburg & Bosch (1990) argued that one species can have
a different average abundance (different ecological groupings) in different
topographic units and between edaphic variations within the same unit. For example,
ephemeral forbs and grass species are usually classified as undesirable Increasers
176
with a low grazing value (Fourie et al. 1987). However, in deserts these species are
important in the diets of herbivores (Dieckmann 1980).
The less desirable grass species consisted of Increaser Ilb and IIc species and
according to Vorster (1982), the grass species in the Increaser lIb group increase
with severe over utilization, while the Increaser IIc groups increase in abundance
with severe over-utilisation. This group comprised annual and perennials with a low
palatability as perceived by the pastoralists. Some of the grass species in this group
are found on disturbed areas and can, therefore, be used as indicators of poor
rangeland condition. For example, species like A. adscensionsis, T. berteronianus, C.
dactylon, D. aegypticum, D. ternata, S. verticellata, U panicoides, P. snowdenii, S.
ustilata, P. pycnothrix were reported to occur on disturbed areas (CADU 1976;
Tainton et al. 1976; Ibrahim & Kabuye 1987; Barker et al. 1989; Van Oudtshoorn
1999).
The perennial grass species are better indicators of the ecological status of an area
compared to annual grass species, because significant changes in abundance between
Decreaser and Increaser perennial species are most likely to take place only after
significant changes occurred in the degradation status of an area (Cable 1975; du
Plessis et al. 1998; Snyman 1999a). In both districts, the percentage of perennial
grasses was higher than the annual grasses. The lower abundance of the annual grass
species, in this study, could be due to the reaction of annual grasses to small changes
in rainfall and grazing pressure. Rainfall has a marked influence on the presence or
dominance of annual species (Cable 1975; Van Rooyen et al. 1990) and is also the
main determinant of forage production (Fourie et al. 1987).
In both districts, many grass species (A. adscenionsis, T. berteronianus, C. dactylon,
D. aegypticum, D. ternata, S. verticellata, U panicoides, P. snowdenii, S. ustilata, P.
pychnothrix, A. adoensis, S. natalensis, E. cilianensis, C. poiflorum, E. indica and
M kunthii) are indicators of changes in species composition of the grass layer that
took place (CADU 1976; Ibrahim & Kabayu 1987). Long-term experiments are
required to separate the effects of different interacting factors that caused the changes
177
in species composition in the study districts. Nevertheless, studies made by others
(Pressland & Graham 1989; Fabregues et al. 1992; Duncan et al. 1993; Milchunas &
Lauenroth 1993; 0' Connor 1994) indicated that drought and overgrazing can
influence the vegetation composition of an area. Frost et al. (1986) and Westoby et
al. (1989) argued that the influence of grazing on the botanical composition of
rangelands is a central tenet of rangeland management, yet the population dynamics
of important savanna grass species appears to be more dependent upon rainfall
variability than upon grazing. The importance of adequate soil water for seedling
recruitment is widely recognized (Glendenning 1941; Harradine & Whalley 1980;
Cox 1984; Peart 1984; Mott et al. 1985; Fowler 1986; Frasier et al. 1987; O'Connor
et al. 2001). Although there is no adequate data base from many sites across the
rangeland sites to ascertain the significance of rainfall the 35 years data collected by
Mathara Estate sugar indicated that in 42.9% of the rainfall years, the rainfall was
above the long term average and in 57.2% of the years, the rainfall was below the
long term average.
It was not possible to determine the stocking rate in the study area. Due to a lack of
land there exist conflict over the use of available resources, further characterized by a
lack of land capability planning, where pastoralists keep too many animals on a
relatively smaller area, which normally leads to overgrazing of the rangeland. This
can lead to wind erosion, further removing the soil from the surface. The influence of
overgrazing on species composition is well documented (O'Connor & Pickett, 1992;
Milchunas & Lauenroth 1993; O'Connor 1995). Livestock first select the most
palatable grass species and if the grazing pressure continues, a decline in the quality
and the quantity of the rangeland occurs (Lazenby & Swain 1969; Cossins & Upton
1985). This causes reduced vigour, less seed production and eventually plant death.
Overgrazing can lead to extensive sheet and gully erosion (Pratt & Gwynne 1977).
Therefore, to enable optimal, sustainable use of the rangeland there should be an
adequate rest period for the grass species to grow, to seed, and be able to accumulate
reserves for the next growing season.
178
In conclusion, a more comprehensive understanding of the environmental factors and
population processes primarily responsible for population and community changes
will, therefore, only be achieved within the context of long-term field experiments.
5.4.2. Basal cover, bare ground and estimated soil erosion values
The results of the study showed that in general, the % basal cover of the studied
rangeland sites in both districts was low. In relative terms, the only exceptions on
the Mars side were Top of bud and Degage with basal cover values of 4.8% and
4.6%, respectively. Kolbayu and Tulu dimitu with basal cover values of 4.7% and
4.6% from the rangeland sites in the side of the Oromos can be said to have relatively
better cover values. Though the intensity of soil loss varied from one rangeland site
to another, in general there was a loss of soil from the surface that was also shown in
the estimated soil erosion values and percent bare ground. The results further
indicated that, the % bareground was relatively high at Asobadad and Samayu
(Awash-Fantale), and Mogassa, Beseka and Chopi (Kereyu-Fantale).
Normally the basal cover of excellent vegetation is expected to be greater than 12%
(Baars et al. 1997) and if the basal cover obtained in the study districts is compared
with this value, the mean basal cover values for Awash-Fantale (3.43%) and Kereyu-
Fantale (3.23%) were very low. Snyman (1997) reported a low basal cover for
vegetation in poor condition in a semi-arid climate. Many different factors can lead
to a low basal cover and high soil loss from the surface. From this study it would
appear that overgrazing (Milchunas & Lauenroth 1993; O'Connor et al. 2001),
drought (William et al. 1990), poor grazing practices (Snyman & Fouche, 1993) and
high tree densities (Smit et al. 1999; Smit 2002) are the most likely determinants.
These factors are interrelated and it is difficult to separate the effect of one from the
other without a long-term experiment.
Overgrazing is widely believed to cause a decline In the basal cover of the
herbaceous vegetation (O'Connor 1985; Thrash et al. 1991). Several studies
reviewed by O'Connor (1985), on the effects of grazing by livestock showed that the
179
herbaceous stratum basal cover decreased when subjected to severe grazing. Thrash
et al. (1991) upon their study of the impact of provision of water for wildlife on the
basal cover of the herbaceous vegetation showed that the relative basal cover of the
herbaceous stratum is more sensitive to the impact of the dam than the woody
species. Snyman (1999a) concluded that the basal and canopy cover provided by a
vegetation community protects the soil surface from raindrop impact, which causes
surface sealing and splash erosion by reducing the considerable kinetic energy from
the falling raindrops. The protection, which grass plants afford, the soil is normally
much greater than that given by most non-grass plants. This is so because, for a given
crown cover, grasses have a much greater basal cover than shrubs (Roux 1981) and
provide an extremely complex network of roots immediately below the soil surface
(Skinner 1964; Theron 1964; Roux 1969; Pressland 1973; Roux 1981; Elkins et al.
1986). In general, vegetation reduces erosion in a variety of largely unrelated ways,
where the density, cover, botanical composition, canopy height and perenniality of
the plant community are the most important (Snyman 1999a).
The study districts had a relatively higher proportion of perennials than annuals. It is
known that stands of perennial grasses provide better cover than the annuals, because
they provide a much denser and more stable cover. Basal cover has an important
influence on the rate at which water infiltrates into the soil, and by promoting
infiltration it reduces erosion wash (Snyman 1999a,b). O'Connor et al. (2001)
indicated that a low basal cover also signifies an increased area of exposed soil,
which is in agreement with the findings of this study. Bare ground is a good
indication of over-utilisation and degree of the degradation of the vegetation. Du
Plessis et al. (1998) made a similar conclusion from their study in the Etosha
National park in Nambia.
Although the action of the hoof of animals on soil status was not addressed in the
study, one can speculate that this action may have an influence on the soil condition.
Because the animals are concentrated in smaller areas with a high intensity of
movement in search of food it is likely to increase hoof action. In the more arid
180
regions, the hoof action of grazing domestic animals is taken as an important factor
for the initiation of erosion (Snyman 1999a).
Historical evidence indicates that natural climatic patterns produce cycles of drought,
followed by periods of relatively higher rainfall. However, Snyman (l999a) reported
that drought can be intensified by injudicious grazing practices or overgrazing, which
can lead to man made droughts. Degradation of the rangeland ecosystem,
accompanied by the loss of the soil, can result in an increased intensity of
management-induced droughts and this kind of drought is usually localized (Snyman
& Fouche 1991, 1993). This, in turn, results in the reduction of infiltration of
rainwater into the soil, which reduces the vigor of the vegetation and this in turn,
increases the severity of flooding by increasing surface runoff Nevertheless, this
argument, does not exclude the existence of sporadic climatic droughts due to a low
amount of rainfall. There have been six extensive drought episodes on the African
continent in the last three decades: 1965-66, 1972-74, 1981-84, 1986-87, 1991-92,
and 1994-95. Drought affects the livelihood of pastoralists that in turn determine
their responses to cope with the condition (Zinash et al. 2000). At these times, forage
resources dwindle to very low levels. Plant cover also declines as plants are killed by
the combined pressure of heavy grazing (Abel 1993) and a lack of plant-available
water (Snyman et al. 1987, 1998, 1999; O'Connor & Bredenkamp 1997). The
situation of the study districts with regard to rainfall was discussed in section 5.4.1.
Considering some rangeland sites in the study districts (Mogassa, Samayu and Chopi
and Bulga riverside), these sites had a higher tree density compared to the others as
characterized by their high ETTE ha-l values. Thus, their low basal cover could be
also associated with high tree densities. Smit et al. (1999) reported that many
savanna areas are water-limited ecosystems and an increase in tree density (bush
encroachment) is considered a major factor contributing to the low occurrence or
even total abscence of herbaceous plants in severe cases.
The relatively better basal cover at Kolbayu and Degage can be related to the fact
that these rangeland sites are found at the junction between the Mars and the Oromos
181
where animals cannot graze as freely as they graze the other rangelands.
Furthermore, because of the shortage of water in these areas during the dry season,
these areas are not grazed as often as the other rangeland sites.
5.4.3. Grass DM yield, grazing capacity and rangeland condition
The mean yield of grass, for the study districts, 437.08 kg ha" for Awash-Fantale and
421.92 Kg ha" for Kereyu- Fantale districts were relatively lower than that of
Amsalu (2000) who reported a yield of 533 kg ha-I from his study in the mid rift
valley of Ethiopia that has a comparable climatic situation. The relatively lower yield
in this study could be due to fluctation in species abundance and herbaceous yield as
determined by rainfall, soil type and the tree-grass ratio (0'Connor 1991).
In both districts, the yield of grass was the highest and the lowest in rangelands with
woody vegetation (Degage, Kolbayu, Samayu and Mogassa). The main reason for
the increase in grass production at Degage and Kolbayu could be due to the relatively
better basal cover value, low tree density and low grazing intensity in these
rangelands as was discussed in section 5.4.2. Increased grass production due to better
plant cover is in agreement with the findings of' O'Connor et al. (2001). O'Connor et
al. (2001) from their study concluded that maintaining basal cover is very important
for grass production, because basal cover is the primary control of water retention
and growth opportunity, but composition and species identity play an additional role.
The decrease in grass production at Samayu and Mogassa, when compared to the
other sites, could be associated with the high tree densities in these rangeland sites as
was expressed by the high ETTE ha".
Determining grazing capacity is a controversial subject due to an abundance of
factors influencing its determination (Vorster 1981a,b; Roe 1997). When one factor
changes, grazing capacity changes to a greater or lesser extent. Furthermore, the
system of land ownership in the pastoral community is communal, where numerous
households, each making more or less independent management decisions, are using
182
the same grazing resource. Though environmental factors determine the inherent
grazing capacity (0'Connor & Bredenkamp 1997), management factors (Danckwerts
& Tainton, 1996) and climatic variation (Danckwerts 1982; Van den Berg 1983; Van
der Westhuizen 1994; Snyman, 1998; Van der Westhuizen et al. 2001) also influence
the degree of variation from the potential long-term grazing capacity. Accordingly,
the mean grazing capacity of 23.6 (10.7 to 35.6ha LSU1) and 14.7 (6.7 to 22.2 ha
TLU1) in Awash-Fantale district and the mean grazing capacitity of 24.8 ha LSU-1
(11.2 to 54.1 ha LSU1) or 14.5 ha TLU-1 (7 to 28 ha TLU-1) in Kereyu-Fantale
district implies that more land is required to sustain an LSU or TLU without
damaging the rangeland ecosystem. The estimation of grazing capacity in this study
was undertaken to give an indication of the extent of the problem in the study
districts. In comparison to others, the grazing capacity of Madala1, Top of bud1,
Samayu, Bulga riverside, Della kara (Awash-Fantale district) and Mogassa, Chopi
and Kara (Kereyu-Fantale) were low. This could possibly be ascribed to the high tree
density in these rangeland sites. An increase in the woody plant density beyond a
critical density was reported to result in the suppression of herbaceous plants, mainly
due to severe competition for available soil water, thus lowering the yield of the
herbaceous layer and also the grazing capacity (Dye & Spear 1982; Moore 1989;
O'Connor 1991; Smit & Rethman 1998b; Richter et al. 2001). This suppressive
effect of bush encroachment is the major reason why clearing or thinning of trees is
considered as a management option (Donaldson 1973; Van Niekerk & Kotze 1977;
Jacoby et al. 1982; Gammon 1984; Moore et al. 1985; Heitschmidt et al. 1986;
Moore & Odendaal 1987; Scholes 1987).
Rangeland condition in this study was calculated based upon the species composition
and the desirability of the species under consideration. In both districts, compared to
the other rangeland sites, rangeland sites along Lake Beseka and Asobabad had a low
rangeland condition score. The low rangeland condition score along Asobabad and
Lake Beseka could be attributed to the high percentage bare ground (8.1 to 10.8%),
the high abundance of forbs (17%) at Asobabad and the fact that these rangeland
sites were also dominated by S. spicatus which was less desirable grass species.
183
From the results it is clear that some rangeland sites such as Madala 1, Samayu, Della
kara and Bulga riverside (Awash-Fantale district) and Harole1, Kara, Mogassa and
Chopi (Kereyu-Fantale district) had a lower rangeland condition score compared to
the others. This is due to the presence of forbs, increased bare ground and less
desirable grass species, which lower the rangeland condition score either on their
own or in combination. The reasons for this and the arguments were discussed in
sections 5.4.1 & 5.4.2. In some rangeland sites, the presence of desirable and highly
desirable grass species resulted in a rangeland condition values as high as 90%. This,
however, indicates that the use of a single factor such as grass species composition
alone may not be a good indicator of rangeland condition since it was based upon the
subjective allocation of the grass species to different categories. Anderson (1985)
and James et al. (1991) stated that the method of rating rangeland condition by
species composition alone appeared not to be reliable for assessing rangeland
condition. Therefore, a combination of factors such as soil parameters and woody
layer (where available) must be used to arrive at a more accurate range Iand condition
value.
5.4.4. Woody vegetation composition
The woody vegetation composition in the study area was dominated by species of
Acacia notably A. senegal and A. nubica. Futhermore, species of Acalypha, Solanum,
Vernonia and Grewia were also relatively abundant. The woody plants identified in
this study were more or less similar to those the reported by Schloeder and Jacobs
(1993) from their study in the Awash National park.
Some of the woody species identified in the study districts such as Solanum species
give an indication of the condition of the rangelands. According to Dale and
Greeenway (1961) and Irvine (1961), Solanum species, are indicators of a change in
the condition of the rangeland towards deterioration and are also considered as
poisonous plants in Ethiopia (Mekonnen 1994).
The low density of some of the woody species such as A. nilotica, A. tortilis and B.
aegyptica could be associated with the usage of these woody plants for fire wood and
184
charcoal. Species of woody plants not found in the other rangeland sites, were found
in the Fantale grazing land and this could possibly be associated with the higher
elevation (1 650 m.a.s.l) of this area than the others. This result concurs with the
findings that altitude with its influence on temperature and precipitation has a
primary influence on vegetation composition of an area (Alemayehu 1979; Pratt &
Gwynne 1977; Alemu 1987; Ayana & Baars 2000).
5.4.5. Browse production and browsing capacity
Browse plants play an important role in the diets of ungulates living in arid zones
(Walker 1980; Nott & Savage 1985). Owing to this fact, livestock production in the
study districts is based on the integration of grazers (cattle and sheep) and browsers
(camel and goats). The availability of forage that can support the different livestock
species is considered to be the most important factor influencing habitat selection by
large herbivores (McNaughton 1987). Furthermore, the fluctuations in live biomass
are extremely important in determining food habits (Kutilek 1979; Jarman & Sinclair
1979), movement patterns (Talbot & Talbot 1963; Maddock 1979), and habitat
utilization (Lamprey 1963) by large mammalian herbivores. Therefore,
determination of browse production for stock usage is very important.
Compared with research efforts on the grasslands in East Africa, the browse
component of savanna ecosystems has for long been neglected (Pellow 1983).
Accordingly, there are limited published results of browse production to make
comparison of the results of this study with others. Rutherford (1982) reported a
production of 1 100 kg ha-t of leaves from a study of the above ground biomass of
Burkea africana-ochana pulchra savanna and Scholes (1987) reported a yield of 801
kg ha-t of leaves for C. mopane. In addition, the study by Walker (1980)
demonstrated that production varies strongly owing both to inherent site differences
(mainly soil depth) to management and reported a yield of 594 kg ha-t to 2 121 kg ha-
t from a C. mopane arid shrub and tree savanna in Zimbabwe, while Dekker and
Smit (1996) reported a yield of 1 224 kg ha-t to 2 672 kg ha-t for C. mopane.
Therefore, in light of the existing knowledge, the total leaf DM at peak biomass
185
produced across the rangeland sites (196 kg ha" to 3311 kg ha-I) can be considered
to be within acceptable ranges.
In the study districts, of the total leaf DM produced by woody plants, 4 773 kg ha-l
(31.2%) in Awash-Fantale and 4 003 kg ha-l (39.87%) in Kereyu-Fantale, was found
to be below 1.5 m which implies that this amount, at least theoretically, is available
for browsing by goats. Nine thousand two hundred and forty six kg ha-l (60.34%) in
Awash-Fantale and 5 707 kg ha-l (56.84%) in Kereyu-Fantale was found above 1.5
m up to 5 m. Although, the mean height of camels is about 3.5 m, it can be said that
most of this browse is within their reach. Smaller browsers will only have access to
these leaves when they dry and fall to the ground. The abundance of browse in the
height above 1.5 m is in agreement with the report of Kozlowwski & KeIler (1966)
who reported that woody species growth is terminal and greater at the top of the
plant, which implies that greater productivity occurs in the upper canopy than the
lower canopy. The remaining 1 125 kg ha-l (8.52%) in Awash-Fantale and 330 kg ha-
l in Kereyu-Fantale was found greater than 5 m height. Of the amount that was
beyond 5 m, the woody plants along the Bulga riverside, mainly, A. nilatica and A.
tortilis that can only be used by breaking the branches, contributed 85.6%. The
browsing heights considered are mean heights and not maximum browsing heights
and were only used to draw comparisons. It is known that large individuals are able
to reach higher than those mean heights, e.g., 2.5 and 5.5 m for kudu and giraffe
respectively (Dayton 1978), while breaking of branches may enable some browsers
to utilize browse even at higher strata (Styles 1993). Furthermore, both pastoral
groups lop browse trees during the dry season (refer Chapter 4).
Among the studied rangeland sites in Awash-Fantale (Awah-FantaleDulcha junction
site, Bulga riverside, Top of budl, Samayu, Madala1, Della kara) and Chopi,
Harolel and Mogassa in Kereyu-Fantale had a higher browse production at peak
biomass than the other grazing lands and this was mainly attributed to the higher
density of woody plants. Possibly more important than the amount of available
browse at peak biomass is the continuity of browse availability over the seasons.
Leaves retained on the trees in a younger phenological state followed by the early
186
emergence of new season's leaves are likely to be of greater value to browsers than
where the trees are leafless for some time with only dry leaves available on the
ground. To this effect, a detailed study was not undertaken on the phenological
behavior of the identified woody species. Nevertheless, the preliminary investigation
undertaken revealed, except along the Bulga riverside, Awash-Fantale Dulcha
junction and Fantale, most woody plants shed their leaves during the dry season, as
the study districts were also dominated by deciduous species of Acacia.
Deciduousness is the feature of trees within the savannas across the world with
pronounced seasonal reduction in canopy cover as the dry season progresses
(Williams et al. 1997). There appears to be a considerable variation in ecosystem
level patterns of deciduousness in the world's savannas. African savannas are mainly
deciduous (Menaut & Cesar 1979; Chidumayo 1990; Smit 1999a) which confirms
that the availability of browse to the browsers will be low during the dry season.
In Awash-Fantale district, E. abyssinica, C. grandiflora, C. rotundfolia, S. perisca
and P. juliflora showed a higher phenological value during the dry season as well.
This might be attributed to their access to water along rivers where they can get
water through their deep taproots. The abundance of browse along the rivers, in turn
implies the dependency of the pastoralists to these areas as a source of browse to
their stock during the dry season. However, E. abyssinica and C. grandiflora were
indicated as poisonous plants by the pastoralists, which limits their use as a source of
feed to livestock and also their contribution to the totalleafDM was small.
Some species (0. rochetina, P. africanum, B. oleiodes, R. communis and D.
ajromontana) along Mount Fantale also displayed higher phenological values.
Nevertheless, these woody plants were rated as unpalatable by the Oromo
pastoralists and hence their use as a source of browse is very limited, which concurs
with the report of Grunow (1980) who indicated that browsers select among plant
species as markedly as grazers do. Furthermore, chemical defences of plants may
also limit the use of browses (Robbins et al. 1987; Bryant et al. 1992). The variation
in leaf DM, even for a similar species across the rangelands with the same ecological
condition can possibly be due to variation between the different vegetation types
187
(Hall-Martin & Fuller 1975; Guy et al. 1979) and can therefore be considered
important factors that influence the distribution of herbivores.
In conclusion, in both districts, the various species of Acacia were the main source of
browse for livestock and hence of the browsing capacity of the rangeland sites.
5.4.6. Bush encroachment
Bush encroachment is one of the most serious and conspicuous results of imbalances
in savanna ecosystems (Richter et al. 2001) and remains one of the major structural
problems limiting optimum animal production (Bester & Reed 1997). The study
confirmed that substantial bush encroachment had taken place in both districts,
which were dominated by species of Acacia (A. senegal, A. nubica and A. mellifera)
and P. juliflora in the case of Awash-Fantale district. Furthermore, C. rotundfolia
and Acalypha species in Awash-Fantale and G. bicolor and G. villosa in Kereyu-
Fantale had a relative higher ETTE ha" value than other species. Just considering
only A. senegal, five rangelands across the study districts were covered mainly with
this encroaching plant.
In most areas of East Africa, the natural vegetation is dominated by woody species,
where edaphic factors do not prevent growth of these life forms (pratt & Gwynne
1979), bush encroachment is a wide spread phenomenon (Prins & Van der Jeugd
1992). The causal factors for bush enroachment are not absolutely clear (Prins & Van
der Jeugd 1992) rather; they are diverse and complex (Smit 2002). Nevertheless Smit
(2002), reported that an increase of woody plant abundance is primarily a function of
two processes. The first is by an increase in the biomass of already established plants
(vegetative growth) and the second is by an increase in tree density, mainly from the
establishment of seedlings (reproduction). Teague & Smit (1992) further explained
that, man has modified the determinants of savanna systems in most situations, either
directly or indirectly. These determinants may either be primary (such as climate and
soil) or secondary (fire and the impact of herbivores).
188
The encroachment by species of Acacia has been reported from different countries
e.g., Ben- Shahar (1991), Stuart-Hill & Tainton (1999); Smit (1999a); Hagos (2001)
in South Africa; Arntzen & Veenendaal (1986), Mcleod (1992); Dahlberg (2000) in
Botswana; Prins & Van der Jeugd (1993) in Tanzania, Mugasi et al. (2000) in
Uganda. A host of other woody species were also reported; e.g., species ofMay tenus,
Euclea, Grewia and Dichrostachys cinerea. Acalypha fruticosa was also indicated as
an encroacher woody plant (Prins & Van der Jeugd 1993) in Tanzania.
Ben-Sharar (1991) studied successional pattern of woody plants in catchment area in
a semi-arid region in South Africa. He reported that A. senegal became dominant in
areas previously dominated by A. tortilis. The successful establishment of A. senegal
seedlings seemed to depend on effective utilization of open spaces and probably also
supported by a fast growth rate. Knoop (1982) also observed that on a site dominated
by Acacia species, a large number of seedlings germinated and survived in an area
cleared of vegetation, but that few were found in an uncleared area. Although, the
role of roots was not covered in this study, other studies (Adams 1967; Hagos 2001)
done on A. mellifera indicated that their root systems were extending laterally
beyond the canopy cover and could play a part in the spacing of trees. The dispersion
of roots could play a major role in the local depletion of water and nutrients around
the canopy and beyond. It is generally conceded that trees are able to make more
effective use of deep water than are the grasses (James 1967; Wu et al. 1985; Wilson
1988; Prins & Van der Jeugd 1992) so that any management actions that increase
water penetration to a depth in the soil profile should stimulate growth in the already
established trees.
Effective dispersal could also be brought about by means of birds or mammals that
travel long distances (Gwynne 1969; Coe & Coe 1987). Although, the seeds of A.
senegal are soft dehiscent seeds, which can be damaged when passing through the
buccal cavity and alimentary systems of the herbivores (Ben-Shahar 1990), this point
is worth mentioning when one looks at the migratory pattern of the different
livestock species in the study districts. Camels are mobile throughout the year and
the other livestock species are also moved from place to place in search of feed and
189
water In an area of resource use conflict where shortage of land is a critical
constraint. Compared, to the intensity of migration, there is a chance that seeds of
different woody plants can be transported from place to place.
Although contradictory reports were given to the wide spread general view that long
term heavy grazing is a prerequisite for an increased rate of woody plant
establishment, work done by Smit & Rethman (1992) has shown that heavy grazing
is one of the most important contributing factors. The study area is an area of
resource use conflict with a shortage of land. The rainfall in the study districts for
many years has also been lower than the long-term average. The situation in the
study districts can also incline to the common view that the destruction of grasses by
overgrazing during drought offers a window of opportunity for the increase of the
woody component because of reduced fire and competition (Du Toit 1972). In
general, evidence from the literature and research work in different countries suggest
that bush encroachment takes place as a result of the exclusive use of moisture by
encroachers, high soil nutrient concentrations, low fire frequencies and high cattle
selectivity (Moleele et al. 2002; Smit 2002).
In the Afar region, P. juliflora is known as undesirable noxious plant (alien species)
and this plant is becoming a major problem. In Awash-Fantale district, this plant is
found localized near the town of Awash and at the border areas neighbouring Awash-
Fantale district and is spreading into the grazingland of the pastoralists living in
Awash-Fantale. It was indicated in the study report of Gebru (2001) that this woody
plant was introduced first by an Indian researcher for the purpose of soil
conservation, especially targeted to the eroded areas of Western Dire Dawa in
Ethiopia and later on it spread to various places of the country. A British consulting
firm first introduced this plant to the Afar region either from the neighboring Sudan
or Ethiopian workers during the period between 1971-1974 around the state farms in
the middle Awash valley of Ethiopia. The pastoralists use the palatable pods of the
plant for livestock feeding and use the plant as a source of fuel wood. Nevertheless,
they strongly complained about its rapid expansion and rapid coverage of large areas
of rangelands, which decreases the size of the rangelands available for grazing
190
animals. Reynolds & Glendening (1949) indicated that between 12 to 45% of the
hard P. juliflora seeds pass undamaged through livestock. Furthermore,
Glewndening & Paulsen (1950) reported that 27% of the total seeds fed to sheep
germinated after ingestion.
5.4.7. Soil
5.4.7.1. IPH, electrical conductance and texture
The direct comparisons of the soil nutrient levels with other studies have limited
value because of the difference in soil type and methods of analysis. As a result, the
discussion below will be more generalized. The pH of the soil and associated
properties strongly influence plant nutrient availability and soil productivity (Tisdale
et al. 1993). Accordingly, the pH values of the soil in the study districts were
generally alkaline. Although, sodium carbonate and bicarbonate were not analyzed in
this study, the higher pH at Lake Beseka and at Asobabad could possibly be
associated with the frequent occurrence of ground water containing high proportions
of sodium carbonate and bicarbonate where the pH exceeds 9.5.
The size of the mineral soil particles vanes across a wide continuous spectrum.
Although, there were variations from one place to another, the soils on the surface
generally were with a good proportion of sand and silt. Similar to the pH, the
electrical conductance was found to be highest at Asobabad and along lake Beseka.
Individual soluble salts were not quantified in this study, yet it is possible that the
high electrical conductivity values at these places could be due to the accumulation
ofNa + Ca ++, Mg ", cr and S04-- (Thompson & Frederick 1978).
5.4.7.2. Percent total Nitrogen
Nitrogen is a vitally important plant nutrient and is the most frequently deficient of
all nutrients (Tisdale et al. 1993). Accordingly, the lowest % N was obtained along
Lake Beseka and Asobabad. This could possibly be associated with the low
accumulation of organic materials that can contribute to the low soil N in these
grazing lands. The variations in N from one rangeland site to the other, shown in the
191
results, might be associated with the contribution of trees to N, nitrogen fixation by
Rhizobium bacteria on the roots of leguminous woody plants and also there might be
the influence of grazing animals. Peterson et al. (1956) argued that grazing animals
ingest a large amount of N in the feed, but 75% is excreted. In general, Bauer et al.
(1987) indicated that large amounts of soil total nitrogen in grazed condition could
result from a source of N extraneous to the grazed system.
5.4.7.3. Organic carbon
Soil organic matter is a key component of any ecosystem and any variation in its
abundance and nature may have profound effects on soil processes. A High
concentration of organic matter in soil is important, because soil organic matter
concentration is one of the most important factor in the storage of nutrients and it is
also important for maintaining good soil structure especially in fine textured soils
(Du Preez & Snyman 1993).
In the study districts, the percentage OC was the lowest at Asobabad and along Lake
Beseka and these soils are saline soils. The low percentage OC in these grazing lands
could be due to the low accumulation of plant materials that can contribute to soil
enrichment. The results also indicated that in both districts, the percent OC was
higher in grazing lands with woody plants than those without woody plants and the
relatively higher content of soil OC in the former demonstrated the soil enriching
capacity of trees. The contribution of woody plants to soil nutrients is well known
(Bosch & Van Wyk 1970; Kennard & Walker 1973; Tiedemann & Klemmedson
1973; Berhard-Reversat 1982; Kamara & Haque 1992; Smit & Swart 1994; Wezel et
al 2000; Smit 2002). The reasons for the higher content could be complex (Vettas
1992). Leaf litter from leaf fall has been mentioned as a possible source (Bosch &
Van Wyk 1970; Stuart-HilI et al. 1987; Belsky et al. 1989; Weltzein & Coughenour
1990). Dust accumulated on tree leaves and branches by wind blowing from the
surroundings could be one possible source of nutrient, as it is washed down to under
canopy soil during stem flow and through fall (Kellman 1979). Woody species have
deeper taproots and extensive spreading lateral roots and such root systems are
probably important in concentrating nutrients under the canopy (Kellman 1979).
192
Knoop & Walker (1985) suggested that nutrients found in low concentration
throughout the soil profile may be taken up by the root system of mature trees and
shrubs and these nutrients return back to under canopy soil when tree leaves fall and
decompose (Garcia-Moya & McKell 1970).
If one looks at the grazing lands without woody vegetation (Aleka, Harole, Madala
and Top of bud), except those mentioned above (Asobabad and lake Beseka), the
organic carbon is not particularly low. This might be due to the effect of manure,
although this was not investigated in this study. In the study districts animals are
concentrated on a relatively small area, where the effect of manure might influence
the soil nutrient status. Glaser et al. (2001) noted the influence of manure on the soil
nutrient status in a semi-arid region of Northern Tanzania.
5.4.7.4. Carbon Nitrogen ratio
The zone of maximium soil OC and N accumulation is restricted to the upper soil
depth (0-50 mm) (Derner et al. 1997) and the CN ratio defines the relative quantities
of these two elements, while in determining the soils N - potentialiy, the CN ratio of
the organic matter is of importance. The CIN ratio of undisturbed topsoil in
equilibrium with its environment is about 10 or 12 and a narrow CIN ratio indicates
that the organic matter is rich in N (Tisdale et al. 1993). Based upon this, in Awash-
Fantale district, of the studied rangeland sites, 44.4% of them had a CN ratio of 8,
while the remaining had a value of 10 (11.1%), 9 (11.1%), 7 (22.2%) and 6 (11.1%).
Correspondingly, in Kereyu- Fantale district, 30%, 30%, 20%, 10% and 10% of the
rangeland sites studied had a CN ratio of 9, 8, 10, 7 and 13 respectively. Barnes et al.
(1991) argued that a relatively narrow ratio of C to N in the soil would favor N
release from the organic matter.
5.4.8. Correlation among the variables studied and multiple regression between
grass yield and other parameters
It was shown in the results that, in both districts, grass DM yield was negatively
correlated with bare ground, estimated soil erosion value, ErrE ha-1 and positively
193
correlated with basal cover, and altitude (Kereyu-Fantale district). The stepwise
regression analysis indicated that grass yield was dependent on ETTE ha-1 and basal
cover in Awash-Fantale, while on percentage bare ground, ETTE, CN ratio and
available K in Kereyu-Fantale district. The difference between the two districts lies
in the fact that, the total ETTE value in Awash-fantale district (65 315) was higher
than that of Kereyu-Fantale district (44 221). The effect of tree density, expressed as
ETTE ha-Ion grass yield was discussed in section 5.4.3. The influence of percent
basal cover and percent bare ground on grass yield differs in such a way that, while
increase in basal cover results in an accompanied increase in grass yield, on the other
hand, increase in percent bare ground will decrease grass yield. These influences on
grass yield might be through their influence on the water status of the soil, which was
also discussed in section (5.4.3).
The positive correlation of grass yield with CN ratio could be associated with a better
ratio of C to N. Barnes et al. (1991) argued that a relatively narrow ratio of C to N in
the soil would favor N realiie from soil organic matter, while the negative
correlation of pH on grass yield was in agreement with the general indication
suggested by Barnes et al. (1991) that grass yield could be influenced by pH.
5.4. CONCLUS][ONS
Communal rangelands support cattle and other domestic livestock in a continuous
grazing system, with several pastoralists having free access to the same grazing area.
No clear management objectives have been stated for these areas, but the unwritten
objective is to maximize the number of animals owned by each individual.
Accordingly, the following conclusions are made from this study .
•:. The rangelands studied were dominated by C. plumulosus and the
different species of Sporobolus. Except at the junction site between the
Oromo and the Afars, the yield, basal cover and the grazing capacity of
the rangeland sites was low which agrees with the perceptions of the
pastoralists about the condition of the rangelands.
194
0:0 The studied districts are bush encroached notably with A. senegal, A.
nubica and P. juliflora (Awash-Fantale district). This finding is also in
line with the pastoralists concern of bush encroachment.
0:0 The results of this study support convectional wisdom, which states the
effect of high stocking rates are generally an undesirable changes in
species composition, reduced productivity and increase erosion (Pluhar et
al. 1987). Therefore, the study area must have a proper land use plan with
rangeland improvement practises such as bush control and resting of the
rangelands with stock exclusions for some period.
195
CJHJ:APTER 6
CONDiTION OlF THE COMMUNAL RANGELANDS
iN lRELATiON TO BENCHMAlRK SiTES
6.1. INTRODUCTION
A benchmark can be defined as a rangeland site (native pasture) with the best
possible botanical composition and cover (excellent condition) in relation to
prevailing climatic condition (Tainton 1999). Benchmarks are needed for rangeland
managers to determine the condition and trend of the vegetation, which can provide
pastoralists with information on the kind and amount of species that can be expected
when vegetation is protected from grazing. It is also important for rangeland
managers to be able to separate the impacts of grazing (or other consumptive uses)
on the species composition and productivity of vegetation from those caused by
climatic variations, explosions of insect populations or pollution (Taylor & Whalley
1976).
Danckwerts (1982) suggested that a benchmark should be an example of vegetation
that is considered to provide the highest possible sustainable animal production for
the rangeland type under consideration. This implies that it would have been well
managed in the past. Unfortunately, under the prevailing condition in Ethiopia, there
are few areas that can be classified as being in excellent condition. To overcome this
problem, in relative terms, the best sites in respect of vegetation condition were
identified in the study area, which could serve as a benchmark for this study.
Accordingly, protected grazing reserves locally known as 'kalo' and the fittle used
junction site between the Afars and the Oromos etnic groups were used for the study.
The use of multiple benchmarks has been recommended by Wilson (1984) to
minimize the bias arising from area selection. This study was, therefore, undertaken
to compare the protected plots (benchmarks) with the communal rangelands and
therefore get an insight into the condition of the grazing areas in relation to the
protected plots (Benchmarks).
196
6.2. PROCEDURE
6.2.1. Identification of benchmark sites
Three benchmarks sites were identified in the study area based upon similarity in
altitude, soil, and nature of the vegetation to the rangelands to which they will be
compared. Two of the benchmark sites were small (0.35 to 1.5 ha) area of land
protected by the Ororno pastoralists for use during the dry season, either by
harvesting or allowing the animals to graze. The two sites were protected for 5 and
15 years respectively. The third one is known as the 'land of fear' locally called
"Lafa soda" (Figure 6.1). This area is located at the junction between the Oromo and
Afar boundaries and has a very good vegetation cover for the mere reason that both
pastoral groups do not graze their animals freely and continuously in these area in
fear of prosecusion by the other group. As a result, the intensity of the grazing is low.
Six sampling sites (four from the Oromo and two from the Afars) were compared
with the benchmark sites.
Figure 6.1. Vegetation condition of the junction area between Afar and Oromo
ethnic groups.
197
The rainfall both at the benchmarks and the sampling sites ranged from 550 to 620
mm and the altitude from 950 to 1040 m.a.s.l. The major focus in this study was the
grass layer based on the rationale that the cattle and sheep are more affected by the
change in the vegetation compared to the camels and the goats, which are also
browsers.
Comparisons were made between benchmarks (Dhebeti and D/Galcha = Benchmark
1) and the sample sites (Madala 1, Kachachilo and Harole 1= First group). The second
group contained one benchmark (Dakaakae= Benchmark II) and four sampling sites
(Top of bud, Aleka, Harole 2 and Madala 2 = second group).
6.2.2. Data collection and analysis
In each of the benchmark and sampling site, 500-point observations was undertaken
using the wheel point apparatus to assess the herbaceous species composition. The
assessment factors and their measurements were carried out as described in Chapter
5 (i.e., herbaceous species composition, grass DM yield, basal cover, soil parameters
and estimated soil erosion). In addition, soil compaction was measured using a rod
penetrometer (ELE pocket penetrometer). At each of the sample site, three
representative plots of 1 m x 1 m were identified and watered artificially to a depth
of approximately 10 mm. Six hours after wetting, an estimation of the soil
compaction was obtained by penetrating to a depth of 6 mm of the topsoil surface
(Friedel 1987). Donaldson et al. (1984) and Donaldson (1986) waited for 18 hours
after wetting, while the measurements were taken six hours after wetting in this
study. The reasons were the sandy nature of the soil and the high ambient
temperature in the study area.
The herbaceous species composition; the percentage basal cover and the grazing
capacity were calculated according to the method described in Chapter 5. The
benchmarks and the sampling sites were compared for vegetation attributes
following the method described by Danckwerts (1982). The method of Danckwerts
. (1982) is preferred over that of Tainton (1999) since the latter imposes a limit to the
198
species composition of the sample site relative to the benchmark, while the former
does not. In this study, the benchmarks are relative and therefore it was not deemed
necessary to impose a limit on the species composition of the sample sites. The
species at the benchmarks and the sample sites were expressed in percentage
composition and were classified into highly desirable, desirable, less desirable and
undesirable species. The relevant weighting score of 10, 7, 4 and 1 for highly
desirable, desirable, less desirable and undesirable (Bare ground and forbs)
respectively, was multiplied with the percentage species composition of each species
to give the individual species score and these were added up for each sample site to
give the total score for the site. The total score of the site, expressed as a percentage
of the total score of the comparable benchmark, gave the percentage score of the site.
The rest of the data (grass DM yield, basal cover, texture, soil compaction,
percentage total Nand OC) were presented in graphs using Microsoft Excel (1997).
The benchmark sites were small in size and the sites could were not replicated, no
statistical analysis was carried out.
6.3. JRESUL TS
6.3.1. Rangeland condition assessment
The condition of the sample sites (Harolel, MadalaI and Kachachilo= group 1),
when calculated as a percentage of the average of the benchmarks (Benchmark = 1),
was lower than the benchmarks (Table 6.1) and it ranged from 65.58% (Madaal) to
90.08% (Harole 1). The condition of the second group of sample sites in relation to
the respective benchmark was 97.19% (Aleka), 90.96% (Harole2), 94.27%
(Madala2) and 99.09% (Top of bud). In terms of rangeland condition, Top of bud
was comparable to the benchmark (Table 6.2).
6.3.2. Grass dry matter yield and basal cover
The DM yield of the grass (kg ha") of the three sample sites, compared with the
average of Dire galcha and Dhebeti (Benchmark 1) was 37.73% (Kachachilo),
48.20% (Madalai) and 40.79% (Harole 1). This implies that the benchmarks had a
higher yield than that of the sample sites. Similarly, the yield of the grasses, for the
second group of sample sites, was 44.31% (Harole 2),52.36% (Madala 2),50.54%
199
Table 6.1. Rangeland condition scores in the first group of sample sites (Kachachilo,
Madala 1 and Harole 1).
Species Species Factor BMI BMI
category (Dhebeti) (Dire galcha) Kachachilo Madala I Harolel
010 score 010 Score 010 score 010 Score 010 Scor
e
Highly Chrysopogon lO 71 710 86.6 866 36.79 368 54.7 547 62.4 624.
desirable plumulosus
desirable Bare ground I I I 0.6 0.6 7.31 7.31 5.18 5.18 5.6 5.6
Less Coelachyrus 4 I 4 1.38 5.52 3.59 14.36 6.6 26.4
desirable Poiflorum
Less Sporobolus 4 2.8 15.2 8 32 48.61 194.4 0.4 1.6 20.6 82.4
desirable natalensis
Less Panicum 4 1.58 6.32 0 0 3.6 14.4
desirable snowdenii
Less Forb 1 1.6 1.6 2.96 2.92 0.8 0.8 0.8 0.8
desirable
Less EIeusine indica 4 0 0 0.2 2
desirable
Highly Cenchrus 10 0.4 4 0.98 9.8 1.6 16 0.2 2
desirable setigerus
Highly Cenchrus 10 0 0 0 0
desirable ciliaris
Desirable Heteropogon 7 20.4 142.8 0 0 0 0
contortus
Less Setaria 4 0.2 0.80 0.19 0.76 0 0 0 0
desirable verticellata
Desirable Cymbopogon 7 3.4 23.8 1 7 0.8 3.2 0
commutatus
Less Cynodon 4 2 8 0 0 0 0
desirable dactylon
Highly Tetrapogon 10 0 0 0 0
desirable cenchriformis
Highly Panicum 10 0 0 0 0
desirable coloratum
Desirable Chloris 7 0 0 0 0
roxburghiana
Desirable Paspalm 7 0 0 0 0
_glumaceum
Highly Chloris gayana 10 1.4 5.6 0 0
desirable
Less Aristida 4 0.2 0.8 0
desirable adscenionsis
Less Dactyloctenium 4 0 0 0 0
desirable aegypticum
Desirable Cymbopogon 10 0 0 0 0
excavatus
Less Sorghum 4 0 0 0 0
desirable purpureo
cericeum
Less Enneapogon 4 0 0 0 0
desirable schimperanus
Desirable Eragrostis 7 0 0 0 0
racemosa
Less Tragus 4 0.2 0.80 0.8 3.2 0 0
desirable benerontonus
Desirable Pennisetum 7 30.53 213.71 0 0
stramineum
Less Eragrostis 4 0 0 0 0
desirable cilianensis
Total score 100 900.8 100 916.0 100 595.8 100 811.45 100 758
Percentage 65.58 90.08 83.4
score
relative to
the
benchmark
200
Table 6.2. Calculation of rangeland condition in the study districts (Harole 2, Aleka,
Madala2 and top of bud).
Species Species Factor Dekaakae Harole Aleka Madala2 Top of bud
category 2
0/0 Scor % Score 0/0 Scor 0/0 Score 0/0 Scor
e e e
Highly Chrysopogon 10 43.6 436 73.1 731 84.3 843 71.23 712 85.3 853
desirable plumulosus 2
Less desirable Bare ground I 0.2 0.2 7.1 7.1 5 5 0.6 0.6 3.07 3.1
Less desirable Coelachyrus 4 2.6 10.4 6.4 25.6 2 8 0.09 0.36 1.06 4.2
Poiflorum
Less desirable Sporobolus 4 2.8 11.2 4.6 18.4 5.12 20.5 9.05 36.2 8.28 33.1
natalensis
Less desirable Panicum 4 0 4.4 17.6 1.4 5.6 0 0
snowdenii
Less desirable Forb I 1.2 1.2 0.8 0.8 1.58 1.58 0.8 0.8 0.66 0.7
Less desirable EIeusine indica 4 0 0 0 0.4 1.6 0 0
Highly Cenchrus 10 1.8 18 0.8 8 0.2 0.8 0 0 0.2 2
desirable setigerus
Highly Cenchrus 10 6.2 62 I 10 0.2 2 0.25 2.5
desirable ciliaris
Desirable Heteropogon 7 0 0
contortus
Less desirable Setaria 4 0 0
verticellata
Desirable Cymbopogon 7 0 0
commutatus
Less desirable Cynodon 4 0.4 1.6 0.6 2.4
dactylen
Highly Tetrapogon 10 15 150 0 0
desirable cenchriformis
Highly Panicum 10 9.8 98 0 0
desirable coloratum
Desirable Chloris 7 4.2 29.4 0 0
roxburghiana
Desirable Paspalm 7 1.6 11.2 0.2 1.4 0 0
giumaceum
Highly Chloris gayana 10 2 20 1.01 7.07
desirable
Less desirable Aristida 4 0.2 0.8 0.8 3.2
adscenionsis
Less desirable Dactyloctenium 4 0.2 0.8 0.6 2.4 0 0 1.16 4.64
Aegvpticum
Desirable Cymbopogon 7 5 35 0 0
commutatus
Cymbopogon 10 2.2 22
excavatus
Less desirable Sorghum 4 0.6 2.4 0 0
purpureo
cericeum
Less desirable Enneapogon 4 0.4 1.6 0 0
schirnperanus
Desirable Eragrostis 7 I 7 0 0
racemosa
Less desirable Tragus 4 2.82 1l.3
berteronianus
Desirable Pennisetum 7 10.79 75.5
strarnineum
Less desirable Eragrostis 4 2.01 8.04
cilianensis
Total score 100 912 100 829.3 100 886 100 859.5 100 903
Percentage 90.96 97.2 94.27 99.1
score relative to
the benchmark
201
(Top of bud) and 53.11% (Aleka) of that of the benchmark (Dakaakae). Similar to
the yield, the basal cover of the benchmark was higher than that of the sample sites in
both cases (Figure 6.2).
1200
'o:s
.c 1000
ee 800
C 600
32
Il) 400
~ 200
0
Benchmark and sample sites
30
23.9
i:' 25 22.1 22.1'u ,-..
o0:s .0- 20
o:s (/)
C) .-l
eo 15
o:s
.~ 10
ao:s C- 5
0
Benchmark and sample sites
,-.. 7
'$. 6
'-"
.....
Il) 5
:>
0 4
C)
ëij 3
en
es 2
..0
0
Benchmark and sample sites
Figure 6.2. Grass dry matter yield, grazing capacity and basal cover of sample sites
and the respective benchmark sites.
202
6.3.3. Grazing capacity
In both benchmark sites and across the sample sites (Figure 6.2), the grazing
capacity improved from as low as 23.9 ha LSU-1 (Kachachilo) to 8.2 ha LSU1
(BMI) and this result indicated that, given proper management of the rangelands,
there can be a considerable improvement in the grazing capacity of the rangelands.
6.3.4. Soil parameters
6.3.4.1. Percent sand, silt and clay
The benchmark sites (Benchmark 1) and their corresponding sample sites
(Kachachilo, Madalal and Harolel) in the first group had a comparable sand, silt and
clay fractions. It was also similar to the second group of sample and benchmark sites
(Figure 6.3).
6.3.4.2. Percentage total nitrogen
Compared with the benchmark sites (Benchmark I), sample sites (Kachachilo and
Madala I) had a lower percent total N, whereas, Harole I .had a slightly higher
percent N. Similarly, two of the sample sites (Harole2 and Top of bud) had slightly
lower N than their corresponding benchmark site and two of the sample sites
(Madala 2 and Aleka) had slightly higher percent N than Dakaakae (Figure 6.3).
6.3.4.3. Percentage Organic carbon (OC)
Kachachilo, Madalal and Harolel had lower percentage OC than their corresponding
benchmark sites (Benchmark 1). In the second group of sample sites (Harole2,
Madala2, Top of bud and Aleka), three of them had lower OC than the benchmark
site (Dekaakae), while Aleka had a slightly higher percent OC than the Benchmark
site (Figure 6.4).
203
50 _ _"IV
........ 40 33 r- r-- r-- r-- -
C 30 24 28
-g - -
24 26 r-
20 - -
"=
'" 10
Sample sites and benchmarks
4550 .---4-0------4-0-.~---,~-----4-2----------------------------_,r-- silt 40
I ,,40 r- r- - 34 38 36r- 32 - -
38
~ 35 r-
~ 30 - -
~ 25
20
15
10
o5 +-~L,~-L,_~~~~_.~-L~~L,-L~_r~~,_~L,~~~
Benchmark and sample sites
40 ...
.. 35 - ';2 ,;~ ';2 JJ
~ 30 - - r- r- 26 clay
'-...'... ')- 22 r- 22_)
~ 20 r- - 18 18
U 15 - r-
10
5
0
~;;-. ""v, .~c ~;- ~,,'. "'r-<V ~;;. '\. O~ ~">
~" ; o ~ c ~"> '."
.~~ ~""JJ CJ' 0">CJ'~ ~~ ~"><;. <:)"'r- ~"><;. 0">~~ " "'r-
~ ~'. +~
~'. ~
~ Benchmark and sample sites
0.2
C 0.16
.., 0.16
0.142
c: 0.129
eco 0.12
e 0.08
~s 0.04
0
Figure 6.3. Soil characteristics at the sample and benchmark sites in the study
districts.
204
6.3.4.4. Soil compaction
The compaction of the topsoil surface was lower at benchmark 1 than the
corresponding sample sites, namely, Kachachilo, Madalal and Harolel. Compared to
the average of the protected plots, the compaction at the sample sites was higher by
25.58% (Kachachilo), 26.58% (Madalal ) and 4.98% (Harole 1). In the second group,
two of the sample sites had slightly higher soil compaction than the benchmark site
and two of the sample sites had slightly lower soil compaction than the
corresponding benchmark site (Figure 6.4).
1.6
,-.,
;:R 1.4
I:....-
~ 1.2
-0e 1
0o0 0.8
.s~,!
00 0.6
00
CS 0.4
0.2
0
Benchmark and Sample sites
,-., 4.5 j. /15 .).01
00 4 ,..--- 3.57
~ 3.5
3.13
- 2.89
r-- 3.16 3.22 3.08 3.29
3 - r-- - -
,..--- 3.04
~ r-- r--
.9 2.5
t5 2
00
ê' 1.51
0
U 0.50
Benchmarks & Sample sites
Fgure 6.4. Soil organic carbon and compaction values in the sample and benchmark
sites in the study districts.
205
6.4. D][SCUSS][ON
6.4.1. Rangeland condition
With an objective of developing better and faster methods of rangeland evaluation,
numerous researchers have contributed to the refinement of the available techniques
as well as the variables that need to be measured for rangeland evaluation.
Accordingly, the benchmark method was developed as one method of rangeland
evaluation technique. However, limitations to the benchmark method have been
suggested by Mentis (1983); Barnes et al. (1984) and Martens et al. (1996). One of
the drawbacks of the rangeland evaluation techniques is that they usually lack
universal applicability. Thus, the choice of which method to use usually depends on
the local condition. Accordingly, an assessment of vegetation was undertaken using
the benchmark method. Except at Kacachilo, where the rangeland condition was low
relative to the benchmark sites, the rest of the sample sites had a comparable
condition with the benchmark sites. This could be due to the relatively short period
of years that the benchmarks have been protected. Therefore, the grass species are
more or less similar at the benchmark and the sample sites. It can also be concluded
that the current benchmark sites are not a good indication of the true potential of the
area.
6.4.2. Grass dry matter yield, basal cover and grazing capacity
Compared to the benchmark sites, the sample sites had a substantially lower grass
DM yield, basal cover and grazing capacity, which implies that there had been a
deterioration in the condition of the rangelands. The specific reasons for the
degradation are many and varied. However, the most likely reason for the
deterioration in rangeland condition manifested in lower yield, basal cover and
consequently reduced grazing capacity, is associated with overgrazing of the
rangeland by animals. The sample and the benchmark sites are found in the same
ecological unit, where one can assume that the influence of climate is the same for
both. Overgrazing and the associated decrease in basal cover and lowered grazing
206
capacity is well documented (e.g. O'Connor 1985; 0' Connor 1991; O'Connor 1994;
Snyman 1999a; Van der Westhuizen et al. 2001). Lazenby & Swain (1969), Cossins
& Upton (1985), O'Connor (1985) and Snyman (1999a) argued that overgrazing lead
to a decline in the quality and productivity of the rangelands. Though it is very
difficult to determine grazing capacity of rangelands under the pastoral production
system where communal ownership of the land predominate the management, it is
essential to estimate grazing capacity. This can serve as a guide for sustainable
rangeland utilisation. The improvement in the grazing capacity from the protected
plot is a clear indication that, given proper management of the rangeland, there is
room for improving the condition of the rangeland.
6.4.3. Nitrogen, Organic Carbon and soil compaction
With regard to percent total N, it was shown in the result that, in some cases, the
benchmark sites had a slightly higher N than the sample sites and in other cases the
sample sites had a slightly higher N than the Benchmark sites. Such slight difference
in soil is not unexpected as the soil can vary naturally within a short range.
According to Manley et al. (1995) the available data on the response of ac and N to
grazing does not indicate any single or consistent response to grazing and this might
be the reflection of the different environments, soil and grazing managements. The
slight increase in N in the sample sites compared with the benchmark sites can be
associated with the influence of animal excreta and the length of time the benchmark
sites have been protected. Lavado et al. (1996) argued that grazing influences the
flux of nutrients in grasslands in different ways (trampling, consumption, excreta
deposition and redistribution and export). An increase in nitrogen is related to excreta
returns, and losses are related to net export of nutrients or erosion. These processes
are a function of the stocking density and the initial nutrient status of the soils. In
most of the sample sites, the ac was lower than the benchmark sites and this could
be associated with the influence of grazing.
Compaction is a most complex soil feature having significant interrelationships with
most of the recognized physical, chemical and biological properties of soils as well
as with environmental factors such as climate (McKibben 1971). Both internal and
207
external factors affect the compaction of a given soil (Bennie & Krynauw 1985) and
of the internal factors affecting soil compaction, organic matter is one of them. In
general terms, the findings of this study indicated that those sites with a low OC
content had a higher compaction. Woodward (1996) argued that negative effects on
seedling establishment and growth are often attributed to changes associated with
soil compaction, including reduced water infiltration rates, decreased diffusive and
mass flow of nutrients and solutes through the soil, anoxia and increased resistance
to root penetration (Greacen & Sands 1980). The loss of vegetative and litter cover
with degradation (Warren et al. 1986; Thurow et al. 1988; Halm et al. 2002) allows
direct impact of raindrops on soil (Lal & Elliot 1994; Russel et al. 2001) and may
also produce hydrophobic substances that can reduce infiltration (DeBano et al.
1970, 1976; Emmerich & Cox 1992; Snyman 1999 a,b).
6.5. CONCLUSION
0:0 The most noteworthy differences between the benchmark and sample sites
were found to be in basal cover and grass yield, which indicates that, with
proper management, it is possible to improve the grazing capacity of the
study rangelands. Two of the benchmarks used in this study are plots
protected by the community and such efforts need to be encouraged and
supported technically. It is also possible to organize the community in the
form of cooperatives so that the community is involved in conservation
activities.
0:0 In both the benchmark and the sample sites, Chrysopogon plumulosus was
the dominant grass species. Therefore, future studies should focus on the
ability of this grass species to resist heavy grazing pressure.
0:0 The result of this study demonstrates that, given proper management like
resting of the grazing land, the rangeland will bounce back remarkably well.
208
ClillAJP1rJEJR 7
THE INlFLUENCE OlFTR.EES AND LIVESTOCK GRAZING ON
GRASS SPECIES COMPOSITION, YIELD AND SOIL
7.1. lINTRODUCTION
Trees may have negative and positive effects on their immediate environment and
the net result of the negative and positive interactions is dependent on tree density.
Established trees create sub-habitats, which differ from the open habitat and which
exert different influences on the herbaceous layer (e.g., Kennard & Walker 1973;
Stuart-HilI et al. 1987; Belsky et al. 1989; Frost & McDougald 1989; Smit & Swart
1994; Asferachew et al. 1998). Trees are also known to improve the soil status under
their canopies and thereby increase crop productivity when isolated trees are found
on the farm. Depending on tree density, they also reduce soil erosion, improve soil
fertility and reverse desertification (Nair 1984; Young 1989).
Compared to their neighbouring grasslands, soil under tree crowns have higher
concentrations of organic matter, higher concentrations of available nitrogen and
other important nutrients, better physical structure, and better water infiltration
(Tiedemann & Klemmedson 1973; KelIman 1979; Bernard-Reversat 1982; Belsky et
al. 1989; Young 1989; Vetaas 1992; Smit & Swart 1994).
Most studies related to the importance of tree canopies have been conducted in
lightly grazed areas and to this effect, work done in the Awash National Park of
Ethiopia under lightly grazed conditions on floristic composition and soil
characteristics has also indicated a similar pattern (Asferachew et al. 1998).
However, no information is available on the effects of trees along a grazing gradient,
which includes the communal grazing lands where the influence of livestock grazing
is also considered in Ethiopia. The only study that can be cited along this line is that
of Belsky et al. (1993) in Kenya. However, wide variability exists between sites and
countries. This study was, therefore, undertaken with the objective of studying the
209
effects of trees in light, medium and heavily grazed sites on grass species
composition, yield and some soil parameters.
7.2. PROCEDURE
7.2.1. Site, trees and sub-habitats selection.
Three sites having a similar altitude (957-1064 m.a.s.l), annual rainfall (550-620
mm) and sandy soils (FAO 1965; UNDP 1984) were identified for the study based
on their grazing histories and the presence of widely spaced, mature trees of Acacia
tortilis and Balanites aegyptica. The grazing histories were determined by
interviewing herders and employee of the Awash National Park who have I stayed'
long in the area. The inclusion of these two woody species is based on the fact that
previous studies had indicated that these tree species are important for the function of
the ecosystem (Asferachew et al. 1998). Two of the sample sites were located in the
Awash National Park of Ethiopia and the third site in the Kereyu pastoral grazing
land. The three sampling sites represented three different grazing gradients, namely
heavy, medium and light. The distance from the heavily grazed site to the medium
site was approximately about 25 km and from the medium to light site approximately
12 km. The three sites were characterised by an extended flat land and the dominant
tree species across three sites were A. tortilis, A. senegal and B. aegyptica, while the
dominant grass species were Chrysopogon plumulosus, Cymbopogon commutatus,
Cymbopogon exacavatus, and different species of Sporobolus. Legumes such as
Tephrosia subtriflora and Crotolaria spinosa are also found (FAO 1965) A brief
description of each of the sites is given below.
7.2.1.1. Lightly grazed site
The specific site representing light grazing is located about mid way between the
head office of the park (Gotu) and the Kereyu hotel. The site serves as a grazing area
for wildlife and is grazed only for a few months intermittently and thus the vegetal
cover and the composition is in good condition.
o
210
7.2.1.2. Medium grazed site
The specific site for the medium grazing was located about 16 km North East of
Mathara town, within the territorial zone of the Awash National Park on the opposite
side of the main gate of the park on the way to the hot spring (Filwoha). It is located
approximately 3 km from the main gate. This site was located adjacent to the
neighbouring pastoral grazing lands and because of the nearness of the area, illegal
grazing is more common throughout the year. Thus, this area serves as a grazing site
for both wild herbivores such as oryx., summering gazelle, warthog, rabbit, dik dik
and the pastoral's domestic animals such as cattle, donkeys and camels.
7.2.1.3. Heavily grazed site
The heavy grazed site was located in the Kereyu' spastoralists grazing land.
According to the information obtained from the surrounding pastoral people, this
area used to be open grassland with no woody vegetation and the herbaceous layer
consisted of very dense and tall grasses of diverse species, providing year round
grazing to the pastoral animals. However, the former excellent grass cover that used
to provide an excellent carpet to the soil gradually vanished due to continuous heavy
grazing. The animal species seen grazing were cattle, sheep, goats, donkeys and
camels, while game animals such as wart hogs, rabbits and dik diks were also
reported.
7.2.1.4. Tree and sub-habitat selection
At each of the grazing intensity sites, five trees of each species were selected
randomly and the heights and canopy diameters of these trees are given in Table 7.1.
In three of the sites, the data needed for the study was collected from two sub-
habitats, i.e., under the tree canopies (located directly beneath the tree crown) and in
the open grassland (located beyond the influence of the roots). Fieldwork was carried
out during the middle of August, 2001 at the time when most plants were in
flowering stage.
211
Table 7.1. Heights and canopy diameters of the tree species studied.
Tree species Height (m) Canopy diameter
(m)
Acacia tortilis 6.10 ± 0.60 1l.90 ± l.80
Balanites aegyptica 5.10 ±l.10 6.60 ± l.1
7.2.2. Sampling of herbaceous vegetation
Grasses and other herbaceous species were harvested in quadrats (0.5m x 0.5m),
randomly placed in each of the two sub-habitats (20 quadrats per sub-habitat) at the
time when the plants where approximately 50% flowering. Each quadrat was clipped
at the ground level using hand shears by species. The harvested material was
weighed using a portable scale, placed in paper bags and labelled. The samples were
subsequently sun-dried until such time that it was oven-dried at 105°C for 24 hours.
During sampling, care was taken that neighbouring trees, their shade and other
influences were avoided.
7.2.3. Soil sampling and analysis
Topsoil samples at two depths (0-5 cm, 5-10 cm), were taken at each tree, at ten
random locations per sub-habitat using a sharp knife and one meter tape. Initially, the
sample site was dug to a depth of 15 cm; after which the first 10 cm depth of the soil
was sub-divided into two parts by marking them with a sharp knife. Thereafter, the
soil sample was taken carefully by digging the respective depth. The samples were
kept in plastic bags, labelled, sealed and transported to the National soil laboratory in
Addis Ababa where it was analysed for pH, texture, electrical conductance (BC),
percent total nitrogen (N) and percentage organic carbon (OC). The method of
analysis was described in Chapter 5.
212
7.2.4. :Data analysis
The sample sites, were classified into 6, namely, heavy grazing canopy (1), heavy
grazing open (2), medium grazing canopy (3), medium grazing open (4), light
grazing canopy (5) and light grazing open (6). The sample sites and the grass species
recorded in each of the six sample sites were ordinated using DECORANA (Hill
1979). Furthermore, grass DM yield and soil parameters data were analysed across
sites (to compare under-canopy and grassland zones) by analysis of variance
(ANDV A) using a randomised complete block design with replications nested within
a site (SAS 1987). In the presence of significant F-values, treatment means were
separated by the least significant difference (LSD) method at the 5% significance
level.
7.3. lRJESULTS
7.3.1. Grass species composition
Fourteen grass species were identified at the heavy grazed sites (Table 7.2). The
grass species recorded under tree canopy and in the open grassland were more or less
similar and the yield of each species was also low (Table 7.2). Furthermore, most of
the grass species were annuals. The grass species composition in the medium and
light grazing plots, both under the canopies of the two tree species and in the open
grassland were more or less similar. The biggest difference between these two
grazing intensities lies in the presence of Panicum maximum under canopy. in the
lightly grazed plots (Table 7.2).
The result of the ordination indicated a clear separation of the sample sites at the
heavy grazing site from the medium and the light grazing sites (Figure 7.1). The
separation of the medium and the light grazing sites was not very distinct. A similar
trend was also observed in the grass species ordination (Figure 7.2).
213
Table 7.2. Grass species (yield kg ha") occurring under tree canopies and in open
grass lands in three sites having different grazing intensities in and near
Awash National Park.
Grazing site Under canopy Grassland zone
Species Yield kg ha' Species Yield kg ha
II
Heavy Tragus 239 Tragus 262
berteronianus berteronianus (A)
CA)
Sporobolus 255 Sporobolus 235
natalensis CP) natalensis CP)
Cenchrussetigerus 234 Cenchrus setigerus 234
CP) CP)
Eragrostis 315 Eragrostis 295
cilianensis CA) cilianensis CA)
Cynodon dactylon 313 Cyndon dactylon 281
CP) CP)
Eleusine indica CA) 217 Eleusine indica 186
CA)
Coelachyrum 243 Coelachyrum 199
poiflorum poiflorum
CA) (A)
Urochloa panicoides 255 Urochloa 233
CA) panicoides CA)
Digitaria temata 221 Digitaria temata 164
(A)
Chrysopogon 377 Chrysopogon 387
plumulosus CP) plumulosus
CA)
Dactyloetenium 206 Chloris pychnothrix 211
aegypticum (A) CP)
Aristida 162 Cenehrus eiliaris 188
adscensionis CA) CP)
Aristida adoensis 183
CP)
Mediwn Sporobolus 1 021 Sporobolus 1 809
natalensis CP) natalensis (P)
Eragrostis 1 548 Eragrostis 64
cilianensis CA) eilianensis CA)
Chrysopogon 2648 Chrysopogon 1257
plumulosus (P) Plumulosus CP)
Cymbopogon 1 749 Cymbopogon 952
commutatus (P) eommutatus CP)
Setaria vertieel/ata 818
CA)
Light Sporobolus. 2509 Sporobolus 1450
natalensis CP) natalensis CP)
Eragrostis 2818 Eragrostis 341
cilianensis (A) cilianensis CA)
Chrysopogon 3892 Chrysopogon 2295
- plumulosus CP) plumulosus CP)
Cymbopogon 2851 Cymbopogon 2571
eommutatus CP) commutatus CP)
Cymbopogon 3247 Lintonia nutans CP) 868
excavatus- CP)
Panicum maximum I 369
(P)
A= annual; P= Perenrual
,.
Ordill~liOJl 1'101
170 ~4
165
160
155
150
145
140
135
130
125
120
115
110
105
100
~ri:'e 6
90
·VxJ 85
Cd 80
-s0:: 75
o0 70 Sample 3
Q)
C/) 65 &
I
<é 60 Samp!.fahll}2le
U 55
Cl 50 G 61
45
40
35
30
25
20
15
1~0 I sar;le 5
0 20 40 GO 80 100 120 140 160 180 200 220 240
DCA-Fil'st axis
Figure 7.1. Ordination or sample sites by grazing intensity and sub-habitats. Key to the
numbers: I= Heavily grazed under canopy; 2= Heavily grazed open grassland; .N......
3= Medium grazed LIlHIer canopy; 4= Medium grazed open grassland; 5= Lightly -A.
grazed under canopy; 6= Lightly grazed open grassland.
urfllll;lIlOIlI'IOI
--15------- ------ ----,------- --
3BD I 9
360
340
320
300
2BD
260
240
220
200
." 160 0
;.<
« lGD I
-.-0... 140 30 120 5
U~ 100 14 10 6
IJJ
I BO ta o 7
~
U 860 ~
Q 40 9
20 IJ
0 12
·20 J3
4
·40
n
-GO
-60
-100
-120
-140 16
-lGD 20
-160 SI
-50 o ~,O 1(l(l 1~,O 2110 250 300
DCA-First axis
Figure 7.2. Oetrendcd correspondence ordination of grass species under different grazing
intensity in and around the Awash National Park of Ethiopia. Key to the number:
I = Tragus berteronianus; 2= Sporobelus natalensis 3= Cenchrus setigerus: 4=
Eragroslis ciliancnsis; 5= Cynodon dactylen; 6= Eleusine indica: 7= Coelachyrum
poiflorum; 8= Urochloa panicoides; S= Digiraria ternara; 10= Chrysopogon
plumulosus. II = Dactyloctcuium aegypticum; 12= Aristida adscensionis; 13= N,_...
Chloris pychnothrix 14= Cyrnbopogon conunututus; 1.5= Setaria vcrticellata; 16= Vl
Cymbopogon excavatus; 17= Lintonia nuians: I S= (',>n"hr"~ ,-;J; ... ; .. .: I<)= Ari~lirl~
216
7.3.2. Grass dry matter yield
7.3.2.1. Heavy grazed site
The DM yield of the grass at the heavily grazed site varied from 293.91 to 332.81 kg
ha-I. Neither tree species nor sub-habitat had a significant (P>0.05) influence on the
yield of the grasses (Table 7.3).
7.3.2.2. Medium grazed site
There was no significant (P>0.05) difference between the two tree species in terms of
their effect on the grass DM yield, while the yield of grasses (kg ha") was increased
(P<O.001) under tree canopies in comparison to that in the open grassland.
7.3.2.3. Light grazed site
Similar to the result found in the medium grazing site, the DM yield of grass was
significantly (P<O.00 1) higher under tree canopies than the corresponding open
grasslands. There was no difference (P> 0.05) between the two tree species in terms
of their effect on grass yield (Table 7.3).
Table 7.3. DM yield of the natural pasture (Mean & SE) under two sub-habitats
along different grazing intensities in and near Awash National Park (NS=
Non-significant; *** = P<O.OOl).
Grazing DM Yield of the grass layer (kg ha")
Site
Acacia tortilis Balanites aegyptica Tree Zone
species effect
effect (IProb)
(Prob)
Under Grassland Under Grassland
canopy Zone canopy Zone
Heavy 316.92 304.06 293.91 332.81 NS NS
(53.05) (64.39) (71.32) (85.39)
Medium 2051.41 1 149.62 1 1 326.66 NS ***
(92.33) (108.94) 904.21
Light 3 209.70 2350.49 2 2442.04 NS ***
(106.86) (86.44) 947.83 (97.24)
(91.68)
217
7.3.3. Soil parameters
7.3.3.1. Heavy grazed site
The effect of tree species, sub-habitat and soil depth on soil pH was not significant
(P>0.05) (Table 7.4), whereas the electrical conductance (EC) of the soil was higher
(P<O.OOl) under tree canopies compared to the open grasslands. An increase in soil
depth decreased (P<O.01) EC. The soil texture (sand, silt and clay percentage) was
not influenced (P>0.05) by sub-habitat or soil depth. The percentage total nitrogen
was higher (P<0.05) under tree canopies compared to the open and was decreased
(P<O.OOl) with an increase in soil, depth particularly for soil under tree canopies.
The organic carbon (OC) followed a similar pattern to that of percentage total
nitrogen. There was no difference (P>0.05) between the two tree species in terms of
their effect on the soil parameters studied.
7.3.3.2. Medium grazed site
At the moderately grazed site, no difference was observed between the tree species in
terms of their effect on the soil parameters studied except for Oc. Percentage organic
carbon was higher (P<0.05) under the canopy of A. torti/is than under the canopy of
B. aegyptica. Soil under tree canopies had a higher pH (P<O.Ol), EC (P<0.001),
percentage total N (P<O.Ol) and percentage OC (P<0.001), whereas the EC of the
soil, percentage total Nand OC (P<O.00 1) decreased with increase in the depth of the
soil. The texture of the soil was not influenced much (P>0.05) by either sub-habitat
or soil depth (Table 7.4).
7.3.3.3. Light grazed site
Soil under tree canopies had a higher pH (P<0.05), EC (P<0.05), total N (P< 0.05)
and OC (P<0.05) than their corresponding open grasslands, while the effect of depth
of soil on soil pH was non significant (P>0.05). The EC (P<O.OOl), percentage total
N (P<O.OOl) and OC (P<O.OI) decreased with an increase in soil depth. Similar to the
moderately grazed site, A. tortilis had a higher (P<0.05) OC than B. aegyptica. The
218
Table 7.4. Chemical properities of soils occurring under tree canopy and open grasslands at
two depths at different grazing intensity gradients in and near the Awash National
Park of Ethiopia (NS= non-significant; * = P<O.05; **= P<O.Ol; ***= P<O.OOl).
Grazing Soil Soil Acacia tortilis Balanites Tree Sub- Zone
site parameter depth aegyptica specie habitat effect
s effect difference
effect between
depths
Under Grass Under Grass
canopy land canopy land
zone zone
Heavy PH 0-5 cm 8.12 8.12 8.18 8.12 NS NS NS
5-10 cm 8.16 8.16 8.32 8.18
Medium PH 0-5 cm 8.26 8.16 8.24 8.24 NS ** NS
5- 10 cm 8.34 8.18 8.28 8.20
Light PH 0-5 cm 8.22 8.16 8.22 8.12 NS * NS
5- 10 cm 8.34 8.18 8.18 8.16
Heavy Electrical 0-5 cm 0.174 0.132 0.147 0.105 NS *** ***
conductance 5- 10 cm 0.126 0.096 0.106 0.09
(EC)
(ds m")
Medium (EC) 0-5 cm 0.23 0.136 0.221 0.147 NS *** ***
5- 10 cm 0.137 0.102 0.154 0.117
Light (EC) 0-5 cm 0.200 0.160 0.193 0.147 NS * ***
5- 10 cm 0.126 . 0.138 0.142 0.121
Heavy %sand 0-5 cm 54.00 53.20 53.20 53.00 NS NS NS
5- 10 cm 55.00 54.60 54.60 54.00
Medium %sand 0-5 cm 53.40 53.80 53.30 53.60 NS NS NS
5- 10 cm 53.80 54.60 53.80 54.20
Lightly %sand 0-5 cm 53.2 53.60 53.00 53.60 NS NS NS
5- 10 cm 53.4 53.40 53.60 53.80
Heavy % silt 0-5 cm 24.60 25.80 25.60 24.80 NS NS NS
5-10 cm 24.80 24.60 25.60 24.60
Medium % silt 0-5 cm 27.00 27.20 27.40 27.40 NS NS NS
5-10 cm 26.20 25.60 27.20 26.00
Light % silt 0-5 cm 27.00 27.00 27.00 27.00 NS NS NS
5-10 cm 26.00 27.00 27.00 26.00
Heavy %clay 0-5 cm 21.40 21.00 21.40 21.00 NS NS NS
5-10 cm 20.20 20.80 20.40 20.00
Medium % clay 0-5 cm 19.60 20.00 19.40 19.00 NS NS NS
5-10 cm 20.00 19.80 19.00 20.20
Lightly % clay 0-5 cm 19.60 19.20 19.60 19.60 NS NS NS
5-10 cm 20.40 20.00 20.00 19.80
Heavily % total N 0-5 cm 0.172 0.1526 0.172 0.154 NS * ***
5- 10 cm 0.099 0.1294 0.089 0.130
Medium % total N 0-5 cm 0.240 0.172 0.227 0.167 NS ** ***
5- 10 cm 0.105 0.140 0.106 0.149
Light % total N 0-5 cm 0.268 0.193 0.262 0.185 NS * ***
5-10 cm 0.135 0.183 0.160 0.160
Heavy %OC 0-5 cm 1.443 1.171 1.429 1.06 NS * ***
5-10 cm 0.845 0.936 0.657 0.81
Medium %OC 0-5 cm 2.052 1.452 1.912 1.408 * *** ***
5- 10 cm 0.912 1.202 0.971 1.132
Light %OC 0-5 cm 2.24 1.73 1.90 1.49 * * **
5-10 cm 1.43 1.52 1.52 1.39
'---------_._-------------------------------------_.
219
rest of the parameters were not influenced (P>0.05) by any of the variables
understudy ('Fable 7.4.).
7.4. DISCUSSION
7.4.1. Effect of grazing on grass layer and soil parameters
Many of the grass species identified at the heavily grazed site were annuals and
consisted mostly of less desirable species (R.efer Chapter 5). Species such as T
berteronianus, S. natalensis, E. cilianensis, C. dactylon, E. indica, C. poiflorum, U
panicoides and D. ternata were reported by others as indicators of deteriorating
rangeland condition (CADU 1976; Tainton et al. 1976; Ibrahim & Kabayu 1987;
Barker et al. 1989; Van Oudtshoorn 1999). The ordination results of the grass species
indicated a change in grass species composition as the grazing intensity increases.
The change in species composition and the associated decrease in grass DM yield
due to grazing effect is in agreement with the report of others (e.g., Frost et al. 1986;
Westoby et al. 1989; O'Connor 1994). The influence of grazing on grass species
composition and yield was discussed in detail in Chapter 5. The number of different
grass species found in the medium and lightly grazed was lower than that of the
heavily grazed, yet most of them were perennials and grasses with a better
desirability value than those found at the heavily grazed site. In relation to annuals,
perennial grass species are better indicators of ecological status of an area (du Plessis
et al. 1998; Snyman 1999a). In general, there was a tremendous increase in grass DM
yield with decreasing grazing intensity (1 608 kg ha-! for medium and 2 737.5 kg ha"
for lightly grazed sites. A higher yield of the natural pasture as the grazing intensity
decreases is in agreement with the report of Bel sky et al. (1989).
The results of this study further showed that, there was an improvement in soil
nutrient status as the grazing intensity decreases from heavy to light. Tiedemann &
Klemmedson (1973); Belsky et al. (1993); Campbell et al. (1994) made a similar
observation in their study. Under the condition of this study, the improvement in soil
nutrient status is mainly attributed to the effect of tree canopies, which is dicussed in
section 7.4.3.
220
7.4.2. Effect of tree species on the grass layer and soil
Except for OC at the medium and lightly grazed sites, the results of the study showed
that there was no difference between the two tree species in terms of their effect on
the studied parameters and this is in agreement with the reports of others relating to
soil (Belsky et al. 1993; Asferachew et al. 1998) and grass yield (Belsky et al. 1993).
Both tree species in this study are leguminous. Differences between leguminous and
non-leguminous species are reported in the literature (Smit & Swart 1994).
7.4.3. Effect of sub-habitats on grass layer and soil parameters
There are contradictory VIews regarding the influence of sub-habitats on grass
species compositon and yield. For example, Kennard & Walker (1973), Stuart-HilI et
al. (1987), Belsky et al (1989), Veenendaal et al. (1993), Smit & Swart (1994) and
Asferachew et al. (1998) reported a grass species turnover from under canopy to the
open areas, while, De Ridder et al. (1996) and Moyo & Campbell (1998) reported
that there was no effect of canopy on grass species. The result of the present study
revealed that there was no difference in grass species composition due to sub habitat
effect at heavy and medium grazing. However, at the lightly grazed site there was a
difference in grass species composition, which was reflected by the presence of
Panicum maximum under tree canopies. Interestingly, the result of this study adds
knowledge to both contradicting ideas with regard to grass species composition. The
existence of P. maximum exclusively under tree canopies agrees with the report of
others (Bosch & Van Wyk 1970; Kennard & Walker 1973). Bosch & Van Wyk
(1970) explained that this association is due to a high soil nutrient status under trees.
The increase in the grass yield could be attributed to the improvement in soil
physical properties (Elwell 1986) and exchange capacity of the surface soil due to the
improvement in soil nutrient status as a result of high volume of leaf fall under trees
(Kennard & Walker 1973; Campbell et al. 1994), droppings from birds and
mammals (Belsky et al. 1989) and the concentration of nutrients from areas beyond
the crowns by means of lateral tree roots (Tiedemann & Klemmedson 1973). The
221
other possible reason is the lower understorey solar radiation and soil and plant
temperatures (Tiedemann & Klemmedson 1973) resulting In reduced
evapotranspiration, which creates mesic under storey patches which are colonized by
palatable perennial grass species that have a higher water-use efficiency (Amundson
et al. 1995).
Soil under tree canopies had a higher pH at the moderate and light grazing sites, a
higher EC, OC and N at all levels of the grazing. The importance of canopies in soil
enrichment is well documented (Bosch & Van Wyk 1970; Kennard & Walker 1973;
Tiedemann & Klemmedson 1973; Kellman 1979; Bernard- Reversat 1982; Belsky et
al. 1989; Young 1989; Smit 1994; Smit & Swart 1994; Hagos 2001). Although, the
question and mechanism of soil enrichment under tree canopies remains largely
unexplained (Smit 2002), many theories have been presented, such as leaf litter from
leaf fall (Bosch & Van Wyk 1970; Stuart-Hill et al. 1987; Belsky et al. 1989), stem
flow and through fall (Kellman 1979; Williams et a!. 1987; Potter 1992).
There is a lack of consensus in published literature on the influence of trees on pH.
While Bosch & Van Wyk (1970); Kennard & Walker (1973); Palmer et al. (1988);
Young (1989) reported a higher pH under tree canopies, Belsky et al. (1989) and
Hagos (2001) reported a lower pH at the base of Acacia trees than further from the
trunk. The results of this study support the former view. Based on the positive
association between increases in exchangeable cations and soil pH (high base
saturation) (Barnard & Polseher 1972; Kennard & Walker 1973; Hatton & Smart
1984) a higher pH under canopies of savanna trees conforms more logically with the
higher content of exchangeable cations in this sub-habitat (Smit 2002).
7.4.4. Effect of so in depth on soil parameters
The effect of depth of soil, in this study, was observed in EC, percentage total Nand
percentage OC, which implies that there was a decrease in these soil parameters with
increase in soil depth and this agrees favourably with the report of others for EC
(Asferachew et al. 1998), N (Bernard-Reversat 1982; Asferachew et al. 1998) and
222
OC (Bernhard-Reversat 1982; Asferachew et al. 1998). Bernhard- Reversat (1982)
argued that oe and N content of soils of the savannas in north Senegal were
concentrated in the first few centimetres of soil and increased under tree canopies (A.
senegal and B. aegyptica).
7.5. CONCLUSION
This study investigated the effects of tree species and livestock grazing on the grass
layer and soil parameters. The following conclusions can be made from this study:
.:. Heavy grazing resulted in a decreased grass yield with grasses of less
desirability. There were also more species of annual grasses at the heavily
grazed site than at the medium and lightly grazed sites. There was a
significant improvement in grass DM yield as the intensity of grazing
pressure decreased .
•:. Heavy grazing also influenced the soil nutrient status. The percentage N,
percentage OC and electrical conductance were significantly lower at the
heavily grazed site compared to the lightly and medium grazed sites .
•:. Except for OC, there was no difference between the two tree species in
terms of their effect on grass DM yield and soil nutrient status .
•:. Tree canopy m semi-arid areas have often been found to create
micro habitats of improved soil physical and nutrient status. These
encourage the growth mesic, palatable and high yielding perennial grasses
in otherwise arid environment. Accordingly, the significance of the tree
canopy in this study was shown by the increase in grass DM yield,
improvement of the soil nutrient status and the creation of sub-habitat that
is suitable for important grass species P. maximum.
223
CHAPTERS
CONCLUDING REMARKS AND RECOMMENDATIONS
Pastoralism is a significant but declining econorruc sector In North East Africa
(Djibouti, Ethiopia, Somalia and the Sudan). Indigenous resource tenure systems in
Africa have evolved to meet the constraints and opportunities of often-difficult
biophysical environments, while facilitating the operation of complex spatial and
temporal land use patterns. Traditional systems provide security of tenure in
culturally relevant ways that permit adaptation to new circumstances. Pressure on
limited land resources and unsuitable environmental policies have led to a
pastoralist's crisis in the arid rangelands of Africa, for which the pastoralists
themselves are often blamed. The pastoralists are immediately and directly affected
by drought and famine. However, this is not the only reason for the decline in these
communities. Misconceptions have been translated into misguided policies or even a
failure to monitor and account for the sector in the central planning process. On the
other hand, imposed tenure structures in Africa have often not strengthened
individual rights and have blocked tenure development and adaptation in response to
new situations. Pastoralists in Africa have in particular been negatively affected by
the imposition of national tenure systems, which in many cases have often served to
marginalize nomadic populations with repercussions in terms of land degradation,
food security and instability. Land scarcity, changing land and resource use, unclear
land tenure and ownership arrangements create a complex array of problems. The
problems faced by the pastoralists living in the Middle Awash Valley of Ethiopia are
quite similar to those faced by many pastoralists in the arid areas of Africa.
This study has generated results by combining both aspects of social and biological
sciences through a questionnaire survey and the application of different rangeland
evaluation techniques to gain insight into the local people's understanding of changes
in their natural resource base and quantative data on the condition of the rangelands.
This combination of the two types of surveys tends to have a lower degree of bias
associated with the findings, as opposed to findings of pure ecological or socio-
224
economic studies. Thus, a combination of the two furnishes adequate information on
the condition of the rangelands, people's life style and the environment.
The general conclusion that can be made from the results of the four studies (the
pastoralists perceptions and the rangeland evaluation studies using different methods
and arguments), is that the condition of the rangelands has deteriorated. Although
there was variation from one rangeland site to another, generally the rangeland sites
had a low herbaceous basal cover and grazing capacity indicating that the rangelands
are in of need improvement. Furthermore, the ecological destruction will continue
unless corrective measures are taken in time. Encroachment of the rangeland with
woody plants of different species confirms the need for careful and participatory
interventions to improve the condition of the rangeland. Accordingly, the following
recommendations and scope for future research are made in order to improve the
situations existing in the study districts.
The problems facing the pastoralists in the Middle Awash Valley have been created
over many years and the solutions also require time. The solutions to the condition of
the existing rangelands require the clear commitment and full participations from the
side of the pastoralists, government and non-government organizations directly or
indirectly involved in rangeland resources utilisation, management, protection and
other related activities. A better understanding of the situation of the pastoralists and
the different elements of the ecosystem must be made and a co-ordinated strategy be
devised considering the different elements of the ecosystem in a such a way that one
is complementary to the other. There must be a multidisciplinary approach towards
the study of ecological stress, to look at demographic, sociological and political
factors, as the study area is also very important from a conservation point of view.
The conservation of the rangeland resource must receive top priority, whether the
land is used for conservation, livestock production or for that matter any kind of
production.
Pastoralists in the study districts acknowledge the vital role of trees and shrubs since
they provide a range of products and services In their daily life. Therefore, selective
225
protection and management of naturally occurring trees and shrubs, which are of
high importance for livestock feeding and other functions such as Acacia tortilis, A.
nilotica, Balanites aegyptica and others must be undertaken. To this effect, use of the
indigenous technical knowledge of the communities augmented by scientific
methods is crucial in the protection and rational use of woody and other plants in
general. The undiscriminating cut of trees for fuel wood and charcoal, particularly
along the high ways must be controlled through teaching the community about the
importance of conserving the trees and careful use of these woody plants in the
ecosystem. Management and future research should seek ways of regeneration of
preferred multi-purpose trees and shrubs. Future research should also focus on
developing of mechanisms in which disappearing valuable plants could be
propagated. Furthermore, focus must be given on the multiplication of drought
resistant fodder and tree species. The seeds of most important grass species, legumes
and tree species must be collected and conserved.
In order to improve the rangeland condition, the grazmg lands should be given
adequate rest, preferably with stock exclusion. This can be done in parts with the
effort of the community. Short-term protection from grazing would go a long way
toward restoring plant cover. Longer periods of protection would be needed because
rainfall is lower and vegetation might be less persistent in some rangeland sites. In
the lower rainfall areas the provision of seedling rests to improve the botanical
composition of rangeland has proved to be reasonably successful. Rests should be
designed both to promote seed production and to encourage the development of
seedlings. Such protective measures could be enforced by elders and village leaders
and should be adopted as part of a general management plan.
For rangeland improvement, the best is to stick with the most promising native
species of grasses and browse. Indigenous species has the advantage of proven
adaptation to the environment. Herbaceous exotics in particular, appear to have their
growth constrained by low rainfall. Therefore, the use of indigenous plants grasses
and browses must be given due attention if improvements are to be made in the
, rangelands.
226
Interventions must also be undertaken with forage improvement programmes. It is
contended that forage extension should focus on plants that help the people feed
themselves first. To this effect, the use of dual-purpose legumes such as cow pea and
intereropping of legumes with other food crops are important as a source of food for
human beings and feed for the animals. Research works undertaken in the lowlands
of Ethiopia like the Borana and the mid rift valley areas have shown these activities
to be viable and advantageous to the pastoralists. The improvement of grazing areas
through the introduction of high yielding and nutritionally quality species such as the
legumes can be considered along the river lines where usage of irrigation can be
possible. The study area is naturally endowed with many rivers passing through the
study districts and usage of irrigation is already under way for food crops.
Some members of the Oromo pastoralists are already acquainted with the protection
of small plots, locally known as 'kala' for use during the dry season, either by
harvesting or grazing by animals. The better option for calf feeding is the use of the
grass produced in the kala. Such a practice has been shown to be very important in
other rangelands of Ethiopia such as the Borana rangelands. In this study, kala sites
were used as benchmark sites and have proved that they are very important for
studying rangeland conditions as well. Therefore, such efforts of the community need
to be appreciated and encouraged. There is already a good will of the community to
get involved in such practices provided they are encouraged and supported
technically. Furthermore, such practises must be extended to the neighbouring
districts through education and field demonstration.
The yield of browse, their phenology and the DM yield of grasses is affected by
season, therefore such studies needs to be undertaken over years and during different
seasons. Studies related to the persistency of some of the grass species like
Chrysopogon plumulosus under the existing grazing condition in the study districts
need to be carried out. Furthermore, research should include rangeland ecological
monitoring to learn the trend in the vegetation. It is also important to distinguish
between climatic and man-made droughts. Although the pastoralists blame drought
for poor animal production, in most cases the main reason is poor rangeland
227
condition. Rangeland in poor condition will create pseudo-droughts. In addition, the
methods of rangeland assessment used in this study need to be evaluated and be
modified to suit the local condition if necessary. The results of the rangeland
condition study showed that interacting factors such as drought, overgrazing, and
high tree density affect the grass species composition, yield and basal cover of the
rangeland. Therefore, long-term experiments are required to study the relative
importance of each of these factors on the rangeland ecology of the study districts.
One of the areas that deserve attention should be the conservation of soil and water
resources. Soil and water conservation measures are, therefore needed in the grazing
lands. Use of trees and shrubs is a key way for assisting in the rehabilitation of
gullied areas and abandoned fields. Every encampment in the study districts had
tonnes of manure piled next to the corrals from years of corral cleaning. But this
huge amount of manure is not used. Therefore, education on the importance of using
manure as fertilizer and practical demonstration of its use must be undertaken in the
study districts. This manure at least can be used to fertilize land in the backyard
where dual-purpose legumes can be produced and the manure can also be used to
improve the soil status. Future research should also focus on the effect of soil on
vegetation distribution and/or vice versa.
Future water development programs should distinguish between water for human,
calf and other livestock use. The community highly appreciates the contributions
made by government and non-government organisations in water development
activities. Still there is an inadequacy of water supply, both in terms of quantity and
quality. Strong efforts must be made to develop methods of rainwater harvesting as
the water harvested in such way can be used for many activities, including irrigation
of pastures and consumption by livestock. Water development activities must be
increased with due consideration of the effect of establishing watering points on the
vegetation in Close proximity of the watering points.
The community's effort of regular monitoring of rangeland resource problems is not
backed up technically. Therefore, strong efforts must be made to assist the
228
pastoralists technically and by creating forums for discussion among the pastoral
groups. The two pastoral groups have developed a good culture along this line and
the involvement of all groups in the community in seeking solutions and appropriate
resource management practices should deserve attention. The role of women in
handling livestock and rangeland management activities needs to be appreciated and
encouraged. Such type of activities can be more effective in collaboration with the
concerned offices including the Environmental Protection Authority (EPA), the
agricultural offices and other NGO's involved in natural resource conservation. One
of the major problems for proper rangeland use is conflict between the two pastoral
groups. Accordingly, efforts must continue in reducing land use conflicts between
the different ethnic groups by promoting amiable sharing of resources.
From studies undertaken in Ethiopia and elsewhere it is well known that the presence
of woody plants in savanna is associated with positive and negative aspects, which is
closely related to tree density or tree abundance. Considering the decline in the
production potential of rangeland as a result of bush encroachment, it is a national
interest that encroacher plants are controlled. Combating bush encroachment,
however, will require a proper understanding of the mechanisms of invasion, which
has been adequately dealt in the thesis. Contrary to the common believe that bush
encroachment is detrimental to grazers, but not browsers, there are indications that
bush encroachment may also be detrimental to some browsers. High-density stands
may therefore not only be poorly suited to grazers because of reduced growth of the
herbaceous plants, but also to browsers because of their relatively poor browse
supplying characteristics (Smit 1999). Therefore, the control of encroacher plants
must be part of the rangeland management practice.
Control should start with key areas where young tree populations have invaded and
should not be indiscriminate but selective. The problem of bush encroachment in the
study districts can be tackled from two directions, i.e., prevention of bushes from
encroaching and the control woody plants in already encroached areas. In the former
situation, the preventive measures may include occasional hot fires and various
biological control methods such as browsing by browsers. A combination of fire and
229
biological methods will give better result than applying one of them only. Under the
condition of bush encroachment, the control measures that can be used are
mechanical methods combined with chemicals (herbicides) or mechanical methods
combined with repeated cutting or mechanical methods combined with heavy
browsing by goats. The last alternative seems more practical for the study area as it is
possible to use a high density of goats to browse a specific area. After applying
mechanical cutting the goats should graze the area at a high stocking rate. Heavy
grazing will prevent coppice-growth of the woody plants. Big areas can be divided
into smaller areas and the control methods applied accordingly.
Although, the problem of Prosopis juliflora is not wide spread in Awash-Fantale
district, the pastoralists are concerned that this plant will cover wide area in short
period of time. One of the reasons for its rapid expansion is seed dispersal by animals
feeding on the ripe pods and spreading the seeds in their dung. In order to prevent it
from spreading, the above-mentioned methods (hot fires and biological control
methods) can be used. Furthermore, it is preferable not to allow the animals to graze
the plant at the pod or seed stage. The pods (seeds) need to be collected by the
available labour (children) and it should be grounded before feeding it to animals. In
this way the seeds are destroyed, while utilising the high protein content of the pods
and seed. In areas where it is difficult to apply the mentioned methods, economic use
of the encroacher plants can be sought.
In all circumstances the application of bush control methods, the focus should be on
the bushes rather than the trees. The existence of the big trees have numerous
advantages of which one being their ability to suppress the growth of new seedlings
in addition to their stabilization role in the ecosystem. The big trees need to be
preserved. The rapid establishment of tree seedlings after the removal of some or all
of the mature woody plants may reduce the effective time span of bush control
measures. In many cases the resultant re-establishment of new seedlings may in time
develop into a state that is in fact worse than the original state. It is hypothesised a
more stable environment can be created, which is not as prone to the rapid
regeneration of new woody plants by making use of system dynamics (Smit et al.
230
1999). Here the natural functioning of the savanna system is allowed to stimulate the
development of an open savanna comprised mainly of large trees. It is based on the
principle that the distance between a tree and its neighbour of the same species is not
determined purely by chance, but that tree spacing is normally distributed. The larger
the individual, the greater is the distance between it and the nearest individual of the
same species. This is particularly noticeable with.Acacia species (Smith & Goodman
1986; 1987).
It is important for any land manager to realize that there is no quick solution to the
problem of bush encroachment. Effective management of bush encroachment should
not be considered a once-off event, but rather a long-term commitment. This may
involve alternative approaches that are not necessarily the simplest or cheapest. It has
been proven time and time again that the least expensive method of killing trees may
not be the most economical approach in the long term. It is also very important to
avoid or minimise other direct or indirect causes of bush encroachment. Of these
sound grazing management practise, which will ensure a vigorous and competitive
herbaceous layer, is of critical importance.
Development projects should help farmers to improve farming techniques and use
suitable seeds for the lowlands. There is a need to develop farming systems
according to the potential of the land. One of the areas that can be exploited in the
study districts is the development of Apiculture. Owing to the presence of different
kinds of woody plants this area can be a viable enterprise if well managed and can
provide extra income for the households.
The primary constraint on increasing the productivity of livestock in pastoral systems
is the acute shortage of feed during the dry season and the poor quality of what feed
is available. Owing to the abundance of browse plants and their primary role in
livestock feeding in the study districts, evaluation of the nutritive value of browse
species and other feed resources must be undertaken in the study area.
Supplementation strategy in the study area should focus on leguminous trees like
Acacia tortilis, A. nilotica, A. senegal, A. nubica, Balanites aegyptica and others.
231
The study districts are located in between many research centres. Despite this
advantage, studies have not been conducted to investigate the potential of the local
animals and improvement of the performance of the animals through breeding and/or
improved management practices. With the current positive attitude of the
government to the local pastoralists and with the emphasis given to the dry land,
research studies need to be undertaken on ways of improving the productivity of their
animals.
The community appreciates the efforts made by the government and the NGO's
operating in the study area in improving the veterinary services in the study area.
However, the supply of medications and other services is not adequate. This is
evident through sporadic heavy losses of small ruminants (particularly sheep).
Accordingly, improvement of animal health services and disease control measures,
particularly for calves and small ruminants, needs much attention. The use of local
plants as a source of medicine is practised in the community and research should be
undertaken on the importance of such activities.
One of the most interesting results obtained in the study, particularly on the side of
the Afar pastoralists was that their second most important source of income was from
the sale of milk and milk by-products. The milk by-products are processed using
local materials. These products, however, are easily perishable owing to the high
ambient temperature and the long distance that must covered to reach market places.
Therefore, efforts must be made to improve the milk products and by-products
processing techniques through educating and demonstrating better techniques and
equipments. To this effect, there is good expertise in the International Livestock
Research Centre (!LR!) in Ethiopia. Furthermore, the livestock marketing, marketing
infrastructures and supply livestock market information must be improved through
research and development in such a way that livestock marketing maximizes the
income of the pastoralists and at the same time reduce the pressure of the livestock
on the rangeland.
232
One of the major problems in developing the rangelands of Ethiopia in general is the
lack of adequate numbers of trained personnel in the different areas of rangeland
sciences. The country has a high potential to develop its resource base, be it in
livestock production, wildlife, crop production and some other agriculture related
areas as was reviewed in from literature. Yet, it should not be a surprise that most
development projects related to the rangelands are developed by expatriate staffs,
while the country should have developed it's own nationally capability rather than
relying on expatriate staffs. In most of the pastoral districts, it is unlikely to find even
one person trained in rangeland management. Therefore, it will be very difficult to
think of rangeland development in the absence of trained staff. Accordingly,
education must be given at different levels (from national to the grass root level in
different rangeland considerations, conservation, improvement and environmental
management and wildlife management). Therefore, having a college or other
teaching institution responsible to educate and teach at different levels in rangeland,
wildlife and related disciplines must be one of the top priorities. Owing to if s
numerous advantages like accessibility, proximity to the capital city, the existence of
the different elements of the rangeland and the pastoralists, the Awash National Park
can be one possibility. It is also an area where education on the conservation of
natural resources can be effectively taught. Human resource development must be
accompanied by creating a favourable environment for the staff and the provision of
necessary facilities.
Studies in many countries with pastoral communities and supported by the results of
this study have showed that the pastoralists in general have a good knowledge of
rangeland ecology. Nevertheless, owing to the internal and external problems facing
the pastoralists relying solely on indigenous knowledge of the pastoralists is not
adequate. Therefore, traditional experiences, skills and strategies accumulated by the
pastoralists over the centuries should instead complement modem scientific
knowledge rather than be ignored.
Linkages among development entities, research institutions, the pastoralists and other
non-government organisations working in rangeland resource management must be
233
strengthened. Solutions to the problems in the study area are possible through
collaboration between all parties involved. Rangeland policies based on solid
scientific and indigenous knowledge systems need to be formulated that can improve
the condition of the rangelands.
234
CHAPTER. 9
SUMMAlRY
Four studies were conducted in the Middle Awash Valley of Ethiopia in order to
determine the condition of the rangelands commonly grazed by the animals of the
pastoralists. The perceptions of the pastoralists about the rangeland resource must be
taken into consideration in order to develop possible intervention strategies to
minimize further degradation of the rangeland ecosystem. The two pastoral groups
(Afar and Oromo) live in an area characterised by resource use conflicts (state and
privately owned agricultural enterprises, conservation areas and use by the local
people). The study of the pastoralists perceptions was undertaken through group
discussions and interviewing of the individual households using a structured
questionnaires (90 households from the Oromo and 55 households from Afar). In the
second study, evaluation of the condition of the rangeland (l l sites in Awash-Fantale
and 10 sites in Kereyu-Fantale) was undertaken considering aspects of the
herbaceous layer, woody layer and the soil. The variables studied were grass species
composition, basal cover, estimated soil erosion, grass DM yield, woody species
composition, density, browse production and leaf phenology. The soil was analysed
for pH, texture, electrical conductance, percent total nitrogen, percent organic
carbon, available K and available P. In the third study, the condition of the rangeland
sites was evaluated using the benchmark method. Three benchmark sites (two
pastoral protected small plots and one junction site between the two neighbouring
districts) were used. Six sampling sites (four from the Oromo and two from the Afar)
were identified to be compared with the identified benchmark sites. In the fourth
experiment, the influence of tree species and livestock grazing on grass species
composition, yield and soil parameters was studied in three grazing gradients and
two sub- habitats,
The results of the study on pastoralists perceptions generally indicated that there was
deterioration in the conditions of the rangeland evident in an increase in bush
encroachment (mainly different species of Acacia) and a decrease in the abundance
235
of grasses and important legumes. In the case of the Afars, there was also an increase
in the abundance of poisonous plants. Drought and overgrazing were indicated as the
major causes of rangeland degradation. Furthermore, population pressure,
immigrants and the expansion of crop production were also among the factors
influencing rangeland condition. Poisonous plants of both woody and herbaceous
species were identified to affect livestock production in the study area. The
pastoralists, through their long-term tradition, have developed traditional methods of
intervention when toxins of plant origin poison an animal. However, these traditional
methods are not completely successful and need to be studied. Woody plants play an
important role in the farming system and have a wide range of uses. Some of their
uses were very interesting like the use of Grewia tenax bark to soften ladies hair by
soaking it in water overnight. Traditional people use woody plants as a source of
medicine for human and veterinary purposes. Such practices also need to be studied
in depth for better development in the area of ethno botany. The majority of the
pastoralists replied that the grazing lands are bush encroached notably by Acacia
Senegal, A. nubica and the other species Prosopis juliflora. Although it was difficult
for the pastoralists to clearly state the exact causes of bush encroachment, which
actually is also true for the scientific community, they have attributed possible
reasons that are acceptable within the scientific context. The pastoralists also
indicated that the intensity of conflict between them has increased owing to the
nutritional constraints of their animals. The main problem regarding livestock
production inclines more to nutritional constraints.
The results of the rangeland condition study indicated that in general, the dominant
grass species in the study districts were Chrysopogon plumolosus and different
species of Sporobolus. The rangelands sites were also characterised by a high
percentage bare ground, low basal cover, with a loss of soil from the surface and low
grazing capacity. The only exceptions were grazing sites at the junction between the
Oromo and the Afar pastoralists with relatively better cover values and higher grass
yields. The highest browsing capacity for camels and goats was contributed by A.
senegal and A. nubica. The soils around lake Beseka and Asobabad were found to
have a high electrical conductance, pH and low percent total nitrogen and organic
236
carbon. The results of the correlation matrix and the stepwise multiple regression
indicated that grass DM yield was affected by ETTE ha", basal cover, bare ground,
eN ratio and available K. The results of the analysis of the benchmark study revealed
that the sample sites had a substantially lower basal cover and grass DM yield. In
terms of grass species composition and other parameters, the benchmark and the
sample sites were comparable.
The results on the influence of tree species found under different grazing intensity,
revealed that the grass DM yield increased significantly with a decrease in the
grazing pressure. Areas under tree canopies subjected to medium and light grazing
pressures had a higher grass DM yield (P<O.OOl) than the corresponding open
grasslands. The soil nutrient status was also improved due the effect of sub-habitat
and with a decrease in grazing pressure. In general, for the sustainability of the
rangelands, overgrazing must be avoided and the rangeland given a rest period for
recovery. The scientific findings coincided with the pastoralists perceptions of their
environment, which shows that the pastoralists were quite aware of what is going on
in the rangelands.
237
REFERENCES
Abel NOJ 1993. Reducing cattle numbers on southern African communal range: Is it
worth it? In: Range Ecology at Disequilibrium. Behnke RH, Scoones I and Kerven C
(eds) Overseas Development Institute, London. 173-195 pp.
Abel NOJ 1997. Mis-measurement of the productivity and sustainability of African
communal rangelands: a case study and some principles from Botswana. Ecological
Economics 23: 113-133.
Adams ME 1967. A study of the ecology of Acacia mellifera, Acacia seyal and
Balanites aegyptica in relation to land clearing. Journal of Applied Ecology 4: 221-
237.
Addis A 1992. Ethiopia Livestock Development Strategy draft document,
consultancy report, Washington DC UNCEFfUNSO Project.
Aidle I'M 1988. Herbivory as a selecting agent on the timing of leaf production in a
tropical understorey community. Nature (London) 336: 574-575.
Ali M H 1993. The case of Afar. In: Edwads S & Tafesse M (eds) Conference on
pastoralism in Ethiopia, 4-6 February 1993, Ministry of Agriculture, Addis Ababa,
Ethiopia. 1- 4 pp.
Alemayehu M 1979. Grassland condition in Chilalo Awaraja, Arsi Administrative
Region, Ethiopia, an ecological study. Ethiopian Journal of Agricultural
Science 1(2): 116 - 125.
Alemayehu M 1998. The Borana and the 1991-92 drought: A rangeland and
livestock resource study. Institute for sustainable development, French Catholic
Committee against hunger and for development. Addis Ababa, Ethiopia. 90 pp.
Alemu T 1987. Grassland composition ofNekemte Awraja. Proceedings of the First
National Livestock Improvement Conference, 11-13 February 1987, Addis Ababa,
lAR, Addis Ababa.173-175 pp.
Amsalu S 2000. Condition of the major grazing areas in the mid rift valley of
Ethiopia. MSc thesis, University of Alemaya, Dire Dawa, Ethiopia.
Amundson RG, Ali AR & Belsky AJ 1995. Stomatal responsiveness to changing
light intensity increases rain-use efficiency of below-crown vegetation in tropical
savanna. Journal of Arid Environments 29: 139-153.
Anderson EW 1985. Percent composition verses absolute units of measure- a view
point. Rangeland: 247-248.
Archer S 1989. Have southern Texas savannas been converted to woodlands in
recent history? American Naturalist 134: 545-561.
238
Archer S 1990. Development and stability of grass/woody mosaics in a subtropical
savanna parkland, Texas, USA. Journal of Biogeography 17: 453-462.
Archer S, Seifres CJ, Bassharn CR & Maggio R 1988. Autogenic succession in a
subtropical savanna: conversion of grassland to thorn woodland. Ecological
Monographs 58: 111-127.
Archer S & Smeins lFE 1991. Eco-system level processes. In: Heitschmidt RK and
Stuth JW (eds). Grazing management. Timber Press, Portland, Ore. 45-75 pp.
Arntzen JW & Veenendaal EM 1986. A profile of environment and development
in Botswana: Gaborone: National Institute of Development Research and
Documentation (NIR), University of Botswana.
Asferachew A, Masresha IF& Zerihun W 1998. Investigations on canopy feature
of three indigenous woodland tree species of Ethiopia. SINET: Ethiopian Journal of
Science 21(1): 113-132.
Asmarom L :n.973.Gada. Three approaches to the study of African Society, The Free
Press, New York, USA. 340 pp.
Aucamp AJ 1976. The role of browsers in the bushveld of the Eastern Cape.
Proceedings of the Grassland Society of southern Africa Il: 135-138.
Australian Agricultural Management Company (AACM) 1984. Final drafts of
Ethiopian Livestock Subsector Review. Main report and annex. Vol. 3. 65 pp.
Ayana A & Baars RMT 2000. Ecological condition of encroached and non-
encroached rangelands in Borana, Ethiopia. African Journal of Ecology 38: 321-328.
Ayele GM 1986. Economic Adaptation and Competition for Scarce Resources: the
Mar in the North East Ethiopia. Unpublished MA thesis. Bergen: Department of
Social Anthropology, University ofBergen.
Azocar lP 1987. Habitos de pastoreo y de consumo de especies forrajeras del Ganado
caprino en zones aridas. Avances en production Animal 12(1-2): 3-9.
Azocar lP, Lailhacar S, Padilla F & Rojo H: 1.991. Methode d' evaluation de la
phytomasse utilisable des arbustes fourragers Atriplex repanda et Flourensia
thurifera. In: CIRAD (ed), Actes du Quatrieme Congres International des Terres de
Parcours, Montpellier, France: AFP. 512-514 pp.
Baars RMT, Chileshe EC & Kalokoni DM 1997. Technical notes: Range condition
in high cattle density areas in the Western Province of Zambia. Tropical Grasslands
31: 569-573.
Bailey RW 1945. Determining trend of range-watershed condition essential to
success in management. Journal of Forestry 43: 733-737.
239
Barker JR, Herlocker DJ & Young SA 1989. Vegetal dynamics along a grazing
gradient within the coastal grasslands of central Somalia. African Journal of Ecology
27: 283-289.
Barnard RO & Folscher WJ 1972. The relationship between soil pH and
exchangeable cations. Agrochemophysica 4: 53-56.
Barnes DL 1979. Cattle ranching in the semi-arid savannas of east and southern
Africa. In: Management of semi-arid ecosystems, Development in Agricultural and
Managed-Forest Ecology 7, Walker BH (ed). Amsterdam, Elsevier. pp 9-45.
Barnes DL, Beukes BH & Kotze GD 1985. A study of variations in floristic
composition of Highveld in the eastern Transvaal. Journal of the Grassland Society
of Southern Africa 2: 23-27.
Barnes DL, Kotze GD, Beukes BH & Kohlmeyer C 1987. Variations in floristic
and palatability composition, yield of veld and of soil characteristics on a paddock
scale. Paper presented at the 22nd Congress of the Grassland Society of Southern
Africa, January 1987.
Barnes DL, Lyoyd BV & McNeill L 1976. The use of shoot dimensions to estimate
the leaf mass or leaf area of certain indigenous trees in Rhodesia. Proceedings of the
Grassland Society of Southern Africa 11: 47-50.
Barnes DL, Rethman NlFG, Beakes BH & Kotze GD 1984. Veld composition in
relation to grazing capacity. Journal of the Grassland Society of Southern Africa 1:
16-19.
Barnes DL, Swart M, Smith MlF & Wiltshire GH 1991. Relations between soil
factors and herbage yields of natural grassland on sandy soils in the southeastern
Transvaal. Journal of the Grassland Society of Southern Africa 1: 16-19 8(3): 92-99.
Bartels GB, Norton BE & Perrier GK 1993. An examination of the carrying
capacity concept. In: Behnke RH, Scoones I & Kerven C (eds), Range ecology at
disequilibrium: New Models of Natural Variability and Pastoral Adaptation in
African Savanna. Overseas Development Institute, London. 89-103 pp.
Batabyal AA & Godlfrey EB 2002. Rangeland management under uncertainity: A
conceptual approach. Journal of Range Management 55(1): 12-15.
Bauer A, Cole CV & Black AL 1987. Soil property comparisons in VIrgm
grasslands between grazed and nongrazed management systems. Soil Science
Society of America Journal 51: 176-182.
Baxter P 19941.Pastoralists are people: why development for pastoralists, not the
development of pastoralism. Rural Extension Bulletin 4: 3-8.
240
Bayer W & Grell H 1994. Pastoralism fitting desertification. Entwicklung
Landlicher Raum 28(2): 6-9.
Bayer W & Grell H 1995. Pastoralism and the fight against desertification.
Agricultural Rural Development 2(1): 15-18.
Beasom SL & Haucke HH 1975. A comparison of four distance sampling
techniques in south Texas live oak mottes. Journal of Range Management 28: 142-
144.
Behnke RH 1994. Natural resource management in pastoral Africa. Development
Policy Review 12: 5-27.
Behnke RH, Seoones :n: & Kerven C 1993. Range Ecology at Disequilibrium: New
models of Natural variability and pastoral adaptation in African savannas. Overseas
Development Institute, London. 248 pp.
Bell lRHV 1982. The effect of soil nutrient availability on community structure in
African ecosystems. In: Huntley BJ & Walker BH (eds) Ecological studies 42:
Ecology of tropical savannas. Springer-Verlag, Berlin. 193-216 pp.
Belsky AJ 1990. Tree/grass ratios in East African savannas: a comparison of existing
models. Journal of Biogeography 17: 483-489.
Belsky AJ, Amundson R.G, Duxburg JM, Riha SJ, Ali AR &. Mwonga SM 1989.
The effects of trees on their physical, chemical and biological environments in a
semi-arid savanna in Kenya. Journal of Applied Ecology 26: 1 005-1 024.
Belsky AJ, Mwonga SM & Duxbury JM 1993. Effects of widely spaced trees and
livestock grazing on under story environments in tropical savannas. Agro forestry
Systems 24: 1-20.
Bernbridge T & Tapson D 1993. Communal livestock systems. In: Maree C &
Casey NH (eds). Livestock production systems: principles and practice. AGRI
Development Foundation, Pretoria. 361-373 pp.
Bennie ATP & Krynauw GN 1985. Review: Causes, adverse effects and control of
soil compaction. South African Journal of Plant and Soil2: 109-114.
Ben-Sharar R 1990. Soils, vegetation and herbivores in the Sabi-Sand game reserve,
South Africa. PhD thesis, Oxford University, Oxford.
Ben-Sharar R 1991b. Successional patterns of woody plants in catchment areas in a
semi-arid region. Vegetatio 93: 19-27.
Berg .lRH 1966. An evaluation of selected Nevada deer ranges: condition, forage
potential and deer-livestock competition. MSc thesis. University of Nevada, Reno.
241
Bernard-Reversat IF 1982. Biogeochemical cycles of nitrogen ill a semi-arid
savanna. Oikos 38: 321-332.
Beruk Y & Tafesse M 2000. Pastoralism and Agro-pastoralism: Past and present.
In: ESAP Proceedings: Pastoralism and Agro-pastoralism which way forward?
Proceedings of the 8th annual conference of the Ethiopian Society of Animal
Production, 24-26 August 2000, Addis Ababa, Ethiopia.
Bester IF & Reed JH: 1997. Bush encroachment: A thorny problem: Spotlight on
Agriculture. University of the Free State, No.1, Sep. 1997.
Blackburn WH, Knight RW & Wood MK 1982. Impact of grazing on watersheds:
A State of Knowledge. Texas Agricultural Experiment Station. MP 1496. College
Station, Texas, 32 pp.
Blackburn WH & Tueller ]PT 1970. Pinyon and Juniper invasion ill black
sagebrush communities in east-central Neveda. Ecology 51: 841-848.
Bodley JH 1985. Anthropology and contemporary Human problems. 2nd edition.
Palo Alto Mayfield Publishing Company, Mountain view, California Mayfield, 258
pp.
Bookhout TA 1965. Feeding coactions between snowshoe hares and white-tailed
. deer in northern Michigan. Transaction of the Northern American wildlife
conference 30: 321-335.
Borehert R 1980. Phenology and ecophysiology of tropical trees: Erythrina
poeppigiana O.F. Cook. Ecology 61: 1065-1074.
Borehert R 1994c. Induction of rehydration and bud break by irrigation or rain in
deciduous trees ofa tropical dry forest. in Costa Rica. Trees 8: 198-204.
Bosch om 1989. Degradation of the semi-arid grasslands of southern Africa.
Journal of Arid Environments 16: 165-175
Bosch om & Booysen J 1992. An integrated approach to rangeland condition and
capability assessment. Journal of Range Management 45: 116-122.
Bosch om & Gauch HG 1991. The use of degradation gradients for the
assessment and ecological interpretation of range conditions. Journal of the
Grassland Society of Southern Africa 8: 138-146.
Bosch om, Gauch HG, Booysen J, Gouws GA, Nel MV, Stols SHE & Van Zyl
E 1992. Integrated System for Plant Dynamics: user's guide. Potchefstroom
University for CRE.
Bosch om & Kellner K 1991. The use of degradation gradients in the ecological
interpretation of veld condition assessment. Journal of Arid Environments 21: 21-29.
242
Bosch OJB, Van Rensburg FPJ & Truter S du T 1987. Identification and
selection of benchmark sites on litholitic soils of the western grasslands biome of
South Africa. Journal of the Grassland Society of Southern Africa 4: 59-62.
Bosch ora & Van Wyk J.J1P 1970. Die invloed van bosveldbome op die
produktiwiteit van Pantcum maximum: 'n voorlopge verslag. Proceedings of the
Grassland Society of Southern Africa 5: 69-74.
Boshe .IT & Wegge JP 1992. Wildlife conservation in the Ngorongoo conservation
area in Tanzania: Social and ecological implications of increasing pastoralist and
declining per capita livestock populations. Mammalian conservation in developing
countries: a new approach. Proceedings of a workshop held at the 5th Theriological
congress, August 1989, Rome, Italy.
5-45 pp.
Bouder G & Riviere R 1968. Emploi pratique des analysis fourrageres pour
lappre' ciation des paturages tropicaux. Revue d' êlevage et de medecine veterinaire
des pays tropieaux 21 (2): 227-266.
Bourn :0 and Blench R 1999. Can livestock and wildlife co-exist? An
interdisciplinary Approach: Livestock, Wildlife and People in the semi-arid
rangeland of Eastern Africa. ODI Research study. Overseas Development Institute,
Portland House, Stag Place, London. 104 pp.
Bray RB & Kurtz ]LT.1945. Determination of total, organic and available forms of
phosphorous in soils. Soil Science 59: 39-54.
Breman H & Kessler J 1995. Woody plants in Agro-ecosystems of semi-arid
Regions: with an emphasis on the Sahelian countries. Springer-Verlag. Berlin.
270 pp.
Breman II, Van Keulen H & Ketelaars JJMH 1983. Land evaluation for semi-arid
rangeland: a critical review of concepts. In: Sideerius (ed) 229-244 pp.
Breman H & De Wit C 1983. Rangeland productivity and exploitation in the Sahel.
Science 221: 1341-1347.
Bryant JP, Reiehardt PB & Clausen TP 1992. Chemically mediated interactions
between woody plants and browsing mammals. Journal of Range Management 45:
18-24.
Bucher EH 1987. Herbivory in arid and semi-arid regions of Argentia. Revista
Chilena de Historia Natural 608: 265-273.
Buckman RF & Quintus RL 1972. Natural areas of the society of American
foresters. Society of American Forester 38 pp.
243
Butrington LC & Herbel CH 1965. Vegetational changes on a semi-desert
grassland range from 1858 to 1963. Ecological Monographs 35: 139-164.
Burrows WH 1980. Range management in the dry tropics with special reference to
Queensland. Tropical Grasslands 14: 281-287.
Burrows WH, Carter JO, Scanlan JC & Anderson ER. 1990. Management of
savannas for livestock production in northeast Australia: contrasts across the tree-
grass continuum. Journal of Biogeography 17: 503-512.
Cable DR 1975. Influence of precipitation on perennial grass production in the semi
desert southwest. Ecology 56: 981-986.
Cabral DR & West NE 1986. Reference unit-based estimates of winterfat browse
weights. Journal of Range Management 39: 187-189.
Campbell BM, Dore B, Luckert M, Mukamuri B & Gambiza J 2000. Special
section: Land use options in dry tropical woodland ecosystems in Zimbabwe:
Economic comparisons of livestock production in communal grazing lands in
Zimbabwe. Ecological Economics 33: 413-438.
Campbell BM, Frost r,King JA, Mawanza A & MhnallUgaL 1994. The influence
of trees on two contrasting semi-arid soil types at Matopos, Zimbabwe. Agroforestry
Systems 28: 159-172.
Caughley G 1976. Wildlife management and the dynamics of ungulate populations.
In: Coaker TH (ed). Applied Biology Volume 1. London: Acadaemic Press. 183-246
pp.
Cauldwell AE 1998. Management recommendations for Mtendere Game ranch,
Zambia, based on the ecology of fire and the vegetation. MSc (Wildlife
Management) disseration. University of Pretoria, Pretoria.
Central Statistical Authority (CSA) 1994. Summary reports on the 1994
population and housing census of Ethiopia. Various statistical abstracts, Addis
Ababa, Ethiopia.
eSA 1996. Agriculture sample survey 1995/96 (1988 EC), Report on livestock,
poultry and beehives population (private peasant holdings) Vol Vl. Statistical
Bulletin 152 eSA, Addis Ababa, Ethiopia.
eSA 2000. Statistical abstracts. Addis Ababa, Ethiopia.
Chidumayo EN 1990. Above ground woody biomass structure and productivity in a
Zambian woodland. Forest Ecology and Management 36: 33-46.
Chilalo Agricultural Development Unit (CADU) 1974. An illustrated guide to
grasses of Ethiopia. CADU, Asella, Ethiopia.
244
Child B 1995. The practice and principles of community based wildlife management
in Zimbabwe: the CAMPFIRE programme. Biodiversity and conservations
5: 369-398.
Child G 1995a. The role of community-based wild resource management 10
Zimbabwe. Biodiversity and conservation 5: 355-367.
Child G 1995b. Wildlife and people: the Zimbabwean success (How the conflict
between animals and people became progress for both). Harare, Zimbabwe and New
York, USA: Wisdom Foundation.
Child! RD, Heady HF, Peterson WC & Piper RD 1984. Arid and semi-arid lands:
sustainable use and management in developing countries. Winrock International:
Morriltion AR, USA. 205 pp.
Chopra SL & Kanwar 1976. Analytical Agricultural Chemistry. Kalyani University
Publishers, New Dehli.
elements ]FE 1916. Plant succession: an analysis of the development of vegetation.
Publication No. 242, Carnegie Institute, Washington.
CliflfElP 1971. Our research natural areas. American Forestry 77: 36-38, 53, 54.
Coe M & Coe C 1987. Large herbivores, Acacia trees and bruchid beetles. South
African Journal of Science 83: 624-635.
Coe M, Cumming D & Phillipson J 1976. Biomass and production oflarge African
herbivores in relation to rainfall and primary production. Oceologica 22: 341-354.
Coe M & Goudie AS 1990. The conservation and management of semi-arid
range1ands and their animal resources. Techniques for desert reclamation.
Environmental monographs and Symposia. John and Sons Ltd. Chichester, UK.
Cook CW, Taylor K & Harris LE 1962. The effect of range condition and intensity
of grazing upon daily intake and nutritive value of the diet on desert ranges. Journal
of Range Management 15:1-6.
Coppock DL 1994. The Borena plateau of southern Ethiopia: Synthesis of pastoral
research, development and changes, 1980-90. International Livestock Centre for
Africa, Addis Ababa, Ethiopia.
Coppock DL & Reed ID 1992. Cultivated and native browse legumes as calf
supplements in Ethiopia. Journal of Range Management 45(3): 231-237.
Cossins NJ 1985. The productivity and potential of the pastoral systems. ILCA
Bulletin 21: 10-15.
245
Cossins N & Upton M 1985. The productivity and potential of the southern
rangelands of Ethiopia. JESS (Joint Ethiopian pastoralists systems study). Draft
Summary report. International Livestock Centre for Africa. Addis Ababa, Ethiopia.
Coupland RT 1979. Grassland ecosystems of the world. Biological program, No.
18, Cambridge University Press, Cambridge.
COUSilllSB 1996. Range management and land reform policy in post- apartheid South
Africa. In: Land Reform and Agrarian Change in Southern Africa. An Occasional
Paper series, Vol. 2. Programme for Land Development and Agrarian studies,
University of Western Cape, South Africa.
Cox JJR. 1984. Tempertature, timing of precipitation and soil texture effects on
germination, emergence and seedling survival of South African love grasses. Journal
of South African Botany 50: 159-170.
Crouch GL 1968. Effects of simulated deer browsing on Douglas-fir seedlings.
Journal of Forestry 64: 322-326.
Cumming DHM 1990a. Wildlife products and the market place: a view from
southern Africa. Multi-species Animal Production Systems Project Paper 12, Harare,
Zimbabwe: World Wide Fund for Nature.
Cumming DHM 1990b. Developments in game ranching and wildlife utilisation in
Eastern and Southern Africa. Multi-species Animal Production Systems Project
Paper 13, Harare, Zimbabwe: World Wide Fund for Nature.
Cumming DHM 1990c. Communal land development and wildlife utilisation:
potential and options in northern Nambia. Multi-species Animal Production Systems
Project Paper 14, Harare, Zimbabwe: World Wide Fund for Nature.
Cumming DHM & Bond WE 1991. Animal production in southern Africa: present
practice and opportunities for peasant farmers in arid lands. Harare, Zimbabwe:
World Wildlife Fund for Nature: Multi-species Animal Production systems project.
Report to the International Development Research Center, Regional Office for
Eastern and Southern Africa, Nairobi, Kenya.
Dahlberg AC 2000. Vegetation diversity and change in relation to land use, soil and
rainfall-a case study from North-East District, Botswana. Journal of Arid
Environments 44: 19-40.
Dale IR & Greenway PJ 1961. Kenya trees and shrubs. Published by authority of
the government of the Colony and Protectorate of Kenya. Buchanan s Kenya Estates
Limited in association with the Hatchards, 187 Piccadilly, London, W.I. 654 pp.
Danckwerts JE 1982. Veld condition assessment in the Eastern Cape Region.
Dohne-Agric 5(1&2): 9-15.
246
.Danckwerts JE & Tainton NM 1996. Range management. Optimization of forage
production and quality. Bulletin of the Grassland Society of Southern Africa. 7: 36-
42 .
.Danckwerts .JE & Teague WR 1989. Veld management in the Eastern Cape,
Pretoria Pasture research section, Department of Agriculture and Water supply.
Eastern Cape Region .
Daniel K & Tesfaye K 1996. Oromiya Livestock Development Strategy. Addis
Ababa, Ethiopia .
Daubenmire R. 1940. Exclosure techniques in ecology. Ecology 21: 514-515 .
David W, Ngairorue BT, Apollus A & Tjieveze lEI2000. Perceptions and realities
of land degradation in arid Otjimbingwe, Nambia. Journal of Arid Environments 45:
337-356 .
Dawit A 2000. Pastoralism and pastoral production system. In: ESAP Proceedings.
Pastoralism and agro-pastoralism which way forward? Proceedings of the 8th annual
conference of ESAP, 24-26 August 2000, Addis Ababa, Ethiopia. 1-5 pp .
Day lPR 1965. Particle fractionation and particle size analysis. American Society of
Agromony. Agronomy No. 545-567 .
Dayton BR 1978. Standing crops of dominant Combretum species at three browsing
levels in the Kruger National Park. Koedoe 21: 67-76 .
Debano LF, Mann L.D & BamiUollU .DA 1970. Translocation of hydrophobic
substances in soil by burning. Soil Science American Proceedings 34: 130-133.
Debano LF, Savage SM & Hamilton DA 1976. The transfer of heat and
hydrophobic substances during burning. Soil Science American Journal40: 779-782.
De Beer lP, EK Harith A, Deng LL, Semiac-Santos SJ, Chantal B & Van
Grootheest B 1991. A killing disease epidemic among displaced Sudanese
population identified as visceral leishmaniasis. The American Journal of Tropical
Medicine and Hygiene 44: 283-289.
Dekker B 1997. Calculating stocking rates for game ranches: Substitution ratios for
use in Mopane veld. African Journal of Range and Forage Science 14(2): 61-67.
Dekker B & Smit GN 1996. Browse production and leaf phenology of some trees
and shrubs in different Colophospermum mopane savanna communities. African
Journal of Range and Forage Science 13(1): 15-23.
De Leeuw lPN, Grandin BE & Solomon B 1991. Introduction to the Kenyan
rangelands and Kajiado Disrict. In: Solomon B, De Leeuw PN, Grandin BE & Neate
PJH (eds.) Maasi herding: An analysis of the livestock production system of Maasai
247
pastoralists in eastern Kajiado District, Kenya. International Livestock Center for
Africa (Il.Cá), Addis Ababa, Ethiopia. 7-19 pp.
De Leeuw 1PN& Totbill JC 1990. The concept of rangeland carrying capacity in
Sub-Saharan Africa-myth or reality. Pastoral Development Network Paper 29b.
Overseas development Institute, London.
De lRidder CH, JarvellLC & Hobson lFO 1996. Using the dry-weight rank method
to determine the effect of Acacia karroo density on herbaceous species composition
by weight. Congress 31 abstracts, Grassland Society of Southern Africa,
Pietermaritzburg.
De Ridder N & Wagenaar KT 1984. A comparison between the productivity of
traditional livestock systems and ranching in Eastern Botswana. ll.,CA Newsletter 33:
5-7. Addis Ababa, International Livestock Center for Africa.
Derner ID, lBriske DD & Boutton 'fW 1997. Does grazing mediate soil carbon and
nitrogen accumulation beneath C4 perennial grasses along an environment gradient.
Plant and Soil 191: 147-156.
Deshmukh [][( 1984. A common relationship between precipitation and grassland
peak biomass for East and Southern Africa. African Journal of Ecology 22: 181-186.
Desta A 1993. Large-scale agricultural development and survival issues among
pastoralists in the Awash Valley. In: S.Edwards & Tafesse M (eds.). Conference on
pastoralism in Ethiopia, 4-6 February, MOA, Addis Ababa, Ethiopia. 67-72 pp.
Dhont AA 1989. Carrying capacity: a confusing concept. Acta Oecologia 9: 337-
346.
Dieckmann RC 1980. Ecology and breeding biology of the gemsbok Oryx gazella
(Linnaeus 1758) in the Hester Malan Nature Reserve. MSc thesis, University of
Pretoria, Pretoria.
Dikeni :U:, Moorehead R & Scoones ][ 1996. Land use and environmental policy in
the rangelands of Southern Africa: case studies from the Free State and Northern
Province. Land & Agriculture Ploicy Centre. Johannesburg Working paper 38.
Diress TA, Farah KO & Mbuvi DA 1998. Impact of land use on woody vegetation
in the semi-arid area of Abala district, North Afar, Ethiopia. The African pastoral
forum. University of Nairobi, Kenya.
Donaldson CH 1978. Some thoughts on the economics of bush eradiction with
special reference to the Blackthorn problem of the Molopo area. Unpublished report.
Roodeplaat Research Station, Pretoria, South Africa.
Donaldson CH 1986. Simple technique for estimating soil compaction. Karoo
Agriculture 3: 55-58.
248
Donaldson CH, Rootman G & Grossman D 1984. Long term nitrogen and
phosphorous application on veld. Journal of the Grassland Society of South Africa 1:
27-32.
Drummond RB & Moll E 1977. Trees of Southern Africa. C Struik publishers,
Cape Town, Johannesburg, South Africa. 959 pp.
Duncan J, Stow D, Franklin J & Hope A 1993. Assessing the relationship
between spectral vegetation indices and shrub cover in the Jornada Basin, New
Mexico. International Journal of Remote Sensing 14: 18,3 395-3416.
Du Plessis WP, Bredenkamp GJ & Trollope WSW 1998. Development of a
technique for assessing veld condition in Etosha National Park, Nambia, using key
herbaceous species. Koedoe 41(1):19-29.
Du Preez CC & Snyman HA 1993. Organic matter content of a soil in a semi-arid
climate with three long standing veld conditions. African Journal of Range and
Forage Science 10: 108-110.
Du Toit PF 1972. Acacia karroo intrusion: the effect of burning and spanng.
Proceedings of the Grassland Society of Southern Africa 7: 23-27.
Dye ]Pj & Spear ]PT 1982. The effects of bush clearing and rainfall variability on
grass yield and composition in south-west Zimbabwe. Zimbabwe Journal of
Agricultural Research 20: 103-118.
Dyksterhuis EJ 1949. Condition and management of rangeland based on quantative
ecology. Journal of Range Management 2: 104-115.
Dyson-Hudson RN 1980. Nomadic pastoralism. Annual Review of Anthroplogy 9:
15-61.
Edwards PJ & Coetsee G 1971. Growth, vigour and carrying capacity of veld: how
to determine in Farming. South Africa 47: 38-39.
Ekaya WN, Kinyamario Jl & Karue CN 2001. Abiotic and herbaceous
vegetational characteristics of an arid rangeland in Kenya. African Journal of Range
& Forage Science 18: 117-124.
ElkÏllS NZ, Sabol GV, Ward TJ & Whitfordl WG 1986. The influence of
subterranean termites on the hydrological characteristics of a Chihuahaun desert
ecosystem. Oecologia 68: 521-528.
Ellis J 1991. The role of wildlife in production systems in animal agriculture: Sub-
Saharan Africa: Study work group paper for Winrock International (1992):
Assessment of animal agriculture in Sub-Saharan Africa. Fort Collins, Colorado,
Colorado State University.
249
Ellis JE and Swift DM 1988. Stability of African pastoral ecosystems: alternative
paradigms and implications for development. Journal of Range Management 41(6):
458-459.
Ellison L, Lincoin ARA & Bailey R.W 1951. Indicators of condition and trend on
high range-watersheds of the Inter-mountain Region. State Department of
Agriculture Handbook 9.66 pp.
Elwell HA 1986. Determination of erodibility of a subtropical clay: a laboratory
rainfall simulator experiment. Journal of Soil Science 37: 345-350.
Emmerich WE & Cox JR. 1992. Hydrologic characteristics immediately after
seasonal burning on introduced and native grass-lands. Journal of Range
Management 45: 476-479.
Enghoff M 1990. Wildlife conservation, ecological strategies and pastoral
communities: a contribution to the understanding of parks and people in East Africa.
Nomadic Peoples 25-27: 93-107.
Environmental Poteetion Authority (EPA) 1998. Annual Report. Addis Ababa,
Ethiopia. 98 pp.
Ethiopian Agricultural R.esearch Organization (EAR.O) 2001. Dryland Research
Strategy Draft Manual. Pastoral and Agro-pastoral Development Strategy. (EARO),
Addis Ababa, Ethiopia. 65 pp.
Ethiopian wildlife conservation organization (EWCO) 2000. The Awash National
Park pamphlet. Addis Ababa, Ethiopia.
Evangelou, P. 1984. Livestock Development in Kenya's Maasailand: Pastoralists
transition to a market economy. Boulder, Colorado: Westview Press.
EVDSA 1990. Technical report on Lake Basaka water level rise. In: Schloeder C &
Jacobs M (eds). The Awash National Park Management Plan: 1993-97. Wildlife
Conservation Society and the Ethiopian Wildlife Conservation Organization, Addis
Ababa, Ethiopia. 56 pp.
Fabregues BP-de, Fabregues P de B, Fabregues De BP, Chapman GP 1992.
Observations on the ebb and flow of native grasses in the area of the Ekrafane
Ranch, Sahel. Desertified grasslands: their biology and management. Papers
presented at an International Symposium, London, 27-28 February and March 1,
1991. Linnean Society Symposium Series No. 13. 37-46 pp.
Falkengren-Greper U 1989. Effect of sternflowon beach forest soils and vegetation
in southern Sweden. Journal of Applied Ecology 26: 341-352.
250
Falkenmark M & Rockstrom J 1993. Curbing rural exodus from tropical drylands.
Ambio 22: 427-437.
Falkenmark M & Suprapto :RA 1992. Population landscape interactions in
development: A water perspective to environmental sustainability. Ambio 21: 31-36.
Food and Agricultural Organization (FAO) 1965. Survey of the Awash River
Basin. Reconnaissance of the soils of the Awash Basin, Middle Awash Valley Vol.
II-V.
FAO 1975. The ecological management of arid and semi-arid Africa and the near
and the Middle East. AGPC Misc/31. Report of an International Conference, Rome,
Italy.
FAO 1988. Guidelines: land evaluation for extensive grazing. Soil Bulletin. NO.58.
Rome, FAO.
FAO 1992. Ethiopia livestock sector development projects, reports 107/92. CP-ETH
44, YUP, August.
FAO 1995. FAO production year book, Rome, Italy.
Foran BD 1976. The development and testing of methods for assessing the condition
of three grassland types in Natal. MSc Agric. thesis. University of Natal,
Pietermaritzburg, South Africa.
Foran Bf), Bastin G & Shaw KA 1986. Range assessment and monitoring in arid
lands: the use of classification and ordination in range survey. Journal of
Environmental Management 22: 67-84.
Foran BD, Tainton NM & Booysen P de V 1978. The development ofa method for
assessing veld condition in the three grass veld types in Natal. Proceedings of the
Grassland Society of Southern Africa 13: 27-33.
Fortman L 1989. Peasant and official views of rangelands in Botswana: Fifty years
of deviation? Land Use Policy 6(3): 197-202.
Fourie .JH, De Wet NJ & Page JJ 1987. Veld conditions and trends in Kalahari
duneveld under an extensive stock production system. Journal of the Grassland
Society of Southern Africa 4: 48-54.
Fourie JH & Fouche' AHJ 1985. Weidingswaarde en ekologiese status van grasse
en karoo-bossies in die Vrystaatstreek. Glen Agriculture 14:14-18.
Fourie JH, Van Niekerk JW & Fouche HJ 1985. Weidingkapasiteitsnorme in die
Vrystaat-streek. Glen Agriculture 14: 4-7.
Fowler NL 1986. Microsite requirements for germination and establishment of three
grass species. American Midland Naturalist 115: 131-145.
251
Franklin JF & Trappe JM 1968. Natural areas: needs, concepts and criteria.
Journal of Forestry 66: 456-61.
Fratkin E 1997. Pastoralism: Governance and Development issues. Annual Reviews
in Anthropology 26: 235-261.
Frasier GW, Cox JR. & Woolhiser DA 1987. Wet-dry cycle effects on warm-
season grass seedling establishment. Journal of Range Management 40: 2-6.
Freudenberger KS & Wood E 2000. Cattle and cultivators: conflicting strategies
for natural resource management. ILEIA-Newsletter 16(1): 8-10.
Friedel MB 1987. A preliminary investigation of woody plant increase in the
western Transvaal and implications for veld assessment. Journal of the Grassland
Society of Southern Africa 4(1): 25-30.
Friedel MB 1991. Range condition assessment and the concepts of thresholds: A
view point. Journal of Range Management. 44(5): 422-426.
Frost r, Medina E, Menaut JC, Solbrig 0, Swift M & Walker n 1986.
Responses of savannas to stress and distrubance, Biology International, Special issue
10.
Frost WE & McDougald NK 1989. Tree canopy effects on herbaceous production
of annual rangeland during drought. Journal of Range Management 42: 281-283.
Futterknecht C 1995. People/park interactions in the Awash National Park:
opportunities and constraints. Presented for Participatory Wildlife Management
Workshop (16-18 May 1995), Addis Ababa, Ethiopia.
Gadamu F 1990. Pastoralist nomadism and rural development. Ethiopia: Rural
Development Options, Zed Books, London.
Galaty J & Johnson DL (eds) 1990. The World of Pastoralism. Guilford, New
York.
Gammon DM 1984. Effects of bush clearing, stocking rates and grazing systems on
vegetation and cattle gains. Vleisraadfokus May/June 3: 9-18.
Garcia-Moya E & McKell CM 1970. Contribution of shrubs to the nitrogen
economy ofa desert-wash plant community. Ecology 51:81-88.
Garland T 2000. Plant toxins. Journal of AOAC International 83(2): 455-459.
Gebral B 2001. Vegetation survey for the proposed Regional land use plan. MOA,
2001, Addis Ababa, Ethiopia.
252
Gentry AW 1981. Notes on Bovidae (Mammalia) from the Harar formation and
from Amado and Geraru Ethiopia. Kirtlandia no. 33. Kent State University Press. 30
pp.
Getadnew K 2001. Dimensions of pastoral poverty in Ethiopia. In: PFE Proceedings.
Proceedings of the second National conference on pastoral development in Ethiopia.
Poverty reduction strategy and pastoral development, May 22-23, 2001, Addis
Ababa, Ethiopia. 66-80 pp.
Gifford GF & Hawkins RH 1978. Hydrological impact of grazing on infiltration: A
critical review. Water Resource Research 14(2): 305-313.
Gilles JL 1982. Planning livestock development: themes from indigenous systems.
Agricultural Administration 11(3): 215-225.
Gilles JL & Jamtgaard K 1981. "Overgrazing in pastoral areas: the commons
reconsidered". Sociologia Ruralis 21(2): 129-141.
Glaser B, Lehmann J, Fuhrbëter M, Dawit S & Zech W 2001. Carbon and
nitrogen mineralization in cultivated and natural savanna soils of Northern Tanzania.
Journal of Biology Fertility Soils 33: 301-309.
Glendening GJE and Paulsen HA 1950. Recovery and viability of mesquite seeds
fed to sheep receiving 2,4-D in drinking water. Botanical. Gazette. 103: 486-491.
Glewll].(llemrning,G. 1941. Development of seedlings of Heteropogon contortus as
related to soil moisture and competition. Botanical Gazette 102: 685-698.
Glieck J 1987. Chaos: The making of a New Science. Cardinal, London.343 pp.
Geebel CJ & Cook CW 1960. Effect of range condition on plant vigor, production,
and nutritive value of forage. Journal of Range Management 13: 307-313.
Grandin B 1987. Pastoral culture and range management: Recent lessons from the
Maasiland. ~CA Bulletin 28: 7-13.
Greacen JEL & Sands R 1980. Compaction of forest soils. A review: Australian
Journal of Soil Research 18: 163-189.
Griffith D 1903. Forage condition of the northern border of the Great Basin. USDA
Bureau of Plant Industries. Bulletin 15.
Grootenhuis JE 1995. Wildlife utilization study. Report 4: Veterninary. Nairobi:
Conservation of Biodiverse Resource Areas Project: Kenya Wildlife Sevice and
African Wildlife Foundation.
Grossman D, Grunow JO & Theron GK 1980. Biomass cycles, accumulation rates
and nutritional characteristics of grass layer plants in canopied and uncanopied sub
habitats ofBurkea savanna. Proceedings of the Grassland Society of Southern Africa
253
15: 157-16l.
Grunow JO 1980. Feed and habitat preferences among some large herbivores on
African veld. Proceedings of the Grassland Society of Southern Africa 15: 141-146.
Gryseeis G & Anderson FM 1983. Research onfarm livestock productivity in the
Central Ethiopian highlands. Initial results: 1977-1980. ILCA Research Report No.4.
ILCA, Addis Ababa, Ethiopia.
Guevara S, Meave J, Moreno-Casasola P & Laborde J 1992. Floristic
composition and structure of vegetation under isolated trees in neotropical pastures.
Journal of Vegetation Science 3: 655-664.
Guy P.R, Mahlangu Z & Charidza H 1979. Phenology of some trees and shrubs in
the Senwa Wildlife Research Area, Zimbabwe-Rhodesia. South African Journal of
Wildlife Research 9: 47-54.
Gwynne MD 1969. The nutritive value of Acacia pods in relation to Acacia seed
distribution by ungulates. East African Wildlife Journal 7: 176-178.
Haan CD 1999. Future challenges to International funding agencies in pastoral
development: an overview. Proceedings of the Sixth International Rangeland
congress.
Hacker RB 1974. Some concepts in rangeland management. Journal of Agriculture
Western Australia 15(4): 76-80.
Hacker R. 1990. Range condition and shrub encroachment in the SORDU pilot
project area. Report commissioned by the Fourth Livestock Development Project of
the Ethiopian, Ministry of Agriculture. GRM International Pty Ltd. (ILCA library
accession no. 89189). 54 pp.
Hagos MG 2001. The influence of tree thinning and sub-habitat differentiation on
the reproductive dynamics of Acacia mellifera subsp. Detinens. MSc Agric. thesis,
University of the Free State, Bloemfontein, South Africa.
Halcrew S 1989. Ethiopian Valleys Development studies Authority: Master plan for
the development of surface water resources in the Awash Basin. Vol. 1. Executive
Summary.
Hall FC 1978. Applicability of rangeland management concepts to forest-range in
the Pacific Northwest. In: DN Hyder (ed). Proceedings of the First International
Range Congress, Society of Range Management, Denver, Colorado. 496-499 pp.
Hall-Martin AJ & Fuller NG 1975. Observations on the phenology of some trees
and shrubs of the Lengwe National Park, Malawi. Journal of the Southern African
Wildlife Management Association 5: 83-86.
254
Halls LK & Alcaniz R 1972. Growth patterns of deer-browse plants in southern
forests. US Department of Africa Forestry Service Pap. SO-75. 75 pp.
Halls LK, McCarty JD & Wiant HV 1970. Relative browsing of 16 species of
white-tailed deer. Journal of Range Management 23: 146-146.
Harbeson JW 1978. Territorial and Development Politics in the Horn of Africa: The
Afar of the Awash Valley. African Affairs 77: 479-498.
Harbenson JW 1990. Post drought adjustments among Horn of Africa pastoralists:
policy and institution building dimensions. Land Reform, Land Settlement and
Cooperatives. No 1-2: 15-27.
Hard in G 1968. The tragedy of the commons. Science 162: 1243-1248.
Hardly MB & Hurt CR 1989. An evaluation of veld condition assessment
techniques in highland sourveld. Journal of the Grassland Society of Southern Africa
6(2): 51-58.
Hardly MB & Walker RS 1991. Determining sample size for assessing species
composition in grassland. Journal of the Grassland Society of Southern Africa 8: 70-
73.
Harradine AR & Whalley RDB 1980. Reproductive development and seedling
establishment of Aristida ramose R. Br. in northern NSW. Australian Rangeland
Journal 2: 124-135.
Harrington GN, Winson AD & Young MD 1984. Management of Australia's
rangelands. Melbourne: CSIRO.
Batton JC & Smart NOE 1984. The effect of long-term exclusion of large
herbivores on soil nutrient status in Murchison Falls National Park Uganda. African
Journal of Ecology 22: 23-30.
Heard CAB, Tainton NM, Clayton J & Hardy MB 1986. A comparison offive
methods for assessing veld condition in the Natal Midlands. Journal of the Grassland
Society of Southern Africa 33: 70-76.
Heitschmidt RK, Schulz Rl) & Scifres CJ 1986. Herbaceous biomass dynamics
and net primary production following chemical control of Honey Mesquite. Journal
of Range Management 39:67-71.
Helem A 1993. The case of Afar. The case of Afar. In: Edwards S & Tafesse M
(eds), Proceedings of conference on Pastoralism in Ethiopia, 4-6 February 1993,
Addis Ababa, Ethiopia.
255
Hill MO 1979. DECORANA - a Fortran programme for detrended correspondence
analysis and reciprocal averaging. Ecology and systematic, CornelI University,
Ithaca, New York.
Hillman J 1993. Ethiopia: Compendium of Wildlife Conservation Information, Vol.
1. Wildlife Conservation in Ethiopia, New York Zoological Socciety/ Ethiopian
Wildlife Conservation Organization.
Hobbs RJ & Mooney HA 1986. Community changes following shrub invasion of
grasslands. Oecologia 70: 508-513.
l!Iogg RS 1997. Introduction: In: Hogg RS (ed) Pastoralists, Ethnicity and the State
in Ethiopia. Haan Publishing, London. 90pp,
l!Iolechek JL, Pieper RID & Herbel CH 1989. Range management principles and
practices. Englewood cliffs, New Jersey, Prentice Hall. 501 pp.
l!IoBm AM, Bennet LT, Loneragan WA & Adams MA 2002. Relationships
between empirical and nomial indices of land function in the arid shrubland of
Western Australia. Journal of Arid Environments 50: 1-21.
Homewoods KM & Rodlgers WA 1987. Pastoralism, conservation and the
overgrazing controversy. In: Anderson D & Grove R (eds) Conservation in Africa:
People Policies and Practice. Cambridge University Press, Cambridge. 111-128 pp.
Homewoods KM & Redgers WA 1991. Maasailand ecology: pastoralist
development and wildlife conservation in Ngorongoro, Tanzania, Cambridge:
Cambridge University Press.
Horewitz M & Little P 1987. African pastoralism and poverty: Some implications
for drought and famine. In: Glantz, M.(ed). Drought and Hunger in Africa: Denying
Famine a Future. Cambridge: Cambridge University Press. 20-75
Bughes HG, Varner LW & Blankenship LH 1987. Estimating shrub production
from Plant Dimensions. Journal of Range Management 40(4): 367-369.
Hulbert LC 1978. Natural area needs for range research. Proceedings of the 1st
International. Rangeland Congress Denver, Colorado. 263-265 pp.
Humphrey RR 1945. Some fundamentals of the classification of range condition.
Journal of Forestry 43: 646-647.
Humphrey RR 1946. Range condition: a classification of the sand-bunchgrass range
in the East Benton Soil Conservation District. US Soil Conservation Service, USDA,
Portland, Oregon.
Humphrey RR 1949. Field comments on the range condition method of forage
survey. Journal of Range Management 2: 1-10.
256
Humphrey RR. 1962. Range ecology. Ronaid Press, New York. 234 pp.
Hurt CR & Bosch om 1991. A comparison of some range condition assessment
techniques used in Southern African grasslands. Journal of the Grassland Society of
Southern Africa 8: 131-137.
Hurt CR & Hardy MB 1989. A weighted key species method for monitoring
changes in species composition of Highland Sourveld. Journal of the Grassland
Society of Southern Africa 6: 109-113.
Hurt OR, Hardy MB & Tainton NM 1993. Identification of key grass species
under grazing in the Highland Sourveld of Natal. African Journal of Range and
Forage Science 10(2): 96-113.
RUIssein K, Sumberg J & Sedldon D 1999. Increasing violent conflict between
herders and farmers in Africa: Claims and Evidence. Development Policy Review
17: 397-418.
Hyder DN, Bement RE, Remmega EE & Terwilliger C Jr 1966. Vegetation soils
and vegetation grazing relationships frequency data. Journal of Range Management
19: 11-16.
Ibrahim KM & Kabuye CRS 1987. An illustrated manual of Kenya grasses. Food
and Agriculture Organization of the United Nations, FAO, Rome, Italy. 763 pp.
Infield M 1996. Livestock production and wildlife conservation: opportunities for
compatible management and integrated production. East African Livestock
Conference, Small Ruminant Collaborative Research Support Program, University of
California, Entebbe, Uganda.
International Livestock Center for Africa (ILCA) 1990. Livestock systems
research manual. Vol.l. ILCA working Paper 1. ILCA, Addis Ababa, Ethiopia.
Irvine FR 1961. Woody plants of Ghana, with special reference to their uses. Oxford
University Press, London. 868 pp.
Jackson ML 1958. Soil chemical analysi. Prentice Hall, Inc., Englewood Cliffs,
New
Jersey. 183-204 pp.
Jahnke HE 1982. Livestock production systems and livestock development m
tropical Africa. Kieier Wissenschaftsverlag Vauk,' Kiel. 45-80 pp.
James AT, Reldon B & Robin Vanhorn Erect E 1991. Dependence of standing
crop on range condition rating in New Mexico. Journal of Range Management 44:
602-605.
257
James LF, Darwin BN & Panter KE 1992. Impact of poisonous plants on livestock
industry. Journal of Range Management 45: 3-8.
James DA 1967. The relationship of tree overstory and herbaceous vegetation.
Journal of Range Management 20: 247-249.
Janse Van Rensburg FP & Bosch om 1990. Influences of habitat differences on
the ecological grouping of grass species on a grazing gradient. Journal of the
Grassland Society of Southern Africa 7: 11-15.
Jansen DJ 1992. Cattle, wildlife, both or neither: a summary of survey results for
commercial ranches in Zimbabwe. Multi-species animal production systems project
paper 30. Harare, Zimbabwe: Worldwide fund for nature.
Jarman PJ & Sinclair ARE 1979. Feeding strategy and the pattern of resource
partitioning in ungulates. In: Serengeti: Dynamics of an Ecosystem. Sinclair ARE &
Norton-Griffiths N (eds) University of Chicago Press, Chigago. 130-163 pp.
Jerry LH, Carton Rex DP & Herbal H 1989. Rangeland Physical Characteristics.
In: Range management principles and practice.
Johannson DC & White TD 1979. A systematic assessment of Early African
Hominids. Science Vol. 203, No. 4378: 321-330.
Johnsen N 1998. Maasai in a tourists' paraside: a process of poverty and
sedentarisation. Indigenous Affairs 2: 10-19.
Johnson DE, Berman MM, Ben-Ali MN & AIi-MN-ben 1992. Evaluation of plant
species for land restoration in central Tunisia. Journal of Arid Environments 4: 305-
322.
Jordaan FP, Biel LC & du Plessis PIM 1997. A comparison of five range
condition assessment techniques used in the semi-arid western grassland biome of
Southern Africa. Journal of Arid Environments 35: 665-671.
Kabaija E & Little D 1987. Composition of native mineral supplements in Ethiopia.
FNE (Forage Network in Ethiopia) Newsletter 15: 24-28. ILCA, Addis Ababa,
Ethiopia.
Kaitho RJ 1997. Nutritive value of browses as protein supplement (s) to poor
quality roughages. PhD dissertation. Wageningen Agricultural University,
Netherlands.
Kamara CS & Haque I 1992. Faidherbia albida and its effect on Ethiopian
highland vertisols. Agroforestry Systems 18: 17-29.
Kamugisha JR & Staid M 1993. Parks and people pastoralists and wildlife:
environmental degradation in and around lake Mburo National Park, Uganda.
258
Technical Report Number 7. Nairobi Regional Soil Conservation Unit, Swedish
International Development Authority.
Kay BL & Leonard OA 1980. Effect of blue oak removal on herbaceous forage
production in the north Sierra foothills. Proceedings of the Symposium on the
Ecology, Management and Utilization of California Oaks. Pacific southwest Fores
and Range Experiment Station, Berkeley, California. 323-328 pp.
Kellman M 1979. Soil enrichment by neotropical savanna trees. Journal of Ecology
67: 565-577.
Kelly JRD 1973. A comparative study of primary productivity under different kinds
ofland- use in south-east Rhodesia. PhD thesis. University of London.
Kelly RD & Walker BB 1976. The effect of different forms of land use on the
ecology of a semi-arid region in south-eastern Rhodesia. Journal of Ecology 64: 553-
576.
Kennan TeD 1969. The significance of bush in grazing-land in Rhodesia. Rhodesia
Science News 3: 331-336.
Kennard DG & Walker BB 1973. Relationship between tree canopy cover and
Panicum maximum in the vicinity of Fort Victoria. Rhodesia Journal Of Agricultural
Research 11: 145-153.
Kipuri N 1995. Socio-ecomonic concerns on sustainable use and management of
semi-arid lands: the case of pastoral lands of East Africa. In: Garry F & Campbell
BM (eds) Sustainable land management in African semi-arid and subhumid regions:
Proceedings of the Scope workshop, 15-19 November 1993, Dakar, Senegal.
Montpellier, France: CIRAD. 273-282 pp.
Kirkman KP 1995. Effect of grazing and resting on veld productivity. Bulletin of
the Grassland society of Southern Africa 6(2): 9-11.
Kirkman KP & Moore A 1995a. Veld management strategies for livestock farmers
in the sourveld regions of Southern Africa. Bulletin of the Grassland Society of
Southern Africa. 6(2): 25-28.
Kirkman KP & Moore A 1995b. Towards improved grazing management
recommendations for sourveld. African Journal of Range and Forage Science. 12(3):
135-144.
Kirkman KP & Moore A 1995c. Quantification of the effects of defoliation on
grass vigour. African Journal of Range & Forage Science 12: 89-91.
Kiss A 1990. Living with wildlife: Wildlife resource management with local
participation in Africa. Technical Paper 130. Washington DC, The World Bank.
259
Knierim PG, Carvell KL & Gill JD 1971. Browse in thinned oak and cove
hardwood stands. Journal of Wildlife Management 35: 163-168.
Knoop WT 1982. Interactions of woody and herbaceous vegetation in two savanna
communities at Nylsvley. MSc thesis, University of Witwatersrand, Johannesburg,
South Africa.
Knoop WT & Walker BH 1985. Interactions of woody and herbaceous vegetation
in two savanna communities at Nylsvley. MSc thesis, University of the
Witwatersrand, Johannesburg, South Africa.
Kozlowwski TT & Keiler T 1966. Food relations of woody plants. Botanical
Review 32: 293-381.
Kutilek M 1979. Forage-habitat relations of nonmigratory African ungulates In
response to seasonal rainfall. Journal of Wildlife Management 43: 899-908.
Lal R & Elliot W 1994. Erodibility and erosivity. In: Lal (ed) Soil Erosion Research
Methods (2nd ed) Soil and Water Conservation Society, Ankeny, LA. 182-208 pp.
Lamprey H 1963. Ecological separation of the large mammal species m the
Tarangire Game Reserve, Tanganyika. East African Wildlife Journal I: 63-92.
Lamprey HlF 1983. Pastoralism yesterday and today: The overgrazing Problem. In:
Bourliere, F. (ed)Tropical savannas. Ecosystems off the world, vol. 13. Amesterdam:
Elsevier press.
Lane C, Futterknect C & Jarso L 1993. Assessment of land related issues in the
Awash National Park. Consultation draft. lIED and CARE. 120 pp.
Lane C & Moorehead R 1994. Who should own the range? New thinking on
pastoral resource tenure in dryland Africa. London: International Institute for
Environment and Development.
Lane C & Pretty J 1990. Displaced pastoralists and transferred wheat technology in
Tanzania. Gatekeeper series. S.A. 20. London: International Institute for
environment and Development.
Lane C & Swift J 1989. East African pastoralism: common land common problem.
Drylands Issues Paper 8. London: International Institute for Environment and
Development.
Lane CR 1998. Custodians of the Commons: Pastoral Land Tenure in East and West
Africa. Earthscan Publications Ltd., London.
Lauenroth WK & Laycock WA 1989. Secondary succession and the evaluation of
rangeland condition. Westview Press, Boulder, Colo.
260
Lavado RS, Sierra JO & Hashimoto PA 1996. Impact of grazing on soil nutrients
in a Pampean grassland. Journal of Range Management 49: 452-457.
Laycock WA 1991. Stable states and thresholds of range condition on North
American rangelands: a viewpoint: Journal of Range Management 44: 427-433.
Lazenby A & Swain FG 1969. Intensive pasture production. Commonwealth
Agricultural Bureaux, England.
Le Houerou HN 1980. The role of browse in the Sahelian zones. In: Le Houerou
RN (ed) Browse in Africa. ILCA, Addis Ababa, Ethiopia.
Leithead HL 1959. Runoff in relation to range condition in the big Bend-Davis
Mountain section of Texas. Journal of Range Management 12: 83-87.
Lenden C & Lamacraft RR 1976. Standards for testing and assessing range
condition in central Australia. Australian Rangeland Journal I: 40-48.
Livingstone I 1977. Economic irrationality among pastoral peoples: myth or reality?
Development and change 8: 209-230.
Lusigi W & Acquay H 1999. Challenge facing pastoralists in the drylands of Sub-
Saharan Africa. Proceedings of the International Rangeland Congress.
Lutge BU, Hardy MB & Hatch GIP 1996. Plant and sward response to patch
grazing in the Highland sourveld. African Journal of Range and Forage Science
13(3): 94-99.
Macnab J 1985. Carrying capacity and related slippery shibboleths. Wildlife Society
Bulletin l3: 403-410.
MacOnochchie JR & Lange RT 1970. Canopy dynamics of trees and shrubs with
particular reference to arid-zone topfeed species. Transaction Research Southern
Australia 94: 243-248.
Macpherson CNL 1995. The effect of transhumance on the epidemiology of animal
diseases. Preventive Veterninary Medicine 25: 2l3-224.
Maddock L 1979. The migration and grazing succession. In: Serengeti: Dynamics of
an ecosystem. Sinclair ARE & Norton-Griffiths M (eds) University of Chigago
Press, Chigago. 104-129 pp.
Manley JT, Schuman GE, Reeder JD & Hart RH 1995. Rangeland soil carbon
and nitrogen response to grazing. Journal of .Soil and Water Conservation 50 (3):
294-298.
261
Martens JC, Danckwerts JE & Zacharias PJK 1996. Species responses to grazing
in the Smaldeel area of the Eastern Cape. African Journal of Range and Forage
Science 13: 29-36.
Mason LR & Hutching SS 1967. Estimating foliage yields on Utah juniper from
measurements of crown diameter. Journal of Range Management 20: 161-166.
Maree C & Casey NH 1993. Livestock production systems. Principles and practice.
Agri-development foundation. Brooklyn.
Man W & Wiegand D 1987. Limits of traditional veterinary medicine in Somalia:
The example of chlamydiosis and q-fever. Animal Research and Development 26:
29-34.
Mathars Estate Sugar 2001. Annual report. Methara, Ethiopia.
Mathesen W & Ringrose S 1994. Assessment of degradation features and their
development into the post-drought period in the westcentral Sahel using Landsat
MSS. Journal of Arid Environments 15: 12-26.
Mathias-Mundy E & McCorkle CM 1989. Ethnoveterninary medicine: An
Annotated Bibliography. Bibliographies in Technology and Social change 6.
Technology and Social Change Program, Iowa State University, Ames, Iowa, USA.
199 pp.
Mathieu P 1995a. 'Conclusion: Les conditions de la gestion locale de
I'environnement, Cahiers du CIDEP 27, November.
Mathieu P 1995b. 'Rehabilitation et cogestion des ressources naturelles dans la
vallee du fleuve Niger, dans le Gourma malien (1986 -89)', Cahiers du CIDEP 27,
November.
McIntosh RP 1985. The background of ecology. Cambridge University Press,
Cambridge, Mass. 383 pp.
McKenzie FR 1996. The influence of grazing frequency and intensity on the vigour
of Lolium perenne L. under subtropical conditions. Australian Journal of Agricultural
Research 47(6): 975-983.
McKibben EG 1971. Introduction. In: Barnes KK, Carlton WM, Taylor HM,
Throckmorton RI & Van den Berg GE (eds). Compaction in agricultural soils.
American Society of Agricultural Engineering, St Joseph, Michigan. 3-6 pp.
McLeod G 1992. Environmental changes at Bobonong in the Central district,
Eastern Botswana. Botswana Notes and Records 24: 87-12l.
McNaughton 1987. Adaptation of herbivores to seasonal changes in nutrient supply.
262
In: Kacker JB & Ternouth JH (eds) The nutrition of herbivores. Acadamic Press,
Sydney. 39-408 pp.
Meadows SJ & White JM 1979. Structure of the herd and determinants of offtake
rates in Kajiado district in Kenya 1962-1977. Pastoral development Networks paper
No. l Ie. London: Overseas Development Institute.
Medina E 1982. Physiological ecology of neotropies savanna plants. In: Huntley BJ
& Walker BH (ed) Ecology of Tropical savanna, Ecological studies series Number
42. Springer-Verlag, New York, USA. 308-335 pp.
Mekonnen Y 1994. A survey of plants (potentially) toxic to livestock In the
Ethiopian Flora. Sinet, Ethiopian Journal of Science 17(1): 9-32.
Melville BIAS, Cauldwell AE & Bothma Jdu P 1999. A comparison of two
techniques for estimating tree canopy volume. South Africa Wildlife Research
29(4): 113-117.
Mengistu H 1996. Opening speech. In: ESAP Proceedings. Fourth National
conference of Ethiopian Society of Animal Production, 18-19 April 1996, Addis
Ababa, Ethiopia.
Mensat MK 1970. Beef production In East Africa. Ifo Institute fur Wirshaft
Forschung, Munich. 227 pp.
Mentis MT 1981. Evaluation of the wheel point and step-point methods of veld
condition assessment. Proceedings of the Grassland Society South Africa 16: 89-94.
Mentis MT 1983. Towards objective veld condition assessment. Proceedings of the
Grassland Society of Southern Africa 18: 81-83.
Mentis MT 1984. Monitoring in South African grasslands. Pretoria: Council for
scientific and Industrial Research. National scientific Programmes Unit (South
African National Scientific Programs report no. 91).
Mensa! JC & Cesar J 1979. Structure and primary productivity of Lam to savannas,
Ivory Cost. Ecology 60: 1 197-1 210.
Michael AM 1978. Irrigation: Theory and practice. Vikas publishing House PVT
LTD. New Dehli, India.
Miehaik DL & Norton BE 1980. The value of reference areas in the study and
management of range lands. Australian Rangeland JournaI2(2): 201-207.
Microsoft Excel 1997. Microsoft software, Microsoft corporation, USA.
Milchunas DG & Lauenroth WK 1993. Quantative effects of grazing and soils
over a global range of environments. Ecological Monographs 63(4): 327-366.
263
Miller D 1997. Range management and pastoralism: new perspectives and their
implications. ICIMOD, Newsletter of the International Centre for Integrated
Mountain Development No. 27: 4-7.
Minton SJ 1988. The effects of pruning on shoot production and basal increment of
Acacia tortilis. South African Journal of Botany 54: 109-117.
Ministry of Agriculture (MOA) 1996. Ruminant Livestock Development Strategy.
Draft. May, 1996, Addis Ababa, Ethiopia.
MOA 2000. Proceedings on National Pastoral extension System. Addis Ababa,
Ethiopia.
Minson IU & McDOIlunldeK 1987. Estimating forage intake from the growth of
beef cattle. Tropical Grasslands 21: 116-122.
Mohammed KY, Mukasa ME & Demessie I. 1989 Effects of the supplemenation
of lakesoil on the growth performane of Arsi sheep.In: NLIC Proceeding, Third
National Livestock Improvement conference, Addis Ababa, Ethiopia.
Mokwunye U 1996. Nutrient depletion: Sub-Saharan Africa's greatest threat to
Economic Development. Dakar,' Senegal: STAP Expert Workshop on Land
Degradation. 27 pp.
Moleele NM, Ringrose S, Matheson W & Vander C 2002. More woody plants?
The status of bush encroachment in Botswana's grazing area. Journal of
Environmental Management 64: 3-11.
Monasterio M & Sarmiento G 1976. Phenological strategies of plant species in the
tropical savanna in the semideciduous forest of the Venezuelen Iianos. Journal of
Biogeography 3: 325-356.
Monk CD & Gabrielson Jr Fe 1985. Effects of shade, litter and root competition
on old-field vegetation in South Carolina. Bulletin of the Torrey Botanical Club 112:
382-392.
Moore A 1989. Die ekologie en ekofisiologie van Rhigozum trichotomum. PhD
thesis, University of Port Elizabeth, South Africa.
Moore A & Odendaal A 1987. Die ekonomiese implikasies en bosbeheer soos van
toepassing op 'n speenkalfproduksiestelsel in die doringbosveld van die Molop-
gebied. Journal of the Grassland Society of Southern Africa 4: 139-142
Moore A, Van Niekerk JP, Knight rw & Wessels H 1985. The effect of
Tebuthiuron on the vegetation of the thorn bushveld of the northern
Cape - a preliminary report. Journal of the Grassland Society of Southern Africa 2:
7-10.
264
Moore G 1982. Forestry, wildlife and National Park Legislation in Ethiopia. FAO,
Rome, Italy. 128 pp.
Morgan WTW 1981. Ethno botany of the Turkana: Use of plants by a pastoral
people and their livestock in Kenya. Economic Botany 35: 96-130.
Morris CD, Hardy MB & Bartholomew PE 1999. Stocking rate. In: Tainton NM
(ed) Veld Management in South Africa. University of Natal Press, Pietermaritzburg,
South Africa. 186-192 pp.
Morris CD & Tainton NM 1993. The effect of defoliation and competition on the
regrowth of Themeda triandra and Aristida junciformis subsp. junciformis. African
Journal of Range and Forage Science 10(3): 124-128.
MoU .lfJ, WinDiams .lf, Andrew MB & Gillison AN 1985. Australian savanna
ecosystems. Ecology and Management of the world's Savannas. TothilI JC & Mott JJ
(eds). Australian Academy of science, Canberra. 56-81 pp.
Moyo CS & Campbell BM 1998. Grass species composition, yield and quality
under and outside tree crowns in a semi-arid rangeland in southwestern Zimbabwe.
African Journal of Range and Forage Science 15(1 & 2): 23-24.
Mudris AM 1998. Resource deprivation and socio-economic changes among
pastoral households: The case of Kereyu and Ittu pastoralists in the middle Awash
valley of Ethiopia. MSc thesis, Agricultural University of Norway.
Mugasi Sk, Sabiiti EN & Tayebwa BM 2000. The economic implication of bush
encroachment on livestock farming of Uganda. African Journal of Range and Forage
Science 17(1-3): 64-69.
Muir JP & Abrao L 1999. Productivity of 12 Stylosanthes In semi-arid
Mozambique. Tropical Grasslands 33: 40-45.
Muller DJ 1999. Herders of forty centuries: Nomads of Tibetan rangelands In
Western China. Proceedings of the Sixth International Rangeland congress.
Mustafa, K. 1997. Eviction of pastoralists from the Mkornzi game reserve in
Tanzania: an historical review. Pastoral land tenure series drylands programme
International Institute for environment and Development. No. 8. London, UK.
Nair PKR 1984. Soil productivity aspects of agro forestry. International Council for
Research in Agroforestry, Nambia.
NAS (National Academy of science) 1979. Tropical legumes: Resources for the
future. US National Academy of Sciences, Washington DC. 331 pp.
265
National Research Council (NRC) 1994. Rangeland health: new methods to
classify, inventory and monitor rangelands. National Academy of Press, Washington
DC.
Naude 'fW, KeIlerman TS & Coetzer JAW 1996. Plant poisonings and
mycotoxicoses as constraints to livestock production in East Africa: the southern
African experience. Journal of the South African Veterinary Association 67 (1): 8-
11.
Ndikumana J, Stuth JW, Kamadi R, Ossiya S, Marambii R & Hamlett JP2001.
Coping mechanisms and their efficacy in disaster-prone pastoral systems of the
greater horn of Africa: Effects of the 1995-97 drought and the 1997-98 El Nino and
the responses of pastoralists and livestock. 106 pp.
Negarit Gazeta 1944. A proclamation to make provision for the preservation of
Game. No. 61 of 1944. Order No.4 (31 October), Addis Ababa, Ethiopia.
Nerney JJ 1958. Grasshopper infestation in relation to range condition. Journal of
Range Management 11: 247.
Netting RM 1976. What alpine peasants have in common: Observations on
communal tenure in a Swiss village. Human Ecology 4: 135-146.
Nicoiis G & JPrigogine I 1989. Exploring complexity. Freeman WH and Company
New York.
Nott K & Savage MJ 1985. Observations on utilization of a dune succulent by
Namib Faunae. South African Journal of Zoology 20: 269-27l.
Oba G 1998. Assessment of indigenous range management knowledge of the Boran
pastoralists of Southern Ethiopia. Boran low land pastoral development
program/GTZ Consultancy paper. Negelle/Borana May 1998.
Oba G & Post E 1999. Browse production and off take by free ranging goats in an
arid zone Kenya. Journal of Arid Environments 43: 183-195.
O'Connor TG 1985. A synthesis of field experiments concerning the grass layer in
the savanna regions of Southern Africa. South African National Scientific
Programmes Report, 114. CSIR, Pretoria.
O'Connor TG 1991. Local extinction in perennial grasslands: a life-history
approach. American Naturalist 137: 735-773.
O'Connor TG 1994. Composition and population responses of an African savanna
grassland to rainfall and grazing. Journal of Applied Ecology 31: 155-171.
O'Connor TG 1995. Acacia karroo invasion of grassland: environmental and biotic
effects influencing seedling emergence and establishment. Oecologia 103: 214-223.
266
O'Connor TG & Bredenkamp GJ 1997. Grassland. In: Cowling RM, Richardson
DM & Pierce SM (eds) Vegetation of Southern Africa. Cambridge University Press,
Cambridge.
O'Connor TG & Crow VRT 1999. Rate and pattern of bush encroachment in
Eastern Cape savanna and grassland. African Journal of Range and Forage Science
16(1): 26-3l.
O'Connor TG, Halnes LM & Snyman HA 2001. Influence of precipitation and
species composition on phytomass of semi-arid African grassland. Journal of
Ecology 89: 850-860.
O'Connor TG & Piekert GA 1992. Influence of grazing on seed production and
seed banks of some African savanna grasslands. Journal of Applied Ecology 29: 247-
260.
Orban B 1995. An ecological management plan for the Lionspruit Game Reserve,
Eastern Transvaal. MSc (Wildlife management) Disseration, University of Pretoria,
Pretoria.
O'Reagain lPJ & Turner JR 1992. An evaluation of the empirical base for grazing
management recommendations for rangeland in the Southern Africa. Journal of the
Grassland Society Southern Africa 9(1): 39-49.
Palmer AIR, Crook BJC & Lubke RA 1988. Aspects of the vegetation and soil
relationships in the Andries Vosloo Kudu reserve, Cape Province. South African
Journal of Botany 54: 309-314.
Parker KW 1951. Application of ecology in the determination of range condition
and trend. Journal of Range Management 7: 14-23.
Pearce D 1997. An economic overview of wildlife and alternative land uses.
http:/www.uea.ac.uk /menu/acad_depts/env/all/resgroup/cserge/ (GEC 97-05) (1
May 1999).
Peart MB 1984. The effect of morphology, orientation and position of grass
diasporas on seedling survival. Journal of Ecology 72: 437-453.
Pechanec JP & Piekford GD 1937. A weight estimate method for the determination
of range and pasture production. Journal of the American Society of Agronomy 29:
894-904. .
Peddie GM, Tainton NM & Hardy MB 1995. The effect of past grazing intensity
on the vigour of Themeda triandra and Tristachya leuctthrix. African Journal of
Range and Forage Science 12(3): 111-115.
267
Pellew RA 1983. The giraffe and its food resource in the Serengeti. 1. Composition,
biomass and production of available browse. African Journal of Ecology 21: 241-
267.
Pendleton D 1966. Study exclosures. In: Range and Wildlife Technical Workshop
Bureau of land management Proceedings. 51-57 pp.
Perkins JS & Thomas DGS 1993a. Environmental responses and sensitivity to
permanent cattle ranching, semi-arid western central Botswana. Landscape
Sensitivity: Thomas DSG & Allison RJ (eds), John Wiley & Sons, London. 273-286
pp.
Perkins JS & Thomas DSG 1993b. Spreading deserts of spatially confined
environmental impacts - land degradation and cattle ranching in the Kalahari desert
of Botswana. Land degradation and Rehabilitation 4: 179-194.
Perrier GK 1995. A comparative analysis of production goals and strategies in
traditional pastoral systems. Rangelands in a sustainable biosphere: Proceedings of
the fifth International Rangeland Congress. Salt Lake City, Utah, July 23 -28 1995.
440-441 pp.
Peterson RG, Wood house Jr WW & Lucas HL 1956. The distribution of excreta
by freely grazing cattle and its effect on pasture fertility: II. Effect of returned excreta
on the residual concentration of some fertilizer elements. Agronomy Journal 48: 444-
449.
Petrides GA 1961. Wildlife preservation and national parks in Ethiopia. Wildlife,
Nairobi 3(3): 24-26.
Picardi A and Seifert W 1976.A tradegy of the Commons in the Sahel. Technology
Review 78: 45-5l.
Pieper RD 1978. Measurement techniques for herbaceous and shrubby vegetation.
Mimeograph Series, New Mexico State University, Las Cruces. 501 pp.
Pieper RD 1990. Overstory-understory relations in Pinyon juniper woodlands in
New Mexico. Journal of Range Management 43: 413-415.
Potgieter JHJ 1991.Blaaranatomie van grassspesies van die landbouna
vorsingsentrum te Glen. MSc thesis, University of the Orange Free State,
Bloemfontein.
Potter CS 1992. Stemflow nutrient inputs to soil in a successional hardwood forest.
Plant and Soil 140: 249-254.
Powell JM & Waters-Bayer A 1985. Interactions between livestock husbandry and
cropping in a West African savanna. In: Tothill JC & Mott JJ (eds) Ecology and
management of the world's Savannas. Australian Academy of Science, Canberra.
268
Powell JM & Wilniams T 1993. Livestock, nutrient cycling and sustainable
agriculture in the West African Sahel. Sustainable Agriculture Gatekeeper Series
SA37. International Institute for Environment and Development, London.
Pratt DJ & Gwynne MD 1977. Rangeland management and Ecology in East Africa.
Frieger publishing company, Huntington, New York.
Pratt DJ, Le Gall F & De Haan C 1997. Investing in a pastoralism sustainable
natural use in arid Africa and the Middle East. World Bank technical paper.
ISSN9253-7494, NO.365.
Pressland AJ 1973. Rainfall partitioning by an arid woodland Acacia aneura F.
Muell in southwestern Queensland. Austalian Journal of Botany 21: 235-265.
Pressland AJ & Graham 'fWG 1989. Approaches to the restoration of rangelands-
the Queensland experience. Australian Rangeland Journal 11(2): 101-109.
Prins HHT 1992. The pastoral road to extinction: Competition between wildlife and
traditional pastoralism in East Africa. Environmental Conservation 19(2): 117-123.
Prins HBT & Van der Jeugd HP 1992. Growth rates of shrubs on different soils in
Tanzania. African Journal of Ecology 30: 309-315.
Purnell MF 1983. The FAO methods of land evaluation for agriculture and forestry
as compared to extensive grazing. In: Siderius (ed), 133-143 pp.
R.ange Inventory standardization Committee (RISC) 1983. Guidelines and
terminology for range inventory and monitoring.
Reich PB & Borehert R 1982. Phenology and ecophysiology of the tropical tree,
Tabebula neochrysantha (Bignoniaceae). Ecology 63: 294-299.
Reich PB & Borehert R. 1984. Water stress and tree phenology in a tropical dry
forest in the lowlands of Costa Rica. Journal of Ecology 72: 61-74.
Reich PB, VIIn C, Waiters MB & Ellsworth DS 1991. Leaf lifespan as a
determinant of leaf structure and function among 23 tree species in Amazonian forest
communities. Oecologia 86: 16-24.
Reich PB, VIllI C, Waiters MB & Ellsworth DS 1992. Leaf lifespan in relation to
leaf, plant and stand characteristics among diverse ecosystems. Ecological
Monographs 62: 365-392.
Reinhard F & Admasu A 1994. Honeybee flora of Ethiopia. Margraf Verlag,
Germany. 510 pp.
269
Rethman NFG & Kotze GD 1986. Veld condition in the southeastern Transvaal and
its effect on grazing capacity. Journal of the Grassland Society of Southern Africa 3:
134-140.
Reynolds HG & Glen-Dening GE 1949. Merriam kangaroo rat a factor in mesquite
propagation on southern Arizona range lands. Journal of Range Management 2: 193-
197.
Richards FJ 1959. A flexible growth function for empirical use. Journal of
Experimental Botany 10: 290-300.
Richter CFG 1991. Gras-bosinteraksie in die bosveldgebiede van Noord-Kaap.
M.Sc. Agric-thesis, University of the Orange Free State, Bloemfontein.
Richter CCF, Snyman HA & Smit GN 2001. The influence of tree density on the
grass layer of three semi-arid savanna types of Southern Africa. African Journal of
Range and Forage Science 18: 103-109.
Ringrose S, Van der Post C & Matheson W 1996. The use of integrated remotely
sensed and GIS data to determine causes of vegetation cover change in southern
Botswana. Applied Geography 16: 225-242.
R.obbins CT, Mole S, Hagerman AE & H.anley TA 1987. Role of tannins in
. defending plants against ruminants: reduction in dry matter digestion. Ecology 68: 1
606-1 615.
Roberts BR 1970. Assessment of veld condition and trend. Proceedings of the
Grassland Society of South Africa 5: 137-139.
Robertson IJM 1970. Awash National Park: A working plan. EWCO, Addis Ababa,
Ethiopia. 128 pp.
Rocheleau D 1988. Women, Trees and tenure. In: L. Fortman & LBruce (eds),
Whose trees? Proprietary Dimensions of Forestry, Boulder & London: Westtview
. Press.
Roderiek S, Stevenson P & Ndung' J 1998. The production parameters influencing
the composition and structure of pastoral cattle herds in a semi-arid area of Kenya.
Journal of Animal Science 66: 585-594
Roderiek S, Stevenson P & Ndungu J 1999. Factors influencing the production of
milk from pastoral cattle herds in Kenya. Animal Science Journal 68: 201-209.
Roe EM 1997. Viewpoint: On rangeland carrying capacity. Journal of Range
Management 50: 467-472.
270
Rosenschein A, Tietema T & Hall DO 1999. Biomass measurement and monitoring
of trees and shrubs in a semi-arid region of central Kenya. Journal of Arid
Environments 42: 97-116.
Roux PW 1969. Die veld as voedingsborn vir die kleinveebedryf.
Kleinveesimposium, University of the Orange Free State, Bloemfontein.
Roux PW 1981. Interaction between climate, vegetation and runoff in the Karoo.
Karoo Agriculture 2(1): 4-8.
Russell JR, Betteridge K, Costall DA & Mackay AD 2001. Cattle treading effects
on sediment loss and water infiltration. Journal of Range Management 54: 184-190.
Rutherford MC 1979a. Aboveground biomass subdivisions in woody species of the
savanna ecosystem project study area, Nylsley. South Africa National Scientific
Programmes Report 36. 33 pp.
Rutherford MC 1979b. Plant-based techniques for determining available browse
and browse utilization: a review. Botanical Review 45: 203-228.
Rutherford MC 1982. Aboveground biomass categories of woody plants m a
Burkea africana- Ochna pulchra savanna. Bothalia 14: 131-138.
Sandford S 1983. Management of pastoral development in the third world.
Chichester: John Wiley & Sons. 10-60 pp.
Sankary MN & Ranjhan SK 1989. The place of fodder trees and shrubs in
grassland system. In: XVI International Grassland Congress, 4-11 October 1989,
Nice, France. 1761-1767 pp.
Sato S 1980. Pastoral movements and the subsistence unit of the Rendille of northern
Kenya. Senri Ethnological Series 6: 1-78.
Sarmiento G, Goldstein G & Meinzer F 1985. Adaptive strategies of woody
species in neotropical savannas. Biological Review 60: 315-355.
Schapera ][ 1943. Native Land Tenure in the Bechuanaland Protectorate Lovedale:
Alice.
Schloeder CA & Jacobs MJM. 1993. The Awash National Park Management Plan
1993-1997. NYZS and EWCO, Addis Ababa, Ethiopia. 288 pp.
Schofield CJ & Bucher Ell 1986. Industrial contributions to desertification in
South America. Trends in Ecology and Evolution 1: 78-80.
Scholes RJ 1987. Response of three semi-arid savannas on contrasting soils to the
removal of the woody component. PhD-thesis, University of the Witwatersrand,.
Johannesburg, South Africa.
271
Scholes RJ 1991. The influence of soil fertility on the ecology of south African dry
savannas. In: Werner PA (ed) Savanna ecology and management. Australian
perspectives and intercontinental comparisons. Blackwell Scientific Publications.
London. 71-76 pp.
Seholes RJ & Wallker BH 1993. An African savanna: synthesis of the Nylsvley
study. Cambridge: Cambridge University Press. 306 pp.
Sehullz CE & Skonhoft A 1996. Wildlife management, land-use and conflicts. South
African Journal Wildlife Research 26(4): 151-159.
Scoones I 1989a. Economic and ecological carrying capacity: Implications for
livestock development in Zimbabwe's communal areas. Pastoral Development
Network Paper No. 27b. Overseas Development Institute, London.
Seooues I 1989b. Patch use by cattle in dry land Zimbabwe: farmer knowledge and
ecological theory. Pastoral Development Network Paper No. 28b. Oversea
Development Institute. London.
Scoones I 1995a. Policies for pastoralists: New directions for pastoral development
in Africa. In: Binns TB (ed). People and the environment in Africa. 23-30,
Chichester, UK: Wiley.
Seooues I 1995b. Contrasting savanna dynamics: implications for livestock
populations in Zimbabwe's dryland communal areas. In: Garry F & Campbell B
(eds) Sustainable land management in African semi-arid and subhumid regions:
Proceedings of the Scope workshop, 15-19 November 1993, Dakar, Senegal.
Montpellier, France: CIRAD. 353-360 pp.
Seoones I 1995e. New directions in Pastoral Development in Africa. In: Scoones I
(ed) Living with uncertainity: New directions in Pastoral Development in Africa.
London: lIED.
Shackleton CM, Griffin NJ & Banks DI 1994. Community structure and species
composition along a disturbance gradient in a communally managed South African
savanna. Vegetatio 115: 157-167.
Sharon L 2000. Natural resource use and threatened livelihoods: The case ofOromo
and Afar pastoralists in the middle Awash valley of Ethiopia. MSc conservation,
University College London.
Shiflet TN & Dietz HE 1974. Relationship between precipitation and annual
rangeland herbage production in Southeastern Kansas. Journal of Range
Management 27(4): 272-276.
Short LR & Woolfolk EJ 1956. Plant vigor as a criterion of range condition.
Journal of Range Management 9: 66-69.
272
Shukla RP & Ramakrishnan PS 1982. Phenology of trees in a subtropical forest in
northeastern India. Vegetatio 49: 103-109.
Silcock lRG 1977. A study of the fate of seedlings growing on sandy red earths in the
Charleville district, Queensland. Australian Journal of Botany 25: 337-346.
Sillitoe P 1998. What knows natives? Local knowledge in Development. Social
Anthropology 6: 203-220.
Simson G & Mitiku H 1998. Trends of resource use in pastoral and agropastoral
areas: A study of natural resource management of a pastoral group in North-Eastern
Ethiopia: The Wossama Afar. 14-31 pp.
Sinclair ARE & Fryxell JM 1985. The Sahel of Africa: Ecology of a disaster.
Canadian Journal of Zoology 63: 987-994.
Singh .:D'S& Johsi MC 1979. Ecology of the semi-arid regions of India with
emphasis on land use. Management of Semi-Arid Ecosystems: Walkers BH (ed),
Elsevier, Amsterdam. 243-273 pp.
Singh P & Ghosh AN 1993. Forage and animal management systems on communal
rangelands in arid and semi-arid regions. Grasslands for our world. Brougham.
Sintayehu GM 1993. Livestock marketing in the pastoral areas of Ethiopia. In:
Edwards S & Tafesse M. (eds) Conference on pastoralism in Ethiopia. 4-6 February
1993, MOA, Addis Ababa, Ethiopia.
Skarpe C 1990a. Shrub layer dynamics under different herbivore densities in an arid
savanna, Botswana. Journal of Applied Ecology 27: 873-885.
Skarpe C 19901:>.Structure of woody vegetation in disturbed and undisturbed arid
savanna, Botswana. Vegetatio 87: 11-18.
Skarpe C 1991. Impact of grazing in savanna ecosystems. Ambio 20: 351-356.
Skarpe C 1992. Dynamics of savanna ecosystems. Journal of Vegetation Science 3:
293-300.
Skerman PJ 1977. Tropical forage legumes. FAO Plant Production and Protection
series, No.2, FAO, Rome.
Skinner JD & Smithers RHN 1990. The mammals of the southern African sub-
region. (2nd edition). University of Pretoria Press, Pretoria, South Africa.
Skinner TE 1964. 'n Fisiologiese-ekologiese studie van stipagrostis ciliata en S.
obusta. MSc- thesis, University of Pretoria, Pretoria.
273
Slayter RO 1975. Structure and function of Australian arid shrublands. In: Hyder
DN (ed), Arid shrublands-Proceedings. Third workshop United States/Australian
Rangelands Panel. Society of Range Management, Denver, Colorado.
Smit GN 1989a. Quantative descripition of woody plant communities: Part 1. An
approach. Journal of the Grassland Society of Southern Africa 6: 189-191.
Smit GN 1989b. Quantative descripition of woody plant communities: Part 2.
Coumpterized calculation procedures. Journal of the Grassland Society of Southern
Africa 6: 192-194.
Smit GN 1989c. A simple, inexpensive method for estimating linear tree
dimensions. Journal of the Grassland Society of Southern Africa 6: 104-105.
Smit GN 1994. The influence of intensity of tree thinning on Mopani Veld. PhD
thesis, University of Pretoria, South Africa.
Smit GN 1996. Biomass Estimation from Canopy Volume (version 2) - Users guide.
Unpublished report, University of the Free State, Bloemfontein, South Africa.
Smit, GN 1999a. Guide to the Acacias of South Africa. Briza publications, Pretoria,
South Africa. 223 pp.
Smit GN 1999b. Calculation of browsing capacity. Unpublished report. University
of the Free State, Bloemfontein.
Smit GN 2001. The influence of tree thinning on the vegetative growth and browse
production of Colophospermum mmopane. South African Journal of Wildlife
Research 31(3&4): 99-114.
Smit GN 2002. The importance of ecosystem dynamics in managing the bush
encroachment problem in Southern Africa. Inaugural lecture-May, 2002. University
of the Free State, Bloemfontein, South Africa.
Smit GN & Rethman NFG 1992. Inter- related floristic changes associated with
different long-term grazing treatments in Sourish Mixed Bushveld. Journal of the
Grassanl Society Southern Africa 9: 76-82.
Smit GN & Rethman NFG 1998a. The influence of tree thinning on the
reproductive dynamics of Colophospermum mopane. South African Journal of
Botany 64: 38-43.
Smit GN & Rethman NFG 1998b. Root biomass, depth distribution and relations
with leaf biomass of Colophospermum mopane. South African Joumal of Botany 64:
25-29.
Smit GN, Rethman NFG & Moore A 1996. Review article: vegetative growth,
reproduction, browse production and response to tree clearing of woody plants in
African Savanna. African Journal of Range and Forage Science 13(2): 78-88.
274
Smit GN, Richter CGF & Aucamp AJ 1999. Bush encroachment: An approach to
understanding and managing the problem. In: "Veld management in South Africa".
(Tainton NM), University of Natal Press, Pietermaritzburg, South Africa. 246-260
pp.
Smit GN & Swart JS 1994. The influence of leguminous and non-leguminous
woody plants on the herbaceous layer and soil under varying competition regimes in
mixed bushveld. African Journal of Range and Forage Science Il: 27-33.
Smith AB 1992. Pastoralism in Africa: Origin and development ecology. C. Hurst &
Co. (publishers) Ltd., United kingdom .. 283 pp.
Smith EL 1988. Successional concepts in relation to range condition assessment. In:
Tueller P (ed), Vegetation science applications for rangeland analysis and
management. The Hague: 113-133 pp.
Smith EL 1989. Range condition and secondary succession: a critique. In: Hyder
DN (ed), Proceedings of the First International Rangeland Congress, Society of
Range Management, Denver, Colorado.
Smith .JC 1895. Forage conditions of the prame region. USDA Yearbook of
Agriculture. 309-324 pp.
Smith TM & Goodman PS 1986. The effect of competition on the structure and
dynamics of Acacia savannas in the southern Africa. Journal of Ecology 74: 1031-
1044.
Smith TM & Goodman ]PS 1987. Successional dynamics in an Acacia nilotica-
Euclea divinorum savannah in southern Africa. Journal of Ecology 75: 603-610.
Snyman DD, Grossman D & Rethman NFG 1990. Tekortkomings van die
naasteplantmetode en Dyksterhuis-verwante Klassifikasiesisteme om veldtoesstande
in semi-ariede gebied te bepaal. Journal of the Grassland Society of South Africa
7(4): 273-276.
Snymarn HA 1989. Evapotranspirasie en waterverbruiks-doeltreffendheid van
verskillende grasspesies in die sentrale Oranje-Vrystaat. Journal of the Grassland
Society of Southen Africa 6(3): 146-151.
Snyman HA 1993. Strive to optimium veld condition and grazing capacity. South
Africa Koop 9: 25-28.
Snyman HA 1997. The influence of range condition on the hydrological
characteristics in semi-arid rangelands. Proceedings of the XVIII International
Grassland congress, Canada. 2(23): 1-2.
275
Snyman HA 1998. Dynamics and sustainable utilisation of rangeland ecosystems in
arid and semi-arid climates of Southern Africa. Journal of Arid Environments 39:
645-666.
Snyman HA 1999a. Soil erosion and conservation. In: Tainton NM (ed), Veld
management in South Africa. University of Natal Press, Pietermaritzburg, South
Africa. 355-380 pp.
Snyman HA 1999b. Short- term effect of soil-water, defoliation and rangeland
condition on productivity of a semi-arid rangeland in South Africa. Journal of Arid
Environments 43: 47-62.
Snyman HA & Fouche HJ 1991. Production and water-use efficiency of semiarid
grasslands of South Africa as affected by veld condition and rainfall. Water South
Africa 17: 263-268.
Snyman HA & Fouche HJ 1993. Estimating seasonal herbage production of a
semi-arid grassland based on veld condition, rainfall and evapotranspiration. African
Journal of Range and Forage Science 10(1): 21-24.
Snyman HA, Venter WD, Van Rensberg WLJ & Opperman DPJ 1987. Ranking
of grass apecies according to visible wilting order and rate of recovery in the Central
Orange Free State. Journal of the Grassland Society of Southern Africa 4: 78-8l.
Soane BD, Blackwelt PS, Dickson JW & Painter DJ 1981. Compaction by
agricultural vehicles: A review, II. Compaction under tyres and other running gear.
Soil and Tillage Research 1: 373-400.
Society of Range Management (SRM) 1974. A glossary of terms used in range
management. Denver, Colarodo.
Solomon B, De Leeuw PN, Grandin BE & Neate PJH 1991. Maasai herding: An
analysis of the livestock production system of Maasai pastoralists in eastern Kajiado
District, Kenya. ILCA Systems study 4. International Livestock Center for Africa,
Addis Ababa, Ethiopia. 154 pp.
Solomon D 1993. Past and present Pastoral development interventions in Ethiopia.
In: Edwards S & Tafesse M (eds), Proceedings of conference on Pastoralism in
Ethiopia, 4-6 February 1993, Addis Ababa, Ethiopia.
Spence LE 1938. Range management for soil and water conservation. The Utah
juniper. 9: 18-25.
Sporton D, Thomas Jl)SG & Morrison J 1999. Outcome of Social and
Environmental Change in the Kalahari of Botswana: the Role of Migration. Journal
of Southern African Studies 25(3): 441-459.
276
SAS (Statistical Analysis System) Institute Inc 1987. Applied statistics and the
SAS programming language 2nd edition Cary, North Carolina.
SlPSS (Statistical Packages for the Social Sciences) 1996. Cary, North Carolina.
Steen E 1994. Drylands of the third World: Potential for Future Development.
Ambio 23(7): 458-460.
Stiles D & Kassam A 1984. An ethno-botanical study of Gabra plant use in Marsabit
District, Kenya. (ILCA library accession no. 89188). 26 pp.
Stoddart LA, Smith AJ[) & Box TW 1975. Range management. New York,
McGraw Hill.
Stuart-Hili GC 1988. Be earful when using the nearest plant grass survey method in
arid areas. Dohne Agriculture 10: 18-23.
Stuart-Hili GC & Bobson FO 1991. An alternative approach to veld condition
assessment in the non-grassveld regions of South Africa. Journal of the Grassland
Society of Southern Africa 8(4): 179-185.
Stuart-Hili GC & Tainton NM 1999. The management of different types of veld:
Savanna. In: "veld management in South Africa" Tainton NM (ed), University of
Natal press, Pietermaritzburg, South Africa. 312-316 pp.
Stuart-Hili GC, Tainton NM & Barnard HJ 1987. The influence of Acacia karroo
tree on grass production in its vicinity. Journal of the Grassland Society of Southern
Africa 4: 83-88.
Sullivan GM & Farris DE 1976. Survey of traditional livestock industry. Tanzania
Livestock Meat Subsector, Texas A & M University. Vo1.2. 147 pp.
Swallow BM 1994. The role of mobility within risk management strategies of
pastoralists and agropastoralists. Gatekeeper Series SA47. International Institute for
Environment and Development, London.
Styles CV 1993. Relationships between herbivores and Colophospermum mopane of
the northern Tuli Game Reserve. MSc thesis, University of Pretoria, Pretoria.
Tafesse M 1990. Ethiopian traditional veterninary practices and their possible
contribution to animal production and management. Paper presented at a conference
on the Need for a National Policy to promote and develop Ethiopian Traditional
medicine held-in November 1990, Addis Ababa, Ethiopia.
Tafesse M 2000. An overview and analysis of the history of public policy towards
the development of pastoralism in Ethiopia. In: the Proceedings of the National
Conference on Pastoral Development in Ethiopia. February 2 2000, Addis Ababa,
Ethiopia.
277
Tafesse M 2001. What should a pastoral development strategy constitute towards
poverty reduction among pastoral communities in Ethiopia? In: PFE Proceedings.
Poverty reduction strategy and pastoral development. Proceedings of the second
National Conference on Pastoral Development in Ethiopia, May 22-23 2001, Addis
Ababa, Ethiopia.
Tainton NM 1981. The assessment of veld condition. In: Tainton MN (ed) Veld and
pasture Management in South Africa. Shuter and shooter Ltd: Pietermaritzburg,
South Africa. 46-55 pp.
Tainton NM 1988. A consideration of veld condition assessment techniques for
commercial livestock production in South Africa. Journal of the Grassland Society
Southern Africa 5(2): 76-79.
Tainton NM 1999. Veld management in South Africa. In: Tainton NM (ed).
University of Natal Press, Pietermarizburg South Africa.
Tainton NM, Bransby DI & Booysen P de V 1976. Common veld and pasture
grasses of Natal. Department of Pasture Science, University of Natal.
Talbot LM & Talbot MB 1963. The wildebeest in Western Masailand. East Africa.
Wildlife Monographs 12: 1-88.
Taylor JA & Wlltalley ROB 1976. The philosophy of range management research in
Australia. Journal of Australian Institute of Agricultural Science 42: 181-186.
Teague WR & Smit GN 1992. Relations between woody and herbaceous
components and the effect of bush-clearing in Southern African savannas. Journal of
the Grassland Society of Southern Africa 9: 60-71.
Teague WR, Trollope WSW & Aucamp AJ 1981. Veld management in the semi-
arid bush-grass communities of the eastern Cape. Proceedings of the Grassland
Society of Southern Africa 16: 23-28.
Thapa B, Walker DH & Sin clair FL 1997. Indigenous knowledge of the feeding
value of tree fodder. Animal Feed Science Technology 67: 97-114.
Theron GK 1964. 'n Outekologiese studie van Plinthus karooieus. MSc-thesis,
Univeersity of Pretoria, Pretoria.
Thompson LM & Frederick RT 1978. Soil and soil fertility. McGraw-Hill
Publishing Company, New York.
Thompson LM & Treoh FR 1978. Soils and soil fertility. Fourth edition. Tata
McGrew-Hill Publishing Company Ltd. New Delhi.
Thrash I, Nel lPJ, Theron GK & Bothma J du jl' 1991. The impact of the
provision of water for game on the basal cover of the herbaceous vegetation around a .
dam in the Kruger National Park. Koedoe 34(2): 121-130.
278
Thun R, Herrmann R & Knickmann E 1955. Untersuchung von Boden. Handbuck
der Landwirtsch. Versuchs-und Untersuchungsmthodik, 3. Auflage, Neumann Verlag
Rodebeul u.Berlin.
Thurow TL, Blackburn WH & Taylor CA Jr 1986. Hydrologic characteristics of
vegetation types as affected by livestock grazing systems. Edwards Plateau, Texas.
Journal of Range Management 39: 505-509.
Thurow TL, Blackburn WH & Taylor CA 1988. Some vegetation responses to
selected livestock grazing strategies. Journal of Range Management 41: 108-114.
Thurow TL & West NE 1996. Influence of Range Condition on patterns of social
change in trans Mara district, Kenya. Rangelands in a sustainable biosphere:
Proceedings of the Fifth International Rangeland Congress, Salt Lake City, Utah,
USA, 23-28 July, 1995. Volume 1: Contributed presentations, 1996. 557-558 pp.
Tibabu A 1997. Park versus pastoralists: resource conflicts in the Awash valley of
Ethiopia. MSc- thesis, Agricultural University of Norway.
Tidmarsh CEM & Havenga CM 1955. The wheel point method of survey and
mesurement of semi-open grasslands and Karoo vefgetation in South Africa.
Memiors of the Botanical Survey of South Africa. 29.
Tiedemann AR & Klemmedson JO 1973. Effects of mesquito trees on vegetation
and soils in the desert grasslands. Journal of Range Management 30: 361-367.
Tisdane SL, Nilson WL, Beaten ID & Havlin JL 1993. Soil fertility and fertilizers.
s" edition. Prentice-Hall, Inc. Simon & Schuster, A Viacom Company, New Jersey.
Toulmin C 1992. Cattle, women and wells: Managing Household Survival in the
Sahel. Clarendon Press, Oxford.
Toupet C 1975. Le nomade, conservateur de la nature? 'L'exemple de la Mauritanie
centrale' in T. Monod (ed) Pastoralism in tropical Africa. Oxford: Oxford University
Press.
Trollope WSW 1990. Veld management with specific reference to game ranching in
the grassland and savanna areas of South Africa. Koedoe 33(2): 77-86.
TroIIolPe WSW, TroIIolPe LA & Bosch OJH. 1990. Veld and pasture management
terminology in South Africa. Journal of the Grassland Society Southern Africa 7:
52-61.
Tsui AO, Ragsdale TA & Shirwa AI. 1991. The settlement of Somalia nomads.
Genus: 1-2: 132-152.
Tsundle K 1999. Pastoral nomadism in Tibet: between tradition and modernization.
Proceedings of the Sixth International Rangeland Congress.
279
Tueller PT & Blackburn WH 1974. Condition and trend of the big sagebush/needle
and thread habitat type in Nevada. Journal of Range Management 27: 36-40.
United Nation Educational, Science and Cultural Organization (UNESCO)
1964. Saving the wildlife of Ethiopia. Recommendations of UNESCO mission.
Oryx, 7: 247-250.
United Nations Development Programme (UNDP) 1984. Geomorphology and
Soils of Ethiopia: Assistance to land-use planning. Rome, Italy.
UNDP 1997. Drought emergency planning to support pastoralist livelihood m
Ethiopia.
UNDP/Relief and Rehabilitation Commission (RRC) 1984. A study of Nomadic
Areas of Ethiopia. Study Report Part I-VI, Addis Ababa, Ethiopia.
Unruh JD 1990. Integration of transhumant pastoralism and irrigated agriculture in
semi-arid East Africa. Human Ecology New York 18(3): 223-246.
Unruh JD 1995. The relationship between indigenous pastoralist resource tenure and
state tenure in Somalia. GeojournaI36(1): 19-26.
Valentine KA 1967. Seasonal suitability, a grazing system for ranges of diverse
vegetation types and condition classes. Journal of Range Management 20: 395-397.
Van der Westhuizen HC 1994. Die kwantifisering van veldtoestrand in die sentrale
Vrystaar. M:Sc. Agric- thesis, University of the Orange Free State, Bloemfontein.
Vali]der Westhuizen HC, Snyman HA, Vall1Rensburg WLJ & Potgieter JHJ
2001. The quantification of grazing capacity from grazing- and production values for
forage species in semi-arid grasslands of Southern Africa. African Journal of Range
and Forage Science 18: 43-52.
Van der Westhuizen HC, Van Rensburg WLJ & Snyman HA 1999. The
quantification of rangeland condition in a semi-arid grassland of Southern Africa.
African Journal of Range and Forage Science 16(2&3): 49-61.
Van der berg JA 1983. Die verwantskap tussen die langtermyn gemiddelde reenval
en die weidingkapasiteit van natuurlike veld in die droe dele van Suid-Afrika.
Proceedings of the Grassland Society of Southern Africa. 18: 147-150.
Van Oudtshoorn F 1999. Guide to grasses of Southern Africa. Briza publication,
Pretoria, South Africa. 288 pp.
Van Niekerk JP & Kotze T 1977. Chemical control of bush encroachment by a
means of aerial spraying. Proceedings of the Second National Weed Conference of
Southern Africa 165-184.
280
Van Niekerk JP & Kotze T 1977. Chemical control of bush encroachment by a
means of aerial spraying. Proceedings of the Second National Weed Conference of
Southern Africa 165-184.
Van Rooysen N, Bredenkamp G.J & Theron GK 1991. Kalahari vegetation: veld
condition trends & ecological status of species. Koedoe 34(1): 61-72.
Van Vegten .JA 1983. Thornbush invasion in a savanna ecosystem m Eastern
Botswana. Vegetatio 56: 3-7.
Van Zyl EA 1986. Veldtoesstandbepaling (Veld condition assessment).
Potchefstroom, South Africa: Department of Agriculture. 7 pp.
Veenendaal EM, Shushu DD & Scurlock JMO 1993. Responses to shading of
seedlings of savanna grasses with different C4 photosynthetic pathways in
Botswana. Journal of Tropical Ecology 9: 213-229.
Verdlcourt B, Trump EC & Church ME 1965. Common poisonous plants of East
Africa. Collins. St James's place, London. 248 pp.
Vetaas OR 1992. Micro-site effects of trees and shrubs in dry season savanna.
Journal of Vegetation Science 3: 337-344.
Vorster LF 2000. Applied veld management. The University of the Free State,
Bloemfontein, South Africa (compiled dictation).
Vorster M 1981a. 'n Belowende tegniek om veld in die karoogebiede te beoordeel.
Karoo Agric 2(1).
Voster M 1981 IJ. 'n Benadering om weidingkapasiteit vanaf veldtoestand te bepaal.
Karoo Agric. 2(2).
Vorster M 1982. The development of the ecological index method for assessing veld
condition in the Karoo. Proceedings of the Grassland Society of Southern Africa 17:
84-89.
Waktola A & Micheal DG 1999. Institutional development and phasing out of the
Norwegian involvement. The case of Awash Conservation and Development Project,
CARE, Ethiopia. Drylands Co-ordination Group, Agricultural University of Norway.
Walker J & Gillison AN 1982. Australian savannas. Ecology of Tropical Savannas
Iiuntley BJ & Walker BH (eds), Springer, New York. 5-24 pp.
Walker PH 1980. A review of browse and its role in livestock production in
Southern Africa. In: Browse in Africa, the current state of knowledge. Houerou H N
Le (ed), ILCA, Addis Ababa, Ethiopia.
281
Wang X. & Zheng D 1998. Sustainable use of alphine meadow Grassland
Resources on the Qinghai Tibetian platuea: An Abstract. ISQTP, Xining.
Ward D, Ngairorue BT, Kathena J, Samuels R & Ofran Y 1998. Land
degradation is not a necessary outcome of communal pastoralism in arid Nambia.
Journal of Arid Environments 40: 357-371.
Wario G 1993. Veterninary services in the pastoral areas of Ethiopia: Constraints
and options for improvement. In: Edwards S & Tafesse M. (eds), Conference on
pastoralism in Ethiopia, 4-6 February 1993, MOA, Addis Ababa, Ethiopia.
Warren A & Khogali M 1992. Desertification and drought in the Sudano-Sahelian
region 1985- 1991. United Nations Sudano-Sahelian Office (UNSO), New York. 102
pp.
Warren LB, Lauenroth WK, Szaker SR & Kinerson RS 1984. Primary
production and abiotic controls in forests, grasslands and desert ecosystems in the
United States. Ecology 64(1): 134-151.
Warren SD, Nevill MB, Blackburn WH & Garza NE 1986. Soil response to
trampling under intensive rotation grazing. Soil Science Society of American Journal
50: 1 336-1 341.
Weaver JE & Mueller IM 1942. Role of seedlings in recovery of mid-western
ranges from drought. Ecology 23: 275-294.
Weltz MA & Wood MK 1986. Short duration grazing in Central New Mexico:
Effect on sediment production. Journal of Soil and Water Conservation 41: 262-266.
Weltzin JK & Coughenour MB 1990. Savanna tree influence on understorey
vegetation and soil nutrients in northwest Kenya. Journal of Vegetation Science
1: 325-334.
Wentzel JJ 1990. Ekologiese skeiding van geselekteerde herbivoor spesies in die
benede Sabie-Krokodilrivier gebied van die Nasionale Kruger Wildtuin. MSc-thesis,
University of Pretoria, Pretoria.
Western D 1982. The environment and ecology of pastoralism in Arid savannas.
Development and Change 13: 183-211.
Westoby M 1980. Elements of a theory of vegetation dynamics in arid rangelands.
Isreali Journal of Botany 28: 164-194.
Westoby M, Walker B & Noy-Meir I 1989. Opportunistic management for
rangelands not at equilibrium. Journal of Range Management 42: 266-274.
282
Wezel A, Rajort JL & lIerbig C 2000. Influence of shrubs on soil characteristics
and their function in Sahelian agro-ecosystems in semi-arid Niger. Journal of Arid
Environments 44: 383-398.
White R. 1992. Livestock Development and pastoral production on communal
rangelands in Botswana. London: Commonwealth secretariat. Wiens JA (1977). On
competition and variable environments. American Scientists. 65: 590-597.
Widstrad GC 1975. The rationale of nomad economy. Ambio 4(4): 146-153.
Wilding R 1985. The history of Pastoralism and the Emergence of the Borana
Oromo. A review oflssues. ILCAllJEPSS Research Report No. 15.
Wilhite DA & Glantz MB 1985. Understanding the drought phenomenon: the role
of definition. Water InternationallO: 111-120.
William HS, Reynolds JF & Cunningham GL 1990. Biological feedbacks In
Global Desertification. Science: 1 043-1 048.
Williams AG, Kent M & Ternan .:DL 1987. Quantity and quality of bracken
throughfall, sternflow and litterflow in a Dartmoor catchment. Journal of Applied
Ecology 24: 217-230.
Williams MAJ & Balling RC 1996. Interactions of desertification and climate,
Arnold, London.
WiIniams RJ, Myers BA, Muller WJ, Duff GA & Eamus D 1997. Leaf phenology
of woody species in a north Australian Tropical Savanna. Ecology 78(8): 2 542-2
558.
Wilson AD 1984. Points of reference in the assessment of change in vegetation and
land condition. Australian Rangeland Journal 6: 69-74
Wilson AD & Tupper GJ 1982. Concepts and factors applicable to the
measurement ofrange condition. Journal of Range Management. 35(6): 684-689.
Wilson JH 1988. Shoot competition and root competition.Journal of Applied
Ecology 25: 279-296.
Wilson RT 1988. Small ruminant production systems In tropical Africa. Small
Ruminant Research 1: 305-325.
Winrock 1992. Assessment of animal agriculture in sub-Saharan Africa, Morrilton,
Arkansas, USA: Winrock International Institute for Agricultural Development.
Wolde Gelbriel K 2001. Land uselLand cover Survey for the proposed Regional
Land use plan. Bureau of Agriculture, Afar National Regional State, Addis Ababa,
Ethiopia. 55 pp.
283
Workman JP 1986. Range Economics. New York: Macmillan.
World Bank 2001. Pastoral area development in Ethiopia. Issues paper and project
proposal. Discussion draft, April 10, 2001. Addis Ababa, Ethiopia.
World resources Institute, International Institute for environment and
Development (WRI, .DER) 1989. World resources 1988-89. New York, Basic
books Inc.
Woodward A & Reed JD 1989. The influence of polyphenolic on the nutritive
value of browse: A summary of research conducted at rr_CA. rr_CA bulletin 35: 2-1l.
~CA, Addis Ababa, Ethiopia.
Woodward CL 1996. Soil compaction and topsoil removal effects on soil
properities and seedling growth in Amazonian Ecuador. Forest Ecology and
Management 82: 197-209.
Wooten EO 1908. The range problem in New Mexico. New Mexico Agricultural
Experimental Station Bulletin No. 66.
Wright SJ 1991. Seasonal drought and the phenology of understorey shrubs in a
tropical moist forest. Ecology 72: 1 643-1 657.
WlI.lH, Sharp PJH, Walker J & Penridge lLK 1985. Ecological field theory: a
spatial analysis of resource interference among plants. Ecological Modelling 29: 215-
243.
Yacob A 2001. Pastoralism in Ethiopia: The issue of viability. In: PFE Proceedings.
Proceedings of the National Conference on pastoral development in Ethiopia,
February 2,2000, Addis Ababa, Ethiopia.
Yadava PS 1990. Savannas of northeast India. Journal of Biogeography 17: 387-
396.
Young A 1989. Agroforestry for Soil Conservation. International Council for
research in Agro-forestry, Nairobi.
ZhOlll Q, Robson M & Kong G 1998. Comparsion between the results from
different ground vegetation cover estimation methods in a rangeland environment. In
Proceedings of the 9th Australasian remote sensing and photogrammatory conference,
20-24 July, 1998. Sydney, Vol. l. 71 pp.
Zinash S, Abule E, Stuth JW, Jamma A & Nidkumanna J 2000. Early warning
system and coping strategies for pastoralists. In: ESAP Proceedings. Proceedings of
the 8th annual conference of the Ethiopian Society of Animal Production, ESAP 24-
26 August 2000, Addis Ababa, Ethiopia.
284
Zinash S, Tesfaye K, Azage T, & Osuji P 1998. The regional LEWS project: the
contribution and role of Ethiopia. GL-CRSPILEWS (Global Livestock Cooperative
Research Program! Livestock Early Warning System for East Africa)- TAMU (Texas
A and M University) and EARO (Ethiopian Agricultural Research Organization).
Proceedings of the National Workshop on Early Warning System and Monitoring
Livestock Nutrition and Health, 4 February 1998, Addis Ababa, Ethiopia. GL-
CRSPILEWS- TAMU and EARO, Addis Ababa, Ethiopia.
285
Appendix 4.1. Questionnaires used in the household survey
Region Zone District
Pastoral Association (PA)-------------------- Village-------------------------------------------
House ho Id no- -- ---- ---- --- -- -- ---- --- --Date--- -- -- ---- --- -- -- ---- ----- -- -- --- -----
Respo ndants name- -- -- -- -- --- ---- ---- -Age-- --- ---- -- --- ---- ---- --- -- ---- --- -- ---
Sex --- ------ -- ----- ---- ----- -- ---- ---------
Ethnic group 1. Afar/Denebe 2. MarlWaima3. KarayolBaso
4. Karayo/Dulcha 5. Itu- 6. Somalia- 7. Others-
Educational background: 1. Formal education----- 2. Read and write------ 3.
None-----
4. Other (specify)-------------------
PART 1. General
1. For what purpose were you using your land 10-20 years back
1. Crop production------ 2. Grazing only------------- 3. Grazing and crop
production---------
4. Others------------------------
2. What is your current land use? (ha)
For cultivation For grazing (if private)
3. Compared to the past (10-20 years), how is the trend in land utilization for cultivation
and grazing
Trend 1. Increasing 2. Decreasing 3. Constant
Land use Trend Reason
Cultivation
Grazing
4. What is your main source of money (Multiple response question)
4.1.Selllivestock 4.2. Sell milk and milk products 4.3. Other (specify)
PART 2. Natural mineral sources and utilization (Like Haya, Boji ete)
1. Do you give natural minerals to your animals? A.Yes B. No
286
2. If the answer to the above question is yes, mention in the table types of mineral
sources, animal species offered, season., the frequency of offer and the form of offer
(Multiple response question)
Type of Animal species Season Frequency Form
mineral Offered of
sources offer
]PART 3. Poisonous plants
1. Do you know any kind of poisonous plant that affect your livestock production?
Yes--------- No--------Indifferent
2. If the answer is yes, indicate in the table the animal species affected,
season and traditional contro Imeasure
1. Animal species affected
1. Camel 2. Cattle 3. Sheep 4.Goat
2. Season when the effect of poisonous plants on animals is highest
1. Dry 2. Wet season 3. Both
Poisonous plant Animal species Seaso Traditional
name affected n control method
PART 4. Bush encroachment
1. Do you observe an increase in bush/shrubs in your grazing land?
Yes No Indifferent
2. What problem did you face due to these encroaching bushes/shrubs? (Multiple
response question)
1. Decrease production of grasses-------- 2. Accident due to wild animals-----------
3. Difficulty in herding----------- 4. Other (specify)-----------------------------
4. What do you think are the possible reasons for the increase of bushes/shrubs?
(Multiple response question)
1. Increased livestock population leading to overgrazing--------------------
287
2. Uncontrolled livestock and wild life movement from place to place---
3. Lack of burning------------------------------------------
4. Drought ---------------------------------------------------
5. I do not know--------------------------------------------
6. Other (specify )-------------------------------------------
PART 5. Water resources and utilization
1. Mention the sources of water to your livestock by season? (Multiple response
question)
Dry Distance m Wet Distance m
season Km season Km
(hr) (hr)
2. Is the water clean and adequate?
Cleanness Clean Dirty
Quantity Adequa te- -- -- -- --Inadeq uate- -- -- -----
3. If inadequate, specify months of inadequacy (--------------to-----------months)
4. What alternative do you have when you face water shortage
A) Buy water from pump B) Move to very distant water source C) Others (specify)
5. Do people and animals use the same sources of water?
~es-------------------------------
~o---------------------------------
+
If no,
Human water source-----------------------------------------
Livestock water source---------------------------------------
6. Is there any grazing land that you do not utilize in the dry season due to lack of water
to your animals?
Yes ~o
7. If yes, mention their names
Dry season-----------------------
288
PART 6. Feed resource and utilization
1. How many months do each of the following livestock feed resource cover for you in
a year?
(Name the months)
1.Cir~ing------------------------
2.Browse trees----------------------
3. Crop residues------------------
4. Weeds/crop thinning-------------------------
5. From migration---------------------------------
6. Sugar cane by-products/molasses-------------------
7. Other (specify)----------------------------------------
2. Is there period of critical feed shortage? 1=Yes 2= no
3. If yes, state the period and the major measure you take to overcome the problem
Period (Months )-----------------------------------------------------
4. Measures taken:
1= Sale anirnals----------------------------
2= Buy feed ----------------------------------
3= Move animals----------------------------------
4= Use reserve feed ---------------------------(specify)
5= social alliance (explain)-------------------------------------
6= Lopping browse trees (name)----------------------
7= Use non-convectional feeds (name)--------------
4. Do you collect and feed sugar cane tops?
Yes No
5. If you are not feeding sugar cane tops/molasses to your animals what is the reason?
1.Transport problem-----2. Financial shortage---3. Lack of knowledge of its usage--------
4.Other (specify)------------------ 5.Accessibility----------------------------------------
6. Do you give supplement to your animal during the dry season?
Yes-----------------------------------
No------------------------------------
289
7.Ifthe answer to the above question is yes, what type of feed do you give as supplement
and to which class of animals do you give priority? (Multiple response question)
Supplement type Animal given Reasons to
priority priority
8. Do you conserve feed for dry season? 1= Yes 2= No
9. If yes, state the type of feed you mostly conserve (Multiple response question)
1= Standing hay-----------------2= Cut hay---------3= Crop residues----4= Browse (pods,
leaves, ete)
5= Others (name)-----------------6= None
10. Do you use browse for livestock feeding? 1= Yes 2=No
11. What is the major problem in utilizing browses (Choice)?
1. They will dry soon after the rain---------------------------
2. Stage at which they will be poisonous is not known----
3. They are long enough to be reached by animals---------
4. Other (specify)---------------------
12. Do you know browse plants used for medicine purpose? Yes-------- No-------------
If yes, mention their names
13. For which of the following activities do you use trees/shrub/bushes (Rank)
1. House construction
2. Fire wood/charcoal making
3. Browsing
4. Fencing
5. Wild fruit
6. Other (specify)
14. Does such usage (other than browsing) affect the availability of browse to your
animal? Yes No
290
PART 7. Rangeland management
1. Does your community allocate grazing lands according to season?
):es----------------------
N0-----------------------
Indifferent-------------
2. If the answer to the above question is yes, what criteria do they use to sub-divide?
Suitability of the grazing lands in terms of availability of grass and browse .
Suitability to human beings in terms of health----------------------------------
Other (specify)----------------------------------
3.Ifthe answer to the above question is no, what is the reason------------------------
4. Currently are there rules that govern the use of grazing lands and water within your
community?
Type of use Response
):es No
Water
Grazing land
5. If the answer to the above question is yes, how is the grazing land and water use
governed?
6. If the answer to the above question is no, what are the possible reasons?
7. Do you practice range plant harvesting?
):es--------------------------
No---------------------------
If yes how and by whom--------------------
If no, why-----------------------
8. Which range condition improvement practices do you apply?
l. Burning 2. Irrigation 3. Over-sowing 4. Fertilizer use 5. manure use
6.Range exclosure 7. Resting 8. Others (specify) 9. None
291
9. If you are not praeticing any of them, what is the reason-----------------------------
10. Which type of ownership of the grazing land do you like in the future?
Type of ownership 1. Private 2. Communal 3. Both4. Open access
I TY!'" of ownershiE I Reason for the choice
11. Compared to the past how is the trend of grasses, legumes and trees
Trend 1. Increasing 2. Decreasing 3. Constant
Plant category Trend Reason
Grasses
Legumes
Trees/bushes/shrubs
Poisonous plant
12. How is the condition of the rangelands
1. Poor2. Fair 3. Good 4. Very good 5. Excellent
13. If it is a poor, what are the possible consraints (Rank)
Increment in livestock population---------------------------------
Drought (Low rainfall)-------------------------------------------------
Increased human population---------------------------------------
Lack of technical knowledge (Mismanagement )----------------
Little support from the government-------------------------------
Over grazing /overstocking due to shortage of land-------
Immigration----------------------------------------------------------
Bush encroachment-----------------------------------------------
Po isonous plant----- --------------- ------------------------- -----
Poor in soil fertility
Change in the direction ofBulga river-----------------------
Expansion of the towns
None----------- ------------------------
14. Does your community make periodic assessment of the condition of the grazing
land?
292
l.Yes 2. No
If No, give reason-------------------------------
If Yes, by whom and how-----------------------
15. Generally compared to the past 5 years, how is the intensity of the conflict?
A. Increased b. Decreased
16. If decreased, why?-------------------------------------------------
17. If increased why?--------------------------------------------------
18. What do you suggest to improve the condition of the range land?
19. Do wild herbivores share grazing lands with your animals?
Yes
No
20. How much forage do the wildlife animals graze?
1. More than your livestock
2. Same as livestock
3. Less than your livestock
23. Does wildlife utilize the same water sources as livestock? Yes No
24. Do you graze your animals in the national park and when?
1. Yes and when
2. No
PART 8. Migration pattern
l. Do you move livestock to other places in search of feed and water? Yes No
2. In the table given below, mention the type of animals involved, place moved and
distance in Km (hr)
I Animal categol)' IPlace moved I Distance in Km (br)
293
Rank Cattle Sheep Goat Camel
disease disease disease disease
3. What problems do you face in the migration?
1. Death of animals 2. Problem of wild animals
3. Feed and water shortage 4. Disease problem
5. Other (specify)
4. Is the frequency of migration increasing?
~es-----------------------------------------
No _
If yes, why-----------------------------------
5. What do you suggest regarding migration?
PART 9. Livestock
1. Livestock ownership/household
Cattle (Cow, heifers, calves, bulls& oxen)
Sheep (Male, Female)
Goat (male & Female)
Camel (male & Female)
2 What are the major reasons for keeping mixture of livestock?
Ouuuation PIOI
340 --8---·---. __·-.-.·-._.-.·---- .... - . . _ •.. - . .. --_.. - ._. - _ ...
320 m
300 4
280 III
6
260
240 )
220 ~
5
13
El
200 24 ra
180 26 IJ 22 25
160 II 10 ID I!I 2
140 III
32
14
120 9
11 IR s
.r~il 27 30el 11 111100 15
C':I 31 I!I n
"0 80 fJe III
0 60 16
C~J 40 [) 12
00 B [J
< 20 211
U 0 El
Cl I!I·20
·40
·60
·80
·100
·120
·140
·160
·180
20
·200 I
0 50 100 150 200 250 JDD 350 400 450 500
DCA First axis
Appendix 5.1. Ordination of the different grass species using DECORANA in Awash-
Fantale district. Key lo the numbers: 1= Cluvsopogon phunulosus; 2= Cymbopogon
commutatus; 3= Cymbopogon CXC(/\'(l/us4= Tetrapogen cenchriformis, 5= Dactyloctenium
aegypticum; 6= Aristida adoensis, 7= Eragros/is raccmosa; 8= Heteropogen contoruts. ')=
Sporobolus natalensis, I l)= Pantcum coloratum, II = Se/aria verticellata; 12= Cenchrus
ciliaris, 13= Cynodon dactylotr. 14= Urocliloa pauicoides, 15= Urochloa panicoidcs, 16=
Coclachrum poiflorunr, 17= Chloris pychnothrix, 18= Sorghum purpureoosericeum, 19= N\0
Euneappogon schimperanus; 20= Aristida adscensionis, 21 = Tragus berteronianus, 22= ~
Paspalm glll/lIacell",; 23= Sporobolus ioc/ac/os; 24= Cenchrus setigerus; 25= Paspalm
dila/a/llnr;26= Chloris roxburghiana, 27= Lintoma /l1I/(II/S; 28= Sporebolus [estivus, 29=
Sporobelus spica/us; 30= I lyparrhcnia lrir/a) I = Digitaria ternata; 32= Ilurcground.
340 -----------1"112-----1- ------ ------------ -----------------------------------
320 0 lil
300 B I!I
2BO
260 32 2915 13 12
240 15!JD lil 1;1'6
lil -
220 I!I _ IIIJ 14
200 ~
lBO 7 lA'
160 5
140 III 31 30 2 lil 34
120 lil I!
100 25 lil I!I la 3B
WIJD
'" 60 27
';;;J 40 19
« 20o IJ IJ6
"c0: -20 lil 4
o ~O B
~ -60 3
IJ) -BO ~6 10
~ -100 III I!I
U -120
-140
~ -160
-180
-200
-220
-240
-260
-280
-300
-320
-340
-360
-380
-400
. -'4·40-4601 ·!!lIG
-480
-500 II
o 50 100 150 200 250 300 350 400
DCA First axis
Appendix 5.2, Ordination of the different grass species usmg DECORANA 10
Kereyu- Fantale district, Key to the numbers: 1= Urochloa panicoides; 2= TragIIS
berteronianus; 3= Themeda triandra; 4= Tetrapogon cenchrifonnis; 5= Sporobolus spica/us; 6=
Sporobolus natalensis. 7= Sporobelus festivus; 8= Sorghl/III purpureosericeum; 9= Se/aria vstilata; 10=
Setaria verticellata; II = }'''lIIli,I'I'/UIII niezinnunr, 12'~ Paspahn ditatatum; 13= Paspalm glnmacenm, 14=
Pantcum snowdenu; 15~ }'Oll/CI/III coloranon. 16= Mierochtoa kunhtii; 17~ Lintoma nutans; 18=
ll yparrhenia IIir/a: 19= Heteropogen contrortus; 20 = Eragrostis racemosa; 21 = Eragrostis racemosa; N
21'" Eragrostis cilianensis, 22= Enneapogan schunperanns; 23= EIeusine indico; 24= Digitaria ternata; \0
25~ Dignoria nulanjuma; 26= Dactyloctenium aegypticum; 27= Cynodon dactylon; 28= Cymbopogon VI
e_,'cava/,{.\'; 29= Cymbopogon COIIIIIIIIWIIIS; 30= Coelachrum poiflorum; 31 = Chrysopogon p/WIW/OSIIS; 32=
Chlons roxburghiana, JJ= Chloris ga)'ana; 34 = Cenchrus setigerus; 35= Cenchrus ciliaris; 36=
llothriochloa rudicans: ) 7= Aristida adscensionis; J8= Bare ground; 39= Aristida adoensis.
296
Appendix 5.3. Correlation matrix among variables studied in Awash-Fantale district (NS= Non significant*= P < 0.05; ** = P<O.Ol; *** = P <
0.001). Correlation coefficients for the vacant cells were non-significant (P >0.05) (GY; grass orv1 yield kg ha"; BC; % basal cover; BG; % bare ground; ESE; estimated soil erosion; ETTE~ Evapotranspuration
tree equivalent ha"; EC; Electrical conductivity ds m"; Sa; Sand (%); Si; Silt (%) ; Cl; Clay (%); TN%; % total nitrogen; OC~·é % organic carbon; eN; carbon nitrogen ratio; p; available phosphorous; K;
available potassium; AJ; altitude).
Paraniet CY nc DG ESE ETTE I'll EC Sa Si Cl TN% DC% CN P K Al
CI'S
GY 0.59 -0.25 -0.31 -0.65 0.13 0.16 0.32 -0.02 -0.38 (J.oJ -(J.05 -0.16 -0.14 0.06 0.33
BC •• -0.57 -0.13 -0.47 -0.18 -0.22 0.25 0.08 -0.38 0.18 0.12 -0.15 -0.17 0.07 0.18
DG NS •• 0.08 0.04 0.52 054 0.04 -0.29 0.21 -0.42 -0.38 0.06 0.18 -0.05 -0.04
ES": • NS NS 0.49 -0.40 -0.4:1 -0.62 0.29 0.52 0.15 0.46 0.58 0.14 0.11 0.02
ETTE ••• • NS •• -0.33 -0.32 -0.51 0.25 0.43 0.14 0.35 0.41 0.29 0.12 -0.07
Pil NS NS .••. • NS 0.97 0.70 -0.62 -0.31 -0.71 -0.78 -0.18 0.04 -0.07 -0.20EC NS NS • NS ••• 0.69 -0.54 -0.38 -0.67 -0.77 -0.27 0.16 0.05 -0.27 I
Sa NS NS NS ••• •• ••• ••• -0.58 -0.73 -0.46 -0.74 -0.54 -0.06 -0.16 -0.34
Si NS NS NS NS NS ••• •• •• -0.13 0.36 0.41 0.01 0.29 0.18 -0.17
Cl NS NS NS •• NS NS • ••• NS 0.23 0.54 0.68 -0.15 0.04 0.56
TN% NS NS • NS NS ••• ••• • NS NS 0.77 -0.22 -0.13 0.20 0.40
DC% NS NS NS • NS ••• ••• ••• • •• ••• 0.41 -0.10 0.31 0.50
eN NS NS NS NS • NS NS •• NS ••• NS • -0.01 -0.01 0.27
Il NS NS NS NS NS • NS NS NS NS NS NS NS 0.12 -0.24
K NS NS NS NS NS NS NS NS NS NS NS NS NS NS 0.20
AI NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS--
.~
297
Appendix 5.4. Correlation matrix among the variables studied in Kereyu-Fantale district. (NS= Non significant*= P < 0.05; ** = P< 0.01; *** =
P < 0.001); .GY= grass OM yield kg ha"; BC= % basal cover; BG= % bare ground; ESE= estimated soil erosion; ETrE= evapotranspuration tree equivalent ha"; EC= electrical conductivity ds m"; Sa= sand
(%); Si= silt (%); CI= clay (%); TN%= % total nitrogen; OC%= % organic carbon; eN= carbon nitrogen ratio; P= available phosphorous; K = available potassium; AI = altitude).
Parameter GY BC BG ESE ETT PH EC Sa Si Cl TN% OC% CN P K Al
s E
GY 0.61 1-0.69 0.27 -0.50 -0.52 -0.13 -0.13 0.17 0.05 -0.12 0.20 0.68 -0.27 0.59 0.46
BC *** -0.74 0.30 -0.25 -0.31 -0.19 -0.08 -o.u 0.03 0.19 0.25 0.23 -0.16 0.53 0.31
BG *** *** -0.05 0.41 0.36 0.32 -0.14 0.21 -0.02 -0.18 -0.30 0.29 0.24 -0.40 -0.32
ESE NS NS NS 0.20 -0.08 0.06 -0.27 0.27 0.17 -0.17 0.01 0.29 0.08 0.44 0.17
ETTE ** NS * NS 0.15 -0.28 -0.06 0.08 -0.01 0.25 0.19 -0.11 0.54 -0.25 -0.19
PH ** NS * NS NS 0.61 0.39 -0.-t8 -0.16 -o.n -0.55 -0.60 0.08 -0.20 -0.50
EC NS NS NS NS NS *** 0.22 -0.21 -0.16 -0.57 -0.62 -0.14 -0.06 0.09 -0.29
Sa NS NS NS NS NS * NS -0.90 -0.80 -0.26 -0.38 -0.11 -0.37 -0.35 0.04
Si NS NS NS NS NS ** NS *** 0.46 0.21 0.40 0.29 0.37 0.36 0.08
Cl NS NS NS NS NS NS NS *** * 0.26 0.24 -0.15 0.25 0.24 -0.19
TN% NS NS NS NS NS NS *** NS NS NS 0.90 -0.21 0.27 0.02 0.16
OC% NS NS NS NS NS ** *** * * NS *** 0.17 0.20 0.16 0.44
CN *** NS NS NS NS *** NS NS NS NS NS NS -0.21 0.16 0.59
P NS NS NS ** ** NS NS * * NS NS NS NS 0.29 -0.49
K ** ** * * NS NS NS NS NS NS NS NS NS NS -0.09
AI * NS NS NS NS ** NS NS NS NS NS * *** ** NS
- ---- -- - ----- - -- - I