AWilliams PHD Thesis

People cascades, land and livelihoods: Farmer and
herder land-use relations in the Idodi rangelands,

Tanzania
Andrew Williams
PhD Thesis
University College London
1
Thesis submitted in fulfilment of the requirements for the Degree of Doctor of Philosophy,
University College London, 2005.
2
Abstract
Land policies in Africa have often been predicated on marginalising or extinguishing customary land
tenure systems in favour of introducing more ‘efficient’ formal systems of individual titling and
registration of land. This approach has been marked by its frequent failure and high cost. In Tanzania,
recently introduced land laws instead now recognise customary systems and set out a basis for
incorporating them in a village-based land tenure system. Yet there is growing apprehension that
placing an emphasis on recognising customary practices will compound the growing trend of social
differentiation, elite capture and the increasing numbers of landless poor. These issues can be better
understood through investigating who benefits and loses from instances of ‘negotiability’ in access to
land at a local level, particularly in the light of broader political economic and social changes.
Based on field work carried out in central Tanzania, the study traces the socio-environmental outcomes
of herders and farmers living in the Idodi rangelands. Over the last 50 years, a substantial portion of
these rangelands have been taken over by the state for the creation of wildlife conservation areas. The
remaining parts of the rangelands have been settled by successive waves of farmers and herders,
mainly associated with evictions from the creation of protected areas, other state-perpetrated land
alienations in northern Tanzania, and state-enforced villagisation. Over time, the continued immigration
of people into the Idodi villages has added to an already growing population, such that today, key
resources - fertile arable land, grazing and water - are in increasingly short supply. The story of the
Idodi rangelands reflects developments occurring in many other parts of Tanzania. In particular,
wetland areas in the dryland rangelands have become a focus of in-migration and heightened
competition for land and water, as farmers and herders alike converge on these centres of relatively
high fertility and productivity. Often, as in the Idodi rangelands, competition for land and water has
grown sufficiently great for conflict to break out in these polyethnic dryland-wetlands.
The social negotiability of land has remained central for herders’ access to key land and landed
resources. In the Idodi rangelands, herders have used their growing social relations with farmer-based
centres of power to avoid conflict and maintain access to farmland. Contrastingly conflict over land has
occurred when other herders have not sufficiently invested in social relations with farmers over access to
land. Herders continue to remain squatters - albeit socially legitimate ones - on village land, without
firm rights to rangeland resources. In recent years strong social relations have not been sufficient to
guarantee herders’ security in the landscape. It is clear that the land entitlements of marginalised
herder groups may often need safeguarding by the government, but it less clear what the best
approach may be. In Idodi, a more overt expression of pastoralists’ rights to land would likely lead to
polarisation between farmer and herder, and an increase in conflict and competition over land. Too
little consideration has been given by the government to enabling the pluralistic yet equitable
development of locally diverse customary understandings of land tenure. The continued increase in
competition and conflict over access to land - as has occurred in Idodi - strongly suggests that priority
should to be given in land reform processes to the development of locally legitimate dispute resolution
fora that focus on negotiated outcomes wherever imposed adjudicatory decisions can be avoided.
3
Contents
Acknowledgements ..................................................................................................................... 12
1 Introduction .......................................................................................................................... 14
1.1 A study overview ............................................................................................................................................ 14

1.2 The thesis structure .......................................................................................................................................... 16
2 ‘Safeguarding’ environment and productivity ...................................................................... 18
2.1 Received knowledge, past policy and new understandings ................................................................. 18

2.1.1 Received knowledge and past policy ....................................................................................... 18
2.1.2 New understandings ................................................................................................................... 19
2.2 Recent theoretical advances in the society and environment debate ................................................. 22
2.2.1 Collective action: Common pool resource and moral entitlement theories ........................ 22
Common pool resource theory ............................................................................................................... 22
The moral entitlement school .................................................................................................................. 24
2.2.2 Political ecology.......................................................................................................................... 24
2.3 Contending with new paradigms of complexity in environment and society..................................... 27
2.4 Land and institutions: social embeddedness, inequality and conflict ................................................... 30
2.4.1 The evolutionary model of land tenure.................................................................................... 33
2.4.2 Communitarian understandings of land tenure ...................................................................... 35
2.5 Mainstreaming new understandings of society and environment in policy ........................................ 36
2.6 The study approach, analytical framework and methodology ............................................................ 38
2.6.1 The study context ........................................................................................................................ 38
2.6.2 The research questions and analytical framework ................................................................. 39
Research questions ................................................................................................................................... 39
The analytical framework....................................................................................................................... 40
2.6.3 The study methodology .............................................................................................................. 42
3 Tanzanian semi-arid rangelands in perspective................................................................... 44
3.1 Introduction ...................................................................................................................................................... 44

3.1.1 Defining the semi-arid rangelands ........................................................................................... 45
3.2 Some key developments affecting rangeland management in Tanzania .......................................... 48
3.2.1 Colonial partitioning and landscape re-organisation ........................................................... 48
3.2.2 Post independence state control: Ujamaa, deconcentration and decentralisation ............ 50
The early independence years.............................................................................................................. 51
The Arusha Declaration and the onset of Ujamaa ............................................................................ 52
Deconcentration of state power ............................................................................................................ 55
The end of Ujamaa: decentralisation, liberalisation and local government reform .................. 55
3.2.3 Control and space: state and nature ....................................................................................... 57
3.2.4 New Tanzanian land tenure legislation and its implications for herders and farmers ..... 62
3.2.5 People cascades: landlessness, insecurity and wanderings ................................................... 65
3.3 Conclusion......................................................................................................................................................... 69
4
4 The dryland-wetland frontier of Idodi and Pawaga ............................................................. 71
4.1 Introduction ...................................................................................................................................................... 71

4.2 The Greater Ruaha Ecosystem: Idodi and Pawaga ................................................................................ 71
4.3 Idodi and Ikwavila valley ............................................................................................................................. 73
4.3.2 The agro-ecology of Idodi and the Ikwavila valley .............................................................. 77
4.3.3 The people and socio-economy of Idodi and the Ikwavila valley ....................................... 78
4.4 Village-based administration and land tenure ........................................................................................ 82
4.4.1 Village governments in Idodi .................................................................................................... 82
4.4.2 Local level courts......................................................................................................................... 85
4.4.3 Land tenure categories and practices in Idodi ....................................................................... 87
4.5 Pastoralists and village government in Idodi: Living on the edge or edging in? .............................. 92
5 The peoples of ‘Kwigongo’: the old and the new ................................................................ 95
5.1 Alienations and catastrophe: the loss of the old way of life ................................................................ 95
5.1.1 The old hamlets and the way of life in early colonial times ................................................. 95
5.1.2 Increasing epidemiological and ecological challenges ....................................................... 101
5.1.3 Depopulation and eviction ....................................................................................................... 103
5.2 The Ikwavila Valley .................................................................................................................................... 105
5.2.1 The populating of the Ikwavila valley ................................................................................... 105
5.3 Livelihoods..................................................................................................................................................... 111
5.3.1 Households ................................................................................................................................. 111
5.3.2 Farming and the farming calendar: ....................................................................................... 115
5.3.3 Farm holdings ............................................................................................................................ 120
5.3.4 Cultivation, yields and agricultural productivity .................................................................. 128
Cultivation ................................................................................................................................................ 128
Yields ........................................................................................................................................................ 129
Household agricultural productivity.................................................................................................... 130
Summary .................................................................................................................................................. 133
5.3.5 Dryland agriculture: Rainfall, crop yields and soil fertility ................................................ 136
Farmer narratives: changing rainfall patterns ................................................................................. 136
Farmer narratives: declining crop yields ........................................................................................... 136
Understanding locally held wisdoms .................................................................................................. 137
Rainfall and soil fertility: synergistic dynamics? .............................................................................. 138
Expensive investments and uncertain returns .................................................................................... 138
5.3.6 ‘Twililage pe twiwumi!’ - Let us eat while we are still alive!............................................... 138
5.3.7 Wetland agriculture: rainfall, crop yields and soil fertility ............................................... 139
High inputs and high returns: predictability and profits ................................................................ 139
5.3.8 Livelihood options ..................................................................................................................... 140
Livelihood diversification ...................................................................................................................... 140
Debt, market speculation and value adding .................................................................................... 140
5.4 Conclusion...................................................................................................................................................... 141
6 The Ilparakuyo of Idodi: Contending with change ............................................................... 143
5
6.1 The establishment of the Ilparakuyo in Idodi ......................................................................................... 144
6.1.1 Pre-Ujamaa: new frontiers ...................................................................................................... 144
6.1.2 Ujamaa and its consequences.................................................................................................. 147
6.2 The Barabaig in Idodi ................................................................................................................................ 148
6.2.1 The arrival of the Barabaig in Idodi...................................................................................... 149
6.2.2 The Barabaig domestic unit, development cycle and household economy ....................... 150
6.2.3 Barabaig herding in Idodi ....................................................................................................... 151
6.2.4 Barabaig and farmer land-use relations ............................................................................... 152
6.2.5 Social organisation, property relations and jural institutions ............................................ 153
6.3 Ilparakuyo landscape occupancy ............................................................................................................. 154
6.3.1 Pastoral ecology and range use patterns ............................................................................. 155
6.3.2 Livestock disease and range-use patterns ............................................................................. 158
6.3.3 Range agreements, range tenure and farm fields ............................................................... 160
6.3.4 Farming ...................................................................................................................................... 161
6.4 Livelihoods..................................................................................................................................................... 162
6.4.1 Households ................................................................................................................................. 163
6.4.2 Herd structure and management ............................................................................................ 167
Analytical considerations ...................................................................................................................... 169
Herd structure and stock distribution .................................................................................................. 173
Herd transactions.................................................................................................................................... 181
6.4.3 Farming ...................................................................................................................................... 184
The significance of hired labour in cultivation .................................................................................. 188
6.4.4 Livelihood strategies ................................................................................................................. 188
Herd decline and comparative livestock holdings ........................................................................... 195
6.5 Conclusion...................................................................................................................................................... 197
7 Negotiating the political ecology of landed resources ....................................................... 199
7.1 Introduction ................................................................................................................................................... 199

7.2 The socio-ecological context to resource dispute: what does land mean and to whom? ............. 200
7.2.1 Farmers: expanding frontiers, fertility decline, and new technology ............................... 201
7.2.2 Herders: seasonal grazing systems, demarcated landscapes and expanding fields ...... 202
7.2.3 Farmers and herders: complementary and conflicting land-use practices ....................... 203
Defining terms and describing the nature of disagreement.......................................................... 203
Complementary land-use practices .................................................................................................... 204
Conflicting land-use practices .............................................................................................................. 205
Changing land-use priorities ................................................................................................................ 206
Summary .................................................................................................................................................. 208
7.3 Social affiliations, land-use and conflict ................................................................................................. 208
7.3.1 Law and social process............................................................................................................. 208
7.3.2 Social affiliations and negotiated outcomes between herder and farmer ....................... 209
Group: herder - farmer and land-use agreements in Mahuninga ward .................................... 210
Leaving a rights-based dispute unresolved - Mahuninga village ................................................ 213
Resolving disagreement through negotiation - Makifu village ..................................................... 214
6
Networks: individual herder-farmer relations and land-use outcomes ....................................... 215
Groups and networks: ‘on-stage’ and ‘off-stage’ discourses ........................................................ 216
Summary .................................................................................................................................................. 217
7.3.3 Conflict and the undermining of social affiliations.............................................................. 217
The context to conflict: resource use pressure .................................................................................. 218
Explaining Barabaig - farmer dispute .............................................................................................. 218
Crisis and beginning the negotiation of landscape use ................................................................. 220
The case of the Ilparakuyo: undermined social affiliations and landscape occupancy .......... 223
Conclusions ............................................................................................................................................... 224
7.3.4 Controlling the terminology .................................................................................................... 225
7.4 Conclusion: Landscape and livelihood outcomes ................................................................................... 227
8 Conclusion ......................................................................................................................... 231
8.1 Disaggregating trends in land-use relations in Idodi........................................................................... 231

8.2 Negotiated land-use agreements: too much flexibility? ..................................................................... 234
8.3 Looking to the future: the case of herders in Idodi............................................................................... 234
8.4 Business as usual? ......................................................................................................................................... 237
A1: A short overview of Local government in Tanzania ...................................................... 239
A2: A FAO land cover scheme for the Idodi and Pawaga rangelands ................................. 243
A3: A Provisional Gazetteer of the Idodi and Pawaga Villages ........................................... 245
A4: The reported incidence of disease in the Idodi livestock herd ........................................ 254
A5: Age grades and age sets of the Ilparakuyo in Idodi ........................................................ 255
Bibliography .............................................................................................................................. 256
7
Figures
Figure 3.1: The topography of Tanzania, location of major wetlands and general tsetse fly
distribution ....................................................................................................................................................... 46
Figure 3.2: Average rainfall distribution for Tanzania ................................................................................... 47
Figure 3.3: Wildlife protected areas in Tanzania in 2004 [not including forest reserves] (Baldus
and Cauldwell 2004) .................................................................................................................................... 60
Figure 4.1: The location of the study site in Tanzania - the Idodi villages, Iringa District........................ 72
Figure 4.2: The villages of Idodi, herder homesteads and land-use areas ................................................ 74
Figure 4.3: The Lunda-Mkwambi Game Controlled Area looking west from Idelemule Mountain
in Tungamalenga in the dry season ............................................................................................................ 76
Figure 4.4: The Ikwavila valley looking south towards Mahuninga from Tungamalenga in the
dry season........................................................................................................................................................ 76
Figure 5.1: The Ruaha River valley during the British colonial period circa 1950 .................................... 97
Figure 5.2: The Ruaha River valley circa 2003 ................................................................................................ 98
Figure 5.3: The Ikwavila valley during the late British colonial era - circa 1950 .................................. 106
Figure 5.4: The Ikwavila valley today - circa 2003 ..................................................................................... 107
Figure 5.5: The history of field allocation in the Ikwavila valley in the last 50 years ........................... 108
Figure 5.6: The distribution of household sizes in the Ikwavila valley as measured in total
people per household (all adults and children). ................................................................................... 112
Figure 5.7: The distribution of household sizes in the Ikwavila valley as measured in Reference
Adults ............................................................................................................................................................. 112
Figure 5.8: A Wetland ‘bonde’ rice fields with a ‘vynungu’ banana patch in the distance .................. 116
Figure 5.9: Harvested dryland ‘nchi kavu’ fields with nitrogen fixing Acacia albida trees................... 116
Figure 5.10: Overall land tenure patterns in the Ikwavila valley .............................................................. 121
Figure 5.11: The pattern of household access to all arable land in the Ikwavila valley ...................... 122
Figure 5.12: The equitability of overall land holding per household by relative inferred wealth
status .............................................................................................................................................................. 123
Figure 5.13: The origin of land currently owned by households ................................................................ 124
Figure 5.14: The pattern of household access to lower value dryland suitable for wet-season
cultivation only (‘nchi kavu’) in the Ikwavila valley............................................................................... 125
Figure 5.15: The average size of dryland holding for households categorised by relative
(inferred) wealth status. ............................................................................................................................. 125
Figure 5.16: The pattern of household access to high value garden land suitable for dry-
season cultivation (‘bustani’ / ‘vynungu’) in the Ikwavila valley. ....................................................... 126
Figure 5.17: The average size of riverine garden holding for households categorised by
relative (inferred) wealth status. .............................................................................................................. 126
Figure 5.18: The pattern of household access to high value land suitable for rice cultivation
(‘bonde’) in the Ikwavila valley (n=234). Fields were either owned or rented. ............................ 127
Figure 5.19: The average size of irrigated rice field holding for households categorised by
relative (inferred) wealth status. .............................................................................................................. 127
8
Figure 5.20: The proportion land cultivated in relation to total land held by households
categorised by relative (inferred) wealth status. ................................................................................. 129
Figure 5.21: The inter-annual variation in overall crop yields for individual fields (sample sizes
individually indicated for each data set in parenthesis). .................................................................... 130
Figure 5.22: Total individual ranked household crop production for the year 2000 as
measured in Cash Equivalents per Reference Adult. ........................................................................... 131
Figure 5.23: Total individual household crop production for the year 2001 as measured in
Cash Equivalents per Average Adult Male Equivalent........................................................................ 131
Figure 5.24: Total individual household crop production for the years 2000 and 2001
compared as measured in Cash Equivalents per Average Adult Male Equivalent. ...................... 132
Figure 5.25: Inter-annual household dryland crop production measured in cash equivalents per
reference adult and classified into inter-quartile ranges. .................................................................. 134
Figure 5.26: Inter-annual household wetland crop production measured in cash equivalents per
Figure 5.27: Inter-annual household total crop production measured in cash equivalents per
Figure 5.28: Women brewing mbege together - by cooking the maize gruel wort in the early
stages of the brew-process ....................................................................................................................... 135
Figure 5.29: A group of men playing ‘bao’ (a popular board game) at ’kilubu’ (beer club) in
Mahuninga. ................................................................................................................................................... 135
Figure 6.1: Grazing farm fields during the early dry season .................................................................... 157
Figure 6.2: Chamyina mountain and highland scarps looking south from Idodi village
rangelands near Kibikimuno ..................................................................................................................... 157
Figure 6.3: Tending small stock in the enkang’ in the early morning ......................................................... 168
Figure 6.4: Ilaiyok with their favourite oxen ................................................................................................... 168
Figure 6.5: Livestock holdings for each olmarei during the year 2000 .................................................... 176
Figure 6.6: Livestock holdings for each olmarei during the year 2001 .................................................... 176
Figure 6.7: A dam built by an Ilparakuyo pastoralist to irrigate his crops .............................................. 179
Figure 6.8: The irrigated field with a crop of onions nearly ready for market ...................................... 179
Figure 6.9: Areas cultivated by each Olmarei for the two major crops - maize and rice - in
2000. ............................................................................................................................................................. 185
2001. ............................................................................................................................................................. 185
Figure 6.11: The relative comparative index value (in cash equivalents) of combined Ilparakuyo
livelihood strategies for 2000 .................................................................................................................. 189
Figure 6.12: The relative comparative index value (in cash equivalents) of combined Ilparakuyo
livelihood strategies for 2001 .................................................................................................................. 189
Figure 6.13: The inter-annual variation in livestock (both large and small stock) available per
reference adult ............................................................................................................................................ 194
Figure 6.14: The inter-annual variation in area cultivated per reference adult ..................................... 194
9
Tables
Table 4.1: The population of Idodi by village in 2000 .................................................................................. 75

Table 4.2: Population growth in the Idodi villages between 1976 and 2000 .......................................... 75
Table 4.3: Reported livestock numbers in Idodi Division in 1984 and 1999 ............................................. 81
Table 4.4: Land tenure categories in the Idodi villages .................................................................................. 88
Table 5.1: An overview of some significant events in the Idodi and Pawaga rangelands
impacting on people’s livelihoods, 1916-1974 .................................................................................... 103
Table 5.2: The origin of heads of households living in the Ikwavila valley.............................................. 110
Table 5.3: The dependency ratio of households analysed by villages .................................................... 113
Table 5.4: The dependency ratio of households analysed by derived wealth index ........................... 114
Table 5.5: The lowland Hehe farming calendar and Swahili equivalent. ................................................ 117
Table 5.6: The different types of soil recognised by farmers in the Ikwavila valley. ........................... 118
Table 5.7: Overall average inter-annual agricultural cultivation and productivity ............................... 132
Table 6.1: The varying compositional number of households (inkajijik) in Ilparakuyo super-
households (ilmarei) ..................................................................................................................................... 165
Table 6.2: Super-household (olmarei) and household (enkaji) sizes measured in Adult
Equivalents .................................................................................................................................................... 165
Table 6.3: Super-household and household characteristics presented by age-set as an indicator
of life-cycle stage ....................................................................................................................................... 166
Table 6.4: Ilparakuyo large stock (cattle - Bos indicus) herd composition and structure for 2000
and 2001 ...................................................................................................................................................... 171
Table 6.5: Ilparakuyo small stock (goats and sheep) herd composition and structure for 2000
and 2001 ...................................................................................................................................................... 172
Table 6.6: The overall proportion of ilmarei in different livestock wealth categories........................... 180
Table 6.7: Average small stock holding in ilmarei of different large stock wealth categoriesa.......... 180
Table 6.8: The livestock holdings of super-households at different stages of developmental
cycle ............................................................................................................................................................... 181
Table 6.9: The proportion of different transactions reported for large stock leaving the
collective herd. ............................................................................................................................................. 183
Table 6.10: The proportion of different transactions reported for large stock entering the
collective herd .............................................................................................................................................. 183
Table 6.11: The relative increase from stock loans in large stock available to ilmarei of
different livestock wealth categories ...................................................................................................... 184
Table 6.12: The proportion of ilmarei in different livestock wealth categories loaning large
stock in and out ............................................................................................................................................ 184
Table 6.13: Ilparakuyo ilmarei cultivation averages in 2000 by crop area according to
livestock wealth category .......................................................................................................................... 187
livestock wealth category .......................................................................................................................... 187
Table 6.15: The proportion of Ilparakuyo ilmarei cultivating rice .............................................................. 187
Table 6.16: Employment of labour in Ilparakuyo cultivation ....................................................................... 188
10
Table 6.17: Relative importance of farming in the production of individual ilmarei ............................. 190
Table 6.18: Overall total productivity indices (in cash equivalentsa, b) of ilmarei grouped by
livestock wealth categories ....................................................................................................................... 191
Table 6.19: The impact of olmarei (household) size on production and wealth status .......................... 193
Table 6.20: Comparative livestock holdings between the contemporary Ilparakuyo of Idodi, the
Il Parakuyo of Bagamoyo District and the Barabaig of Hanang District ........................................ 195
Table 7.1: An inventory of court cases heard by the Baraza la Mahkama ya Kata for 2000 and
2001 .............................................................................................................................................................. 207
Table 7.2: Ilparakuyo enclosures in Mahuninga Ward in 2000-2002..................................................... 211
Boxes
Box 5.1: The methodology used to generate the derived wealth index for households ...................... 114
Box 6.1: Some of the more important grass, forb and tree/shrub species for Ilparakuyo
livestock in the Idodi rangelands.............................................................................................................. 158
Box 6.2: A description of the wealth categories adopted in relation to Ilparakuyo livelihood
strategies....................................................................................................................................................... 175
Box 6.3: A growing ilmarei exerting greater demands on its livestock herd........................................... 177
Box 6.4 An Ilmarei investing in agricultural production as a central livelihood strategy ...................... 178
Box 7.1: Farmer opinions about pastoralists and herder-farmer land-use conflict ................................ 222
Appendices
Figures
Figure A1: The normative structure and functioning of local government in the Tanzanian
mainland ........................................................................................................................................................ 241
Tables
Table A1: Key developments in village-level government and administration - modified from
Shivji and Peter (2000, 46-53) ................................................................................................................ 239
Table A2: A provisional gazetteer of former settlements in the Ruaha River valley in what is
now the Lunda-Mkwambi Game Control Area and the southern extent of the Ruaha
National Park ............................................................................................................................................... 245
Table A3: The reported incidence of different diseases in the Idodi livestock herd in 1991 .............. 254
11
Acknowledgements
I wish to thank the Central Research Fund of London University, the Graduate School and the
Department of Anthropology at University College London, and particularly the Leverhulme Trust for
supporting this research. In addition, I would like to thank the Institute of Resource Assessment at the
University of Dar es Salaam for supporting my research clearance in Tanzania, and the Commission
for Science and Technology for granting me permission to carry out my research. I would like to
acknowledge the Tanzania National Archives and Rhodes House Oxford for kindly allowing me
access to their archives. I’d like to thank the support staff in the Department of Anthropology – in
particular Cassie Hill, Alena Kocourek, Helen Cooper and Chris Hagisavva for all their help.
I am indebted to the staff of MBOMIPA project, Miss Dorothy Bikurakule, Mr K.A.S. Ngomello, Mr
James Mudalbilwa, Mrs Kezia Oola, Mr Mashaka Omari, Mr Aston Mwela and Mrs Dorothy
Mwandimbile, who were always welcoming whenever I showed up at the MBOMIPA office to pick-up
my mail and to use the office as a base while in Iringa Town. They also kindly welcomed me into their
homes. I would also like to thank members of the District Natural Resource Office, for their advice and
support, including allowing me access to their files. In particular, I’d like to thank Mr A. Mallango the
District Natural Resource Officer and Mr I. Kimaro, the District Game Officer as well as Mr J. Hante
the District Land Officer.
In the Idodi villages, I’d like to acknowledge and thank the village executive officers, village
chairpersons and ward executive officers who allowed me access to their files, entertained my
questions and helped me on my way. In particular I’d like to thank Mr T. Makendi, Mr A. Nganylika,
Mr R. Mkwele, Mr T. Mwampashe, and Mr J. Zayumba. In addition, I would like to warmly
acknowledge the late Daudi Mbarazi, previously District Councillor for Idodi, who kept a paternal
eye on me, and who introduced me to many folks. I would also like to especially acknowledge and
thank Mr V. Mwaikambo, and Mr J. Hango.
I owe a debt of gratitude to my field assistants Mr Luca Chengula Mr Ng’awa Mgemaa, Mr Aucelio
Mbata, Mr Onesimo Kiwele and Mr Deli Deda. Were it not for them, this thesis would not have been
completed. In addition, I would like to sincerely thank Mr George Mtati and Mr Saleh Petwa who
were a pleasure to talk to and an amazing font of local knowledge and folklore. In addition I would
like to thank the following people: Mr Mikaelo Mdindile, Mr Juma Katei, Mr Clemek Temisika, Mrs
Rosina Dalu, Mr Kipilpili Lebere, Mr Mahagano Chambulila, Mr Salum Mtemisika, Mrs Lolyian Katei,
Mr Falimbopelo Mdindile, Mr Juma Katei, Mr Wilson Mpwepwa, Mr Selindoo Mtemisika, Mrs
Mwigavilo Mkuvasa, Mr Assecheck Gissemoda, Mrs Kilwa Sekiwele, and Mr Yussuf Kimojaa.
I would like to thank Dr David Brown at the Overseas Development Institute, London for providing
part-time employment while I was writing up in London, and I’d like to thank Dr Alan Rodgers in East
Africa for the opportunity to work on various projects, partly in support of my extended writing up.
12
There are some people who I will never be able to thank enough – and I only hope that I am able to
pass on their kindness and support to others in the future. In particular, Martin and Asha Walsh very
generously welcomed me into their home (and study!) in Iringa, providing me with a base and a room
to stay in whenever I was in town, and were simply great. In Tungamalenga, I would like to sincerely
thank Jonas Padri, Mr and Mrs Mlewa, Godfrey, Humphrey and Ditrik with whom I lived and who
were so hospitable and generous.
This thesis has been a bit of an unexpectedly epic undertaking – and the subject of numerous
humorous jibes from a number of friends as to the time it’s taken to complete. I thank Mhairi, Clare,
Nadine, Steve, Beth, Fiona and Eshetu at UCL for their collegial support. And I thank Jan, Sharon,
Caspar, Niek, Liam, Alessandra, Sally, Howard, Dave, Grant, Hanneke, Bakari and Mafunda, Thabit,
Fred, Liz, Tim, Jo, Tom, Katie, Chris, Shaffin, Bijou, the Lofters crew, Zoe, Annabelle and not least,
Weh! and Maria, for their amused support and humour. Finally, I would like to especially thank my
kind mother on whom I descended during my final stages of writing up and who wonderfully allowed
me to look after myself and tactfully remonstrated with the worst of my infringements on the English
language.
I would like to very much thank Phil Burnham for his gentle and wise mentoring, and for so effectively
hiding what must have been his considerable frustration with my academic deficiencies and hair-
brained moves between the UK and East Africa. And finally, my sincere thanks go to my main
supervisor, Kathy Homewood, whose patience and faith in my dawdling efforts have been
remarkable, and her support unerring. I am truly grateful for the time Kathy has invested in me – and
I hope that, kwa hali na mali, it’s been – and will be – worthwhile! Ashe oleng!
This thesis is dedicated to my father who did not live to see it completed.
13
1 Introduction
1.1 A study overview
This thesis explores the livelihoods and land-use relations of a diverse group of herders and farmers
living in the Idodi rangelands of northern Iringa District, central Tanzania. The thesis is based on field
work carried out in the Idodi villages over a two year period between 2000 and 2001.
I began to realise after the first few months of having lived in Tungamalenga village, which was my
home for two years, that something was not quite right. I had not expected that my fieldwork talking
to people about their lives and livelihoods would be easy. I did expect a certain amount of wariness
from people until I became a better known character in and around the Idodi villages. After all, why
would a European come to live in Idodi? What was he doing? People were very suspicious and did
not understand. This despite my attempts to explain in different ways that I was a student trying to
understand how people in Idodi lived and how they farmed and used the land. Using the word ‘land’
was my mistake. I should not have mentioned it. Before long, rumours had begun to spread that I was
here in Idodi to steal the land. At the time, when I discovered that these (and other) rumours were
floating around, I was somewhat incredulous. Why ever would people think that I wanted to steal
their land from them? It was only over the next year or so that I truly began to understand the
significance of people’s suspicions. People had traumatic memories of being evicted from their homes
by the wazungu (Swahili: Europeans; Mzungu sing.) a long time ago. What the wazungu had started,
the serikali (Swahili: government) had completed with forcible finality during Tanzania’s years of
Ujamaa. People were simply apprehensive about losing their land again.
Over the last 50 years, a substantial portion of the Idodi rangelands have been taken over by the
state for the creation of wildlife conservation areas. The remaining parts of the Idodi rangelands
have been settled by successive waves of farmers and herders, mainly associated with evictions from
the creation of conservation areas, other state-perpetrated land alienations in northern Tanzania, and
state-enforced villagisation during Ujamaa. Over time, the continued immigration of people into the
Idodi villages has added to an already growing population, such that today, key resources - fertile
arable land, grazing and water - are in increasingly short supply.
The thesis investigates what impacts these major state land-use interventions have had on people’s
livelihoods and on land-use relations in Idodi. The thesis shows that there are a minority of people
who have managed to benefit from the villagisation of the Idodi rangelands, others who have fared
not quite as well, and a great many who have, at least in recent years, experienced significant
declines in their livelihoods and well-being. The thesis finds that, among other factors, the different
livelihood trajectories of herders and farmers living in the Idodi villages are increasingly related to
their ability to access and benefit from irrigated soil fertility. Thus a relative minority of people in
Idodi have grown relatively wealthy from building up their access to fertile and irrigated farmland,
drawing on the labour of a larger number of poor farmers now confined to marginally productive
14
land. Herders, as a collective minority, are struggling to maintain, let alone improve, their insecure
access to range resources for their livestock. Today, the declining availability and quality of key
natural resources in the Idodi villages has led to an escalation of dispute, most notably between
farmer and herder.
The story of the Idodi rangelands holds wider significance as it reflects developments occurring in
many other parts of Tanzania. Odgaard (2002, 77) describes an ongoing process of increasing
population, heavy internal immigration, alienation of large areas for wildlife conservation and/or
plantations, an increased focus by the state on cultivation (and implicitly the marginalisation of
pastoralism), and an expansion of the area cultivated nationally, that has led to substantial pressure
on arable land and pastures. In particular, wetland areas in the dryland rangelands have become a
focus of in-migration and heightened competition for land and water, as farmers and herders alike
converge on these centres of relatively high fertility and productivity. Often, as in the Idodi
rangelands, competition for land and water has grown sufficiently great for conflict to break out in
these polyethnic dryland-wetlands, particularly between herders and farmers.
The picture emerging from Tanzania may also be seen as part of a wider trend in sub-Saharan
Africa. Over the last decade, there has been a significant increase in published research and other
sources that describe mounting competition over land and landed resources across the continent
(Peters 2004, 270; Berry 2002, 639). Although as Peters states, there is much to support a view of
Africa as a continent with a majority of its population living on the land, without clearly defined
classes of landed and landless, she argues that this is only part of the story (Peters 2004, 270). A
growing body of evidence suggests that the other part is a less optimistic tale, filled with instances of
intensifying competition and conflict over land, deepening rifts between and within kin-based, ethnic
and regional groups, and of expropriation of land by local and non-local elites (Peters 2000, 270;
Berry 2002, 639). People’s access to land and landed resources across Africa may be most usefully
understood today as being customarily comprised of bundles of rights that are often negotiable and
flexibly mediated through social relations (Shipton 1994, 349 & 351). Yet as competition over land
continues to intensify, some groups of people continue to accumulate more land and landed resources,
while others continue to lose access, fall into higher levels of poverty and may become landless.
A number of countries in Africa have embarked on land reform processes through developing new
land laws and new approaches to administrating land. For many years land policies in Africa have
been predicated on marginalising or extinguishing customary land tenure systems in favour of
introducing more ‘efficient’ formal systems of individual titling and registration of land, an approach
that has been marked by its frequent failure and high cost. In Tanzania, recently introduced land laws
recognise customary systems and set out a basis for incorporating them in a village-based land
tenure system.
However, given the growing evidence of increasing competition and inequality over access to land,
there is growing apprehension, particularly given weak governance at local level, that placing an
emphasis on recognising customary practices will compound the growing trend of social differentiation
15
and the increasing numbers of landless poor observed across Africa (e.g. Peters 2004, 285). It may
be the case that the strengths of customary land practices (i.e. the social negotiability of land) have
been emphasised at the expense of a more candid analysis highlighting exactly who is included and
excluded and why. These uncertainties need to be better answered if current land reforms promoting
the principle of subsidiarity with increased emphasis on customary or local forms of land tenure are
not to be similarly blighted by the serious problems of elite capture and growing inequality and
insecurity over access to land, which have characterised previous land reform programmes (see
Chapter Two). These issues can be better understood through investigating who benefits and loses
from instances of ‘negotiability’ in access to land, an analysis that, in turn, needs to be situated in
broader political economic and social changes taking place, particularly during the last 30 years
(Peters 2004, 271).
The contemporary polyethnic nature of the Idodi villages has meant that land-use relations are a
complex product of people’s multiple interpretations of customary practice, pursuit of different socio-
economic interests and diverse cultural understandings of how the landscape should be managed.
Having established a detailed understanding of farmer and herder livelihood trajectories, I employ
extended case study to examine some of the processes of contingency, flexibility and negotiability
employed by pastoralists to modulate their land-use relations with farmers. I investigate - through my
examination of livelihood practices and people’s land relations in the Idodi rangelands - whether
there are situations and processes that can limit or end negotiation and flexibility in access to landed
resources for certain social groups and categories (see Peters 2004, 271). Focusing on farmer, herder
and local government relations, I set out to identify what some of these key emerging processes and
situations are that have enabled or constrained access by herders to key resources in the Idodi
rangelands. Finally while maintaining a focus on Idodi, I continually and critically relate the story of
Idodi to past and contemporary developments in Tanzanian natural resource management, macro-
economic and development policy.
1.2 The thesis structure
Having provided a short overview of what this study is about, the thesis begins with a review of some
of the key theoretical debates and contemporary developments in the study of society-environment
relations most relevant to better understanding the livelihoods of people living in the semi-arid
rangelands of Tanzania. At the end of this review, and drawing from the preceding discussion, I set
out the context for the thesis, the key questions, and the analytical framework and methodology
adopted. Chapter Three introduces the Tanzanian rangelands, and provides a review of the political
economy of some key pre- and post-independence state policies that have heavily impacted on
people’s customary occupancy and management of rangelands in Tanzania. The Idodi rangelands, as
the chosen site for field study, are introduced in Chapter Four. The Chapter reviews the human and
physical geography of the field site, and then provides an overview of key aspects of social and
administrative organisation and current land tenure practices that are relevant for the study’s
investigation of land-use relations in Chapter Seven. Two parallel studies are then followed in
Chapters Four and Five. The former is dedicated to the farming peoples of the field site, and a more
detailed context to their current landscape occupancy is constructed before a quantitative analysis of
16
their contemporary livelihoods is presented. The latter chapter is devoted to the herding peoples of
the study area, again following a similar pattern of providing a context to their presence in the
landscape before proceeding to present a quantitative analysis of their current livelihoods. The data
presented on the land-use practices and livelihoods of both herders and farmers in the preceding two
chapters, provides the context for Chapter Seven which focuses on land-use relations between
herders and farmers. I examine, through a series of extended case studies, how different groups of
herders have pursued different strategies towards maintaining their access to key resources, why
some strategies have been more successful than others, and why this has been the case. In Chapter
Eight I conclude by discussing the findings of my data chapters in relation to the main research
questions.
17
2 ‘Safeguarding’ environment and productivity
The existence, nature and causes of environmental degradation in East African rangelands have
become the focus of increasing debate in recent years. Long held perceptions that rural African
farming and herding systems, especially in the context of increasing human population, frequently
lead to environmental ‘degradation’ have been increasingly questioned. As reviewed by Beinart
(2000), recent scholarship has, over the last two decades or more, achieved much in painting over a
perhaps somewhat tired canvas of the certainties of environmental degradation and the established
narratives of the ‘dysfunctional’ relationship between rural African people and the environments in
which they live. In its place a rejuvenated and more dynamic canvas can now be found, portraying a
diversity of intertwining and cross-cutting themes that begin to reveal the complexities and context-
rich nature of people-environment relationships in Africa. Alternative understandings of the ecology
of the semi-arid rangelands of East Africa, new examinations of history and past policy as well as
more sophisticated economic and socio-cultural understandings of rural peoples’ ways of life have led
to different perspectives and interpretations of the status of East African rangelands, their peoples
and their drivers of change.
In this opening chapter, I set out a number of key theoretical themes that underpin past and
contemporary understandings of society - environment relationships which I consider most relevant to
the semi-arid rangelands of East Africa. I begin first with a short overview of recent and
contemporary developments in research on society - environment relations in the African semi-arid
rangelands. I then move on to review the two main methodological approaches that most relevant for
this study. Having chosen and justified my choice of method, I review two key themes that have
perhaps dominated the debate over society and environment in the African rangelands. I then briefly
consider the relationship between research and the development of policy which has heavily
influenced society - environment relations. These themes then lead into a discussion of the research
questions, analytical framework and fieldwork methodology that I adopt in this study.
2.1 Received knowledge, past policy and new understandings
2.1.1 Received knowledge and past policy

Since the early colonial period, received wisdoms1 of environmental degradation wrought by African
farmers and herders have heavily influenced the policies and practices of many contemporary
1 After Leach and Mearns (1996, 8) who define a ‘received wisdom’ as, ‘an idea or set of ideas sustained
through labelling, commonly represented in the form of a narrative, and grounded in a specific cultural policy
paradigm’. Thus ‘received wisdom’ can be viewed as a discourse (after Foucault 1971; 1980) that is often
embedded in particular institutional structures that are active on the ground. This may frequently result in the
persistence or tenacity of (a) received wisdom with real practical consequences for (local) people. Leach and
Mearns (1996, 25) do not necessarily view received wisdom as being of a conspired nature, but more as a form
18
African governments and international development institutions. These narratives, often based on de-
contextualised understandings of ecology, political economy and society, and often emanating from
western science, have been dominated by a number of inter-related themes identifying rural farmers
and herders as frequently practising primitive, environmentally degrading and economically
inefficient land, plant and animal husbandry. Natural resource management policy has therefore
been developed by pre- and post-independence governments with pragmatic intent to correct and
improve economically inefficient and environmentally degrading land-use and land husbandry
practices. With the long-term expansion and intensification of a global neo-liberal economy,
governments have been anxious to increase and to capture rural agricultural productivity, by
attempting to integrate local farming and herding systems into national economies and the wider
international economy.
A persistent theme has been that of the dissonance between wild nature (in particular forests and
wildlife) and rural peoples, with the need to protect the former from the latter. The division of
landscapes into those inhabited by people and those exclusively reserved for wildlife and forest has
often been effected to address the perceived dissonance, and to facilitate state-capture of the
economic value of wildlife and forest resources. Contemporary orthodox biological science has been
used to reinforce this process through its contention that biodiversity - or the variety of life - is
greatest in pristine ecosystems untouched by human impact (Guyer and Richards 1996, 1-2).
Narratives of the society-environment relationship with an African context can also be seen as having
often taken the form of a twin problematic of ‘diversity’ and ‘dearth’ (Schroeder 1999, 360). Fears
have been expressed that Africa’s natural wealth - i.e. its diversity - is in crisis and might be lost
(Anderson and Grove 1987, Adams and McShane 1992, Biodiversity Support Programme 1993,
Bonner 1993, Njiforti and Tchamba 1993, IIED 1994, Jarosz 1996, Neuman 1995b & 1996 cited in
Schroeder 1999, 361). Thus stringent (bio)diversity protection and preservation measures have been
justified and achieved through a broadening of scope for interventions by the state and international
agencies (Schroeder 1999, 361). In contrast, ‘dearth’ has been manifested in the ‘twin spectres’ of
drought and famine, and the seemingly intractable problems of inefficient food production and
appalling rates of environmental degradation on urban and rural landscapes (Franke and Chasin
1980, Watts 1983, Peters 1987, Hjort af Ornäs and Salih 1989, Drinkwater 1992, Ghai 1992,
Massaro 1993, Turner 1993 Beinart 1984 &1996, Hoben 1996, Scoones 1996 cited in Schroeder
1999, 361).
2.1.2 New understandings

Contemporary understandings of society-environment relationships within an African context,
particularly over the last 20 years, have grown and been transformed through the multi-disciplinary
scholarship of Africanist (environmental) historians, anthropologists, geographers, ecologists and
of social construction in which individual actors may unwittingly participate and further reinforce the paradigm
through their interactions with others.
19
political economists. The substantial and diverse gains in our understanding of society-environment
relationships within an African context have recently been usefully reviewed by several scholars.
In the introductory chapter to a volume focusing on African savanna environments, Basset and
Crummey (2003) review a number of cross cutting themes focusing on the historical processes and
discourses that have shaped outsider perceptions of - and interactions with - African landscapes and
their peoples. These have been carefully pieced together by African historians (for example,
Anderson 1984; Grove 1987; and McCann 1999) as well as by anthropologists (for example, Tiffen
et al. 1994; Fairhead and Leach 1996; Basset and Zuéli 2000; Brockington 2002). Within this
context, Basset and Crummey highlight the impacts of different ecological understandings on the
management of savanna ecosystems and landscapes, and the recent emergence of alternative
ecological paradigms that have gained growing currency (for example, Homewood and Rodgers
1987; Scholes and Walker 1993; Scoones 1995). The combined work of these scholars has
demonstrated how African modes of landscape function and management have been often
misunderstood by a wide range of pre- and post colonial actors, and in turn how discourses of
environmental degradation have emerged and continue to be promulgated based on these
misunderstandings. Orthodox classification of landscapes into pristine and human-impacted has
become less credible as environmental histories have indicated that long-term relationships between,
for example, farm and forest-savanna in west Africa (e.g. Fairhead and Leach 1996) as well as
montane forest in east Africa (e.g. Schmidt, 1989), have been more complex than initially thought.
Contemporary understandings of customary African land tenure that convincingly challenge received
wisdoms (which continue to retain currency) have been reviewed by Peters (2004) and draw from the
work of scholars such as Channock (1991), Moore (1986; 1996), Basset (1993), Okoth-Ogendo
(1989; 2000), Pinckney and Kimuyu (1994) Goheen and Shipton (1992), Berry (1989; 2002) and
Toulmin and Quan (2000). There is now an extensive literature on herding - for example, that
reviewed by Fratkin et al. (1994) and Little (2003), and farming societies - for example, that of
Netting (1968; 1993) Richards (1985) and Reij et al. (1996), which has provided detailed and
context-rich insights and understandings of local production systems. Lastly, advances in remote-
sensing technology and productive inter-disciplinary collaboration have allowed ecosystem- and
region-wide studies of contemporary landscape use and trajectories of change - for example those
of Lambin et al. (2001) and Homewood et al. (2001). These studies have begun to highlight the
importance of macro-economic factors and land policy as being particularly important drivers in
landscape use and change trajectories.
While affirming these themes, Beinart and McGregor (2003,17) argue, however, that perhaps
previous critiques of ‘colonial science’ have been rather too condemnatory and ideological, with too
little attention paid to the contexts in which particular ideas were initially shaped and thereafter
reproduced. Importantly, Beinart and McGregor acknowledge that some aspects of colonial science
were sensitive to local knowledge, and made substantial gains in scientific scholarship, despite the
fact that such knowledge was not always sensitively and appropriately acted upon.
20
Many scholars and development practitioners now recognise that the perceptions and narratives of
environmental degradation and economic inefficiency leading to corrective and prescriptive policies
have frequently had an unfortunate impact on herders and farmers living in the East African
rangelands. As a result of these and other past policies, for example in Tanzania, Ujamaa2, herders
and farmers have been alienated from their lands, required to grow specific acreages and types of
crops, forced to enter into commodity transactions at times and in conditions not of their choosing,
drawn to produce extractive surpluses in produce and labour, forced to sedentarise, compelled to
de-stock their rangelands of their herds of livestock and have had unfamiliar systems of land tenure
and husbandry imposed on them. In as much as many of the past interventions and policies have been
partly conceived with an intention of improving, one way or another, the livelihoods of rural herders
and farmers, many have had deleterious impacts. Ironically many such policies have contributed to
deteriorating environmental indicators, declining agricultural production, the undermining of the
capability of rural people to effectively manage their environment, increasing levels of inequality in
wealth and the persistence of high levels of rural poverty.
Thus while understandings of society-environment relationships have grown and have been
productively re-examined by a wide range of disciplines as well as through the interlocution of these
different fields, the advances achieved may, as Beinart (2000, 284) argues, now be sufficiently
robust and secure to withstand (further) examination and extension. Beinart reflects with regard to the
study of the colonial (and subsequent3) legacy of environmental degradation narratives:
Arguments rooted in an anti-colonial and sometimes populist or anti-modernist

discourse can present us with an analytical closure, too neat an inversion, which is not
always appropriate in a post-colonial world where the sources of power have
changed. Clearly it is essential to keep issues of equity at the forefront of analysis, to
combat racial assumptions in respect of resource use, and to understand past
relationships between colonial authority and environmental regulation. But it is equally
important that routes of research and analysis are not disguised by the strength of a new
consensus. (italics added)
Peters (2004, 270-1) in reflecting on the study of land and social conflict in Africa proposes that new
theoretical moves have thus far been inhibited by:
‘[S]ticky’ paradigms - analytical frameworks and theoretical premises that have been
highly productive in the past but that, in fast-changing circumstances, are proving to be
blinkers rather than powerful lenses.
In acknowledging the reservations of Beinart (2000) and Peters (2004), this chapter sets out the key
theoretical components which provide a contextual framework for the thesis.
2 Ujamaa, as described later in this thesis, was a massive state-sponsored social experiment in which about 5
million people were resettled into communal Ujamaa villages and made to work in communal agricultural
production.
3 While Beinart specifically reflects upon the colonial legacy at this point in his paper, his insight is arguably
extendable to post-colonial environmental degradation narratives as well.
21
2.2 Recent theoretical advances in the society and environment debate
The emerging society and environment debate has been addressed by a number of different
theoretical approaches, particularly over the last 40 or more years, with emphases on social
anthropology, political economy, ecological anthropology, political ecology, and environmental and
ecological economics. Within the context of contemporary African society-environment debates in
particular, perhaps the most dominant theoretical approaches are those of political economy and
political ecology. A principal concern of these debates is the development of appropriate policy4 and
interventions ultimately targeted towards rural farmers and herders, for developing local institutions
which facilitate improvements to the equitability, socio-ecological sustainability and increased
productivity of their production systems.
Institutions are the focus of attention since they are most thought to shape human behaviour in relation
to resource use (Cleaver, 2000, 364). Key to understanding institutions is the realisation that they are
embedded in complex social relations - relations that continue to evolve and be negotiated within the
context of socially, historically and ecologically located people(s) who shape, and also are shaped
by, a variety of institutions of varying degrees of formality and organisation (Cleaver, 2000: 362)5.
The two theoretical schools of collective action theory and political ecology are discussed in the
following sections.
2.2.1 Collective action: Common pool resource and moral entitlement theories
Perhaps the most immediately prominent and widely received theory within the society and
environment debate is that of collection action theory. Developed in the late 1980s and 1990s (see
below), but with origins in the 1960s (e.g. 0lson 1965) and the late 1970s (e.g. Popkin 1979), collective
action scholarship has now crystallized into two related but very differently focused and articulated
schools (Johnson 2004; Mosse 1997).
Common pool resource theory

Common pool resource (CPR) theory is heavily based on game and organizational theory as well as
conceptual modelling (Bromley 1989; Ostrom 1990 & 1998; Cousins 1993; Keohane and Ostrom
1994). CPR theory is firmly rooted in an epistemology which seeks to construct general and predictive
theories about common property regimes (Johnson 2004, 423). With affinities to the work of Thomas
Hobbes and Adam Smith that premise people as rational self-interested individuals (homo
4 Policy is defined here after Homewood (2004, 128) and taken to mean the de facto guiding principles which
are implicit and evident in official planning, decision-making, and practice, even where these are not articulated
and published as formal policy.
5 Cleaver (2000: 362) draws on the work of: Giddens (1984; 1989) and Long (1992) about the interaction of
agent and structure; Granovetter (1992) in regard to the concept of the embeddedness of economic
transactions in social life, and; (Douglas, 1987) with regard to the role of institutions in shaping individual
perception and action.
22
economicus) (Mosse, 1997, 469), and based within political science, CPR theory counters the accepted
wisdom that communal resource use inherently tends to lead to Hardin's (1968) 'Tragedy of the
Commons'. The 'Tragedy of the Commons' postulates that individually and economically rational
strategies lead to collectively irrational ecological outcomes and to the ultimate degradation of the
communal resource. The proponents of CPR have, through many empirical examples ranging from
recent water rights agreements in the western United States to local level irrigation arrangements in
the Philippines, shown that CPR institutions have been, and can continue to be, effective in their
operation.
Analysis of empirical examples has led to the construction of framework conceptual models (Keohane
and Ostrom 1994; Ostrom 1990) that deconstruct the processes and interactions that tend to occur
during negotiations leading towards collective action regimes. The analyses are concerned with how
the ‘assurance problem’ can be effectively dealt with (Runge 1981; 1984; 1986 cited in Lane 1998).
How can for example, (i) issues of heterogeneity amongst resource users (their preferences,
expectations and capabilities) as well as differences in information flows be minimised; (ii) free-riding
(cheating) be sanctioned against; (iii) institutions evolve in a gradual process towards maturity and
resilience; and (iv) both individual and group costs incurred and benefits derived from entering into
collective action arrangements be incorporated to enable Cost Benefit Analysis modelling.
Thus Common Pool Resource theory has been developed from drawing insights from a compilation of
a large and diverse number of case studies, reflecting the complexity and variability of situation-
specific contexts. However, the theory and methodological approach have been critiqued in a number
of ways.
Agrawal (2001 & 2003) notes that the diversity and number of CPR case studies has led to the
unsystematic generation of a relatively much larger number of variables which are impossible to
analyse carefully. He argues for the adoption of a more systematic approach based on more careful
study design leading to statistically valid cross comparative analyses. Agrawal latterly also
acknowledges the lack of attention paid to power relations and micro-politics within communities in
CPR analyses as called for by Agrawal and Gibson (1999) and Moore (1998 & 1999), and he
further recognises the diachronic nature of these phenomena with regard to the governance of
common property.
The deductive and positivist nature of CPR theory has been tempered by contemporary experiences
of complexities from the field. For example, Campbell et al. (2001) question the ease of realising
Ostrom’s (1994) postulation that, ‘the case-study literature now demonstrates without a doubt that it is
possible for CPR appropriators (i.e. users) to design, operate, monitor and enforce their own
institutional arrangements’. Their experiences from Zimbabwe attest to the high transaction costs in
developing formal CPR institutions especially where there are already well established informal and
socially grounded customary norms and practices that, irrespective of their efficacy and their
suspected decline, incur low transaction costs. Campbell et al. (2001, 596) reflect other scholars’
reservations (e.g. Peters 1994; Mosse 1997; Leach et al. 1997; Goldman 1998; Cleaver 2000; and
23
Toulmin and Quan 2000) that the nature of institutional economics which frames CPR theory lacks
sufficient attention to spatial settings and temporal contexts of history, micro-politics and socio-
economy. Thus, despite the general validity of its key findings, CPR theory is unable to adequately
provide the grounded socio-economic and cultural nuance often required for sufficiently socially
replete and historically contextualised analysis - necessary for understanding people and
environment relations in diverse but often inter-related or nested contexts (e.g. Mosse 1997, 470).
The moral entitlement school

In contrast to Common Pool Resource theory, an alternative school within collective action scholarship
emphasizes the force of tradition (i.e. history), social rights, value systems and moral codes in
facilitating collective resource management (Mosse 1997, 469). Johnson (2004, 415) views the moral
entitlement approach as differing from the CPR school in three ways:
Firstly, socio-economic equality and poverty reduction, as opposed to the efficiency

and health of the commons, constitute major normative concerns. Secondly, rules are
important in so far as they enhance, not restrict, access to the commons. Thirdly, the
entitlement literature tends to favour a structural-historical approach, in which property
rights and relations are contingent upon contextually-specific forces of change.
Thus the entitlement literature is centrally concerned with the problem of inequality, and the ways in
which formal and informal rules reinforce unequal access to common pool resources (Johnson 2004,
415). While the approach of the moral entitlement school arguably provides indispensable
qualitative dimensions to our understanding of collective action, it is open to criticism for lacking a
methodological framework conducive to deductive hypothesis construction and testing as well as
systematic statistical analysis.
Although part of a longer term and much larger set of philosophical deliberations around relativism
and rationalism (e.g. Gandy 1996, 30), there is currently a debate - as recently reviewed by Johnson
(2004) - as to whether these two different approaches, the positivist hypothetical-deductive school
(CPR theory) which leans towards an epistemology emulating the natural sciences, and the historical
and contextual school (representing a moral entitlement paradigm), can be ‘married’. Within the
context of environment - society discourse, some would argue that this marriage is possible and that it
would offer the best of both worlds leading to synergies (e.g. King et al. 1994) such that, ‘their
assumptions, propositions and conclusions are tested both in terms of their logical coherence and
consistency with empirically knowable facts’ (Johnson 2004, 425). While the details of this debate lie
out with the immediate focus of this thesis, the substance of the debate is relevant to the methodology
and approach that is necessarily taken in cross-disciplinary research - such as the study of society and
environment. Thus in acknowledging the substance of the debate, it is possible to turn to differently
framed approaches to the study of society and environment which offer a useful epistemological
alternative.
2.2.2 Political ecology

Political ecology brings together the natural and social sciences in the analysis of the complex
relationships between people and the environment. Political ecology is characterised by a range of
24
different discourses within a contested domain of thought and action (Brown n.d. cited in Escobar
1999, 25), which can most simply be seen as taking two primary forms (Brosius 1997 cited in Escobar
1999, 17) each tending towards opposite ends of a spectrum of interpretation (for example, see
Demeritt’s 2002 review and typology).
The first takes the form of a fusion between political economy and human ecology, in which a series
of differently empowered actors can be seen as contesting each others’ claims to particular resources
- claims which stem from an ‘unproblematic’ ecological base. Thus in this form a dual set of
epistemologies is adopted simultaneously as (ecological) science is used to understand the
environmental impacts of human activities and then social science is used to understand the human
organization of those activities, their dynamics and origins (e.g. Zimmerer 1996, 178).
The second form, informed by post-structuralist social theory, is characterised by a recognition of

‘nature’ (as well as the identities and interests of various agents) as being both contingent and
problematic. Thus Escobar (1999, 3) defines this ‘anti-essentialist’ political ecology as, ‘the study of
the manifold articulations and the cultural mediations through which such articulations are necessarily
established’. ‘Anti-essentialist’ here means that nature6 is differently produced by different groups or
in different historical periods (Escobar 1999, 5). In other words, this latter definition acknowledges
that knowledge and perception of the bio-physical essence of ‘nature’ and people’s interactions with
it are manifold - i.e. that such sets of knowledge, perceptions and interactions with socially constructed
‘nature’ exist, each ‘set’ being contingent on a complex interplay between history and culture, and
perhaps even overlapping in time and/or in space as ‘hybrids’.
This latter definition particularly differentiates political ecology from ‘collection action’ approaches
(as described above) to the study of society and environment, as political ecology seeks to relocate
‘objective’ (natural) science and its role in understanding society-environment relationships within an
empirical and historically nuanced analysis of socio-cultural context. This analysis crucially recognises
that the natural sciences and technology are neither ahistorial nor non-ideological (Escobar 1999, 3),
and that there is a need to incorporate a greater awareness of what the different discourses on
‘nature’ may be ignoring and politically repressing (Soper 1996, 23). Thus it is the case that the often
sophisticated ecological knowledge of farmers and herders has remained largely untapped and
disregarded by ‘science’ because it lacks the imprimatur of scientific objectivity and expertise (Gilson
et al. 2003; also Richards, 1985).
Discourse analysis has become a key tool in post-structuralist political ecology. Gilson et al. (2003,
777) define discourse as, ‘[A] power-laden set of statements about a referential subject in which there
are no coherent, taken-for-granted or innate facts or structures of meaning’. In this regard, discourses
6 The term ‘nature’ is, as with some other widely used terms such as ‘community’, often amorphous in meaning and
has been described by Raymond Williams (1983, 219 in Demeritt 2002, 777) as, ‘...perhaps the most complex
in the [English] language’. Williams (1983) goes on to distinguish three specific, but closely intertwined, meanings
of the word ‘nature’ - see Williams (1983, 219) or more recently Demeritt (2002, 777-778).
25
are seen as sets of knowledge that emerge to serve a power structure and so re-create it. Following
Foucault in using discourse analysis, it is therefore possible to look at the particular interactions
between knowledge and power which accord validity to economic explanations, ecological models
and certain ideas of ‘community’ to the exclusion of others (Mosse 1997, 471).
Thus an important outcome of the political ecology approach - with regard to East African rangelands
- has been the highlighting of the biased use of perhaps increasingly dated rationalist views of
objective ‘factual’ science in justifying and legitimising particular natural resource management
policies, often in the interests of more powerful interest groups (Blaikie 1995, 7). This latter point,
particularly in the context of understanding the social and political organisation of knowledge in
colonial and post colonial Africa (Gilson et al. 2003, 383) and its impacts on society-environment
relationships in Tanzanian rangelands7, has been documented by a growing number of authors - for
example, Brockington (1998), Brockington and Homewood (1996; 2001), Hodgson (2001),
Homewood and Rodgers (1991), Lane (1991; 1996), Maddox et al. (1996), and Neumann (1992;
1998; 2001).
However, important as the developments and gains in political ecology have been in improving our
approaches to studying and understanding ‘society and environment’, there is a growing need to take
into consideration key issues of political economy (e.g. Peters 2004, 280) - such as the impacts of
structural adjustment on Tanzanian herders’ and farmers’ livelihood strategies (e.g. Bryceson 2002,
728). Thus Homewood (2004, 139) states,
To be able to frame the right questions and interpret findings in an appropriate way
researchers need to combine an awareness of political economy and political ecology,
of environmental discourse and narrative, not only with a natural sciences based
understanding, but also with an understanding informed by local perspectives on
environmental processes and causes of change.
This may often entail drawing from disparate disciplinary and theoretical literatures and from
debates in which there may be no single agreed interpretation; at the same time, it also entails not
falling into the traps of over-selectivity, distortion or naiveté associated with disciplinary boundary
crossing (Peters 2004, 280).
The analytical framework on which this thesis is based and which is discussed later in this chapter,
draws from the interdisciplinary approach outlined above of combining political ecology with an
awareness of political economy.
I now move onto consider three key themes that have dominated society and environment relations,
particularly in the semi-arid rangelands of Africa. The first theme focuses on the debate about the
relationship between population and environment, and whether the relationship is relatively
straightforward or whether it is far more complex than previously acknowledged. The former (neo-
7 Tanzania is given particular focus since it is the country in which field work for this thesis was carried out.
26
Malthusian) position has also heavily influenced popular thinking about the relationship between
environment and conflict – the second theme. In recent years, political ecologists have successfully
challenged these popular and mainstream understandings of environment and conflict, leading to a
much more nuanced and grounded analyses of the relationship between violence, environment and
power. Finally neo-Malthusian understandings of environment and population have also often had an
underlying and strong influence on approaches to African land economy - the third key theme. The
widely and long held view that African customary systems of land tenure should be replaced by
individualised and titled land tenure systems has been successfully challenged, although as scholars
have recently pointed out, the renaissance in approach perhaps needs now to be tempered by a
more critical analysis.
2.3 Contending with new paradigms of complexity in environment and society
A recurring and central controversy in the society and environment debate, that has in turn long
underpinned the formulation of natural resource management approaches and policies - not least in
the East African rangelands - is that between neo-Malthusian (after Malthus, 1798 & 1803 rpt 1960)
and alternative understandings (see below) of the population - environment nexus. While neo-
Malthusian paradigms have heavily informed policy and law, impacting on and often controlling local
people’s resource use patterns, as well as heavily influencing natural resource management
interventions, alternative understandings, including those of ‘post-normal science8’, have only recently
begun to filter through into the mainstream.
In substantial part, neo-Malthusian and science-based understandings of society and environment

have been predicated on, or much associated with, a central assumption - ‘the balance of nature’,
that has been increasingly countered by an alternative paradigm, the ‘flux of nature’ (Gilson et al.
2003, 380). The ‘balance of nature’ paradigm, with a history reaching back to Linnaeus in the
eighteenth century and before, emphasises that nature linearly tends towards single equilibrium points
and stability (equilibrium theory), stable states which human action may often disturb or destroy. In
contrast the ‘flux of nature’ paradigm comprises a much more complex construct (Gilson et al. 2003,
381). The ‘flux of nature’ paradigm, or disequilibrium theory, contends that nature - or an ecosystem -
does not necessarily tend towards a single stable equilibrium point, although stability is not
precluded. Instead ecosystems are viewed as potentially having multiple potential stable states
modulated by processes that generate spatial and temporal heterogeneity, including interactions
between organisms (biotic instability), environmental stochasticity and disturbance. Disturbance in
disequilibrium theory, in contrast to equilibrium theory, is viewed as being the norm rather than the
exception (Gilson et al. 2003, 381). Often ecosystem stability may be scale dependent - i.e. sub-
components within an ecosystem may be in different degrees of flux, but the ecosystem may remain in
8 After Funtowicz and Ravetz (1992), who characterise particular areas of science - not least those of ecology
and climate - as increasingly encountering greater levels of complexity and uncertainty than was previously
‘normal’, as scientific methodology and understanding continues to develop (Shackley et al. 1996, 204).
27
a stable state until such time as the variability of one or more sub-components - or patch dynamics -
act in such a way as to drive the larger scale ecosystem into a different state.
Thus, in large part predicated on the logic of the ‘balance of nature’ assumption, neo-Malthusian
narratives emphasise a relatively straight-forward relationship between population growth and
environmental degradation (e.g. Ness et al. 1993 and Pimenthal et al. 1994 cited in Agrawal and
Yadama 1997, 439). Contrastingly, and as Agrawal and Yadama (1997, 440) note, a growing
number of scholars contend that the relationship is anything but straight forward (e.g. Blaikie and
Brookfield 1987; Netting 1993; Whitmore 1990). Thus the hegemony of neo-Malthusian paradigms
of environmental degradation, not least perhaps the controversial desertification debate (e.g. UNEP
1992)9, has been increasingly countered by a series of empirical studies (for example, Fairhead and
Leach 1996; Linblade et al. 1996; Tiffen et al. 1994; Basset and Zuéli 2000) that have drawn
divergent conclusions to prevailing received wisdoms about the relationship between people and
environment in African rangelands. The contemporary challenge to neo-Malthusian paradigms has
particularly benefited from the growing sophistication of disequilibrium theory as well as
contemporary developments in our understanding of the varied environmental history and socio-
economy of African rangelands and forests. Thus many rangeland ecosystems are now viewed as
being the product of continual disturbance through patchy and unpredictable rainfall, fire, grazing,
browsing and physical disturbance (e.g. Dublin 1995, Behnke and Scoones 1993, Ellis and Swift
1988, Homewood and Brockington 1999 cited in Homewood 2004, 129). In such systems habitat
disturbance is not necessarily detrimental to species survival, as species diversity and survival is based
more on their ability to disperse, colonise and persist in a patchy and unpredictably fluctuating
environment (Davis et al. 1994, Stattersfield et al. 1998, Homewood and Brockington 1999, Laris
2002a&b, cited in Homewood 2004, 129). An implication that follows is that (pastoral) mobility and
opportunistic management offer efficient strategies for coping with arid land ecology - as epitomised
by the customary practice of indigenous pastoral systems (Sandford 1983, Behnke and Scoones
1993, Ellis and Swift 1988, Niamir-Fuller 1999, Turner 1998a,b, 1999a, Sullivan 1999b cited in
Homewood 2004, 133).
These and other studies empirically demonstrate that the dynamics underlying environmental change
can be more complex than a simple neo-Malthusian thesis, and that underlying socio-economic
conditions and other factors such as induced migration and land tenure changes precipitated by
government policies may be key to better understanding the environment-population nexus (e.g.
Lambin et al. 2001). For example, deforestation may occur in localised areas, but on a wider scale,
overall trends towards a more forested landscape can occur (Leach and Fairhead 1996). In
agreement with Boserup’s (1965) thesis, instead of inexorable soil erosion and decline in soil fertility
associated with population expansion, rural farmers may, under the right policy conditions, invest
more in soil and tree conservation practices with the aid of urban remittances as land availability
9 Mortimore (1998) usefully reviews the historical development of the desertification and environmental
degradation debate in some detail.
28
declines and as agricultural intensification rises (Tiffen et al. 1994). These studies, amongst others,
have been landmarks in developing better understandings of the socio-economic and ecological
dynamics of agro-ecological systems.
Yet, it may also arguably be the case that, although the recent development of empirically-based
scholarship questioning widely accepted neo-Malthusian perceptions about environment-people
relationships has done much to advance a set of more nuanced understandings about their
complexity, it is necessary that some of these studies be placed within a wider perspective. In
particular, there has perhaps been a growing temptation to over-extrapolate some of these studies’
conclusions outwith their context. This is especially the case when ground-breaking challenges to
received wisdoms are themselves challenged, or considerably modified, by further field research. For
example, while Tiffen et al. (1994) undoubtedly found a clear relationship between increasing
population, maintained soil fertility and an afforested landscape, a subsequent study in the same
area10 revealed a more complex on going process (Murton 1999). This latter study found that while
those farmers able to invest off-farm remittances in land and land-improvements were able to
increase their yields, other poorer farmers were instead forced through circumstance to invest off-
farm (labour) cash incomes in food security and were thus entrapped within a cycle of diminishing
yields and decreasing investment in land improvement. A growing population was found to be
exacerbating this latter cycle as less land was available and more people were in search of non-
farm employment. Whereas an expected constraint in the relationship between people and
environment had been successfully overcome (in terms of agricultural intensification and its associated
soil status and yield improvements), continued population growth had effectively begun to reverse the
relationship for the majority of poor farmers unable to benefit from the nexus between rural and
urban economies (Murton 1999).
This example demonstrates that, as Leach and Mearns (1996, 4) state, contradictory evidence from
an individual case study cannot entirely refute an orthodoxy. While the questioning of particular
received wisdoms has cast doubt over their wider applicability (Leach and Mearns, 1996, 4), a
balance may need to be struck as further inquiry and critical examination reveals that, although the
paradigms through which received wisdoms are constructed may often be extensively flawed, the
concept of an ultimately finite, albeit far more complex, relationship between people and
environment may be sound. The challenge therefore lies in developing more sophisticated
understandings of the drivers and processes of change underlying people-environment relationships
that, while recognising the limits of generalisation between individual, bounded contexts, afford
comparative insights for wider theoretical advances.
I now move on to consider how popular understandings of environment and conflict have been heavily
influenced by neo-Malthusian thinking and reproduced through the application of deterministic
science. Since conflict is a re-occurring and central theme of this thesis, I provide an overview of the
10 These studies were carried out in Machakos in central Kenya, East Africa.
29
contemporary debate about conflict and environment, and its role in better understanding society and
environment relations.
2.4 Understanding conflict: environment, power and violence
Conflict has long been a recurrent feature of socio-political relations over access to resources in the
semi-arid rangelands of sub-Saharan Africa – most notably between herders and farmers. Current
understandings of how and why conflict more generally occurs can be divided into three schools of
thought:
1. ‘Environmental security’ – part of a wider set of theories which recognise conflict as being
primarily driven by environmental scarcity leading to exacerbating political-economic and
socio-economic factors;
2. ‘Collective Action Theory’ – setting out causality as being heavily associated with institutional
failure manifested by poorly governed and ill-defined resources (see Sections 2.2.1 &
2.2.2);
3. ‘Political Ecology’ – setting out causality as being grounded in wider unequal socio-political
relationships and being primarily driven by interests (often elites) defining and affecting
control over (a) resource(s) (see Section 2.2.3).
The ‘environmental security’ paradigm is strongly underpinned by a neo-Malthusian understanding of

the relationship between environment and population11. Emerging during the 1990s from two
separate research programmes led respectively by Professor Homer-Dixon (1991, 1994, 1995,
1998, 1999) and Professor Gunther Baechler (1996, 1998) (Peluso and Watts 2001, 12), they
generated what Peluso and Watts (2001, 15) view as ‘fundamentally similar’ models. Thus the
‘environmental security’ school lays out the rationale that rising population and increasing resource
scarcity will lead to increasingly severe levels conflict around the world. A brief deconstruction and
critique of the ‘environment security’ paradigm (focussing on the work of Homer-Dixon) is useful for
understanding why alternative approaches – and particularly that of political ecology – may better
locate the place of environment in relation to the occurrence of conflict.
2.4.1 A summary critique of the ‘environmental security’ model

The emergence of the ‘environmental security’ school of environment-conflict analysis coincided with
the aftermath of the end of the cold-war and the belief among its proponents and supporters that
growing resource degradation would contribute to a proliferation of ‘small wars’ (Peluso and Watts
2001, 7) or ‘green wars’ (Twose 1991, 1 cited in Fairhead 2001, 213). While there may not
necessarily be any direct aetiological relationship between the ‘environmental security’ school and
African rangeland policies, the underpinnings of each have a great deal in common. Both are driven
11 Population growth underpins one of the three components of the Environmental Security model, despite
Homer-Dixon’s claims (1999, 28) that he is not a neo-Malthusian.
30
by neo-Malthusian understandings of environment and population, which as I established in Section
2.3, are particularly inappropriate in an African semi-arid rangeland context.
In summary, as reviewed by Hartmann (2001, 40-42), the model asserts that:
1. In certain situations renewable resources can cause civil conflict and instability. Conflict and
instability can cause social effects (such as poverty, migrations and weak institutions) that are
misconstrued as being the immediate causes;
2. ‘Increased demand’ for environmental resources is driven principally by population growth;
3. ‘Degradation’ of environmental resources induces powerful groups to tighten their grip on
them in a process termed ‘resource capture’;
4. ‘Resource capture’ leads to ‘unequal resource distribution’ and thereby intensifies resource
scarcity for poorer and weaker groups;
5. The resulting ‘environmental scarcity’ can force migration of the poorest groups to
ecologically vulnerable areas. The pressure of their numbers and their lack of knowledge
and capital can cause severe environmental scarcity and chronic poverty, a process termed
as ‘ecological marginalisation’ (Homer-Dixon and Blitt 1998, 225 cited in Hartmann 2001,
41).
6. Environmental conflict can be avoided through societies adapting to scarcities through more
efficient resource use or acquiring resources instead through international markets.
Adaptation depends on the sufficient supply of ‘social and technological ingenuity’ to
produce solutions to scarcity. In poor countries, the prospects for ingenious adaptation are
low, with potential for further impoverishment and migration.
7. Environmental scarcity also leads to the undermining and threatening of the state and society,
as increased competition for resources leads to social segmentation and reduced social trust,
as well as an escalation of challenges to the authority of the state.
8. Finally, by contributing to migrations, economic decline (in poor countries), social
segmentation and weakened states, environmental scarcity helps lead to violent ‘ethnic
conflicts, insurgencies and coups d’états’ (Home-Dixon and Blitt, 1998, 227 cited in Hartmann
2001, 43).
Perhaps the biggest weakness of the ‘environmental security’ model, aside from its somewhat
simplistic neo-Malthusian leanings, is that it fatally conflates distinct and very different processes into
the overarching term of ‘environmental scarcity’. Thus, ‘increased demand for resources’,
‘degradation’ and ‘unequal resource distribution’ can each or together cause ‘environmental scarcity’
(Hartmann 2001, 43). The conflating of three very distinct and different processes into one concept is
a shortcoming that Fairhead (2001, 217) understandably laments as being ‘tantamount to analytical
obfuscation’. Fairhead then proceeds to show just how flawed the conflation of distinct processes into
a single term ‘scarcity’ can be. He cites examples (e.g. Tiffen, Mortimore and Gichuki 1994, Murton
1997 and Lindblane, Carswell and Tumuhairwe 1997) where, instead of conflict, scarcity has led to
improvement in the quality of the environment – particularly in densely settled agricultural areas
where land is extremely scarce, land values continue to rise concomitant with the increasing
31
productivity of land associated with investments in ecological improvements (Fairhead 2001, 217-
219). Following on from these examples, Fairhead (2001, 219) appropriately asks, ‘Why conflate
into one concept the very relationships that are interesting to pull apart and explore’? For example,
unequal resource distribution has very little to do with the environment per se and everything to do
with issues of power, different understandings of environment and different forms of violence used to
enforce inequality. Inequality in this context is often driven by resource wealth, rather than resource
scarcity, and the interests of elites enforcing their control and access to those resources. Equally,
Fairhead (2001, 220) argues that the case for the consistent presence of a relationship between
resource degradation and increasing conflict has never been made convincingly. Moreover, the
conflation of degradation with environmental scarcity is misleading – resources can be scarce but not
degraded, and they can be widespread but degraded. In as much as the environment is often
associated with the occurrence of conflict, contrary to Homer-Dixon’s assertion, the former is
frequently not an immediate or fundamental cause of the latter. Rather than enclosed and inwardly
collapsing environmental systems – as intimated by the ‘environmental security’ model, often the
dynamics of access to and conflict over resources are driven by powerful external forces and political
economic processes or power relations.
This latter observation in particular underlines the need to look for alternative approaches to
examining the relationship between environment and conflict. The application of common pool
resource theory (see Section 2.2.1) in the analysis of environment and conflict has similar limitations to
the ‘environmental security’ school (Turner 2004, 865). In particular, and as discussed in Section 2.2.1,
CPR theory embraces an individualistic rational choice analysis of group behaviour (Ensminger 1992,
Mccay & Acheson 1987, Oakerson 1992 and Ostrom 1990 cited in Turner 2004, 865). This
postulates that the behaviour of individuals or groups if unconstrained will likely lead to competition-
driven resource over-use – i.e. socially-produced resource scarcity. The neo-Malthusian underpinning
of this analysis can easily lead to over-simplified interpretations of population-driven competition,
conflict, environmental degradation and resource scarcity (Goldman 1998, Peters 1987, Turner
1999a cited in Turner 2004, 865). The adoption of a CPR approach is therefore likely to lead to
shortcomings similar to those of the ‘environmental security’ school.
2.4.2 Political ecology: power, environmental discourse and conflict

Political ecology (see Section 2.2.3) may likely be the most flexible and powerful analytical
approach to understanding the relationship between environment and conflict.
2.5 Land and institutions: social embeddedness, inequality and conflict
Changing perceptions and understandings of customary land tenure and land management have long
been, and continue to be, a key element of the society-environment debate within an African
rangeland context. These changing understandings have led to the development of a dualism
between ‘modern’ and ‘customary’ forms of property relations and land tenure (Woodhouse et al.
2000, 18). The dualism can be traced through a historical kaleidoscope of four recent periods in
32
modern African history: Colonial establishment and consolidation (1880s - 1930s); ‘late colonialism’
with its developmental thrust (1940s - 1950s); independence and the ‘developmental state’ (1960s -
1970s) and; the era of structural adjustment (1980s - present) (Woodhouse et al. 2000, 2). Given
the frequently poor performance of previous land policies based on ‘modernisation’ (Toulmin and
Quan 2000, 3), the debate is particularly significant and relevant as a number of African countries
(e.g. Tanzania, Kenya, Uganda, Malawi, South Africa and Mozambique) are currently pursuing
programmes of land reform - albeit at different stages of development.
2.5.1 The evolutionary model of land tenure

Thus for the early colonial period, colonial authorities are now viewed (e.g. Ranger 1983) as having
played an important part in the re-creation of customary African land tenure through their
interpretation, re-construction and integration of African land tenure systems of the day into their
administrative systems (Colson 1971; Bassett 1993; McAuslan 2000; Okoth-Ogendo 2000; Wily
1988). This process was an important part of moulding and subsuming perceived and constructed
customary law and power structures into the functioning of colonial administrations and economies.
In the latter part of the colonial period, and as underpinned by the influential report in 1955 of the
East African Royal Commission (e.g. Shivji 1998, 5; Wily 1988, 31)12, colonial administrations
increasingly came to view customary - particularly communal - tenure as an obstruction to productivity
and progress (Bassett 1993, 6). During this period, an evolutionary theory of land rights (ETLR) was
to gain strong favour among policy makers. The approach advocated the gradual replacement of
inclusive forms of customary tenure by individual title and registration, predicated on the assumption
that freehold tenure 13 (or its closest legal equivalent) offered the most propitious conditions for
agricultural investments and productivity (Bassett 1993, 6; Peters 1994, 273; Platteau 2000, 52).
Thus as land scarcity increased, people were expected to demand greater tenure security -
articulated through the emergence of property rights which would evolve towards greater measures
of individualisation and formalisation (Platteau 2000, 52). The approach also held that customary
forms of land tenure were inefficient, did not provide sufficient security of ownership and in turn could
not facilitate the use of land as collateral for credit and land investment. Moreover at the time it was
thought that as human population grew, and with increasing agricultural commercialisation, there
would, in any case, be an inherent tendency towards the individualisation of rights (e.g. Basset 1993,
13).
During the early independence period, newly formed African governments - both capitalist and
socialist - adopted the evolutionary land tenure model for different ideological reasons and
12 Key findings of the East African Royal Commission were supported by the colonial Tanganyikan government
in a policy paper ‘Review of Land Policy’ in 1958, which advocated a highly individualised form of freehold
tenure and the transformation of customary land tenure to such a system (Wily 1988, 74).
13 Although reservations were variably expressed within colonial administrations, particularly in eastern and
southern Africa, that granting freehold status to local farmers would result in the destruction of land if farmers’
agricultural practices went unchecked (Basset 1993, 9)
33
assumptions (Bassett 1993, 11). However an overriding ‘broad-spectrum’ approach was common to
both - that a wide range of agrarian problems could be tackled through land tenure reform (Okoth-
Ogendo 1993). Indeed today, this still remains a widespread and prevailing perception among
governments, development organisations and institutions.
Individual titling and registration (ITR) schemes were promulgated as the best approach supporting
the evolutionary model of land tenure. ITR was viewed as a strong complement to the agrarian
reform programmes during the oil price crisis of the 1970s and into the structural adjustment era of
the 1980s - particularly in post-independence African countries with capitalist economies - such as
Kenya. During this period, African national economies began to increasingly suffer from declining
terms of trade due to increased world oil prices and a general fall in primary agricultural commodity
prices. Tanzania followed a very different path in that it embarked on the enforced communalisation
of land predicated on socialist theory and a misunderstanding14 of the nature of customary land
tenure systems (Wily 1988, 82-83). Agrarian reform therefore became a centre-piece of
governments’ attempts to counter the declining status of national economies and an increasing debt
crisis. Agrarian reform was promoted through encouraging increased production of primary
agricultural commodities (in both socialist and capitalist economies) and the development of land-
markets (in capitalist economies), in order to achieve greater levels of export-led economic growth.
ITR, as a central component of agrarian reform, was viewed by many capitalist governments and key
donor institutions, particularly the World Bank, as a key strategy to promote farmer-based
investment in land improvement, access to credit and as an incentive for increasing agricultural
productivity, particularly through the production of cash crops (see also Quan 2000, 34; Platteau
2000, 52-56).
Yet a growing body of evidence began to be gathered by scholars in the late 1980s and early
1990s (e.g. Bruce 1986; Green 1987; Barrow and Roth 1990; Okoth-Ogendo 1989 & 2000; Migot-
Adholla et al. 1994) that the ITR schemes of the 1970s and 1980s had failed to realise their stated
aims (Peters 2004, 274; Platteau 2000, 35; Quan 2000, 66). For example Barrows and Roth (1990,
290) conclude:
On balance, there is little evidence to support the hypothesis that registration, through
increased tenure security, has increased investment in agriculture, or that …farmers are
willing to risk their land for credit.
Scholars now compellingly argue that ITR programmes have frequently exacerbated conflicts and
patterns of unequal access to land based on gender, age, ethnicity and class in farming systems
(Okoth-Ogendo 1976; Coldham 1978; Pala 1980; Davidson 1988; Shipton 1988; Haugerud 1989;
Attwood 1990; Shipton and Goheen, 1992, 316; Shipton 1994, 364-5; Besteman 1994; 1996; cited
in Peters 2004, 274-5; see also Plateau 2000, 56-62 for an economistic critique). Amanor (2001, 5)
14 This misunderstanding, on the part of Julius Nyerere (the first president of Tanzania), likely arose from a
colonial fallacy that posited African systems of customary land tenure were intrinsically communal and that
individual rights and ownership of land effectively did not exist (Wily 1988, 81; see also Nyerere 1968, 7).
34
provides an illuminating perspective of some of the assumptions implicit in the evolutionary model of
land tenure upon which ITR is predicated:
The problems of defining families, households, units of production, consumption and

reproduction, one parent families and extended networks of kin are not considered.
The farm unit is also not defined and it is presumed that the family farm consists of one
contiguous unit which is passed down from parents to children (probably the assumption
is the father) to children (probably thought of as sons). The concept of a farmer
working on several plots that were acquired in different ways from different people
does not feature in this conception of farm.
A similar literature details the adverse and sometimes disastrous impacts of ITR and range
privatisation in herding systems and range management in Africa (Galaty et al. 1981; Horowitz,
1986; Baxter and Hogg 1990; Behnke et al. 1993; Peters 1994; cited in Peters 2004, 275). Range
privatisation has generally lead to the break-down in range management systems that depend on
larger scale seasonal movements of herds to efficiently utilise the temporal and spatial variability in
productivity in semi-arid rangelands for grazing. But even more critically, privatisation of the range -
for example, through group ranches - has tended to allow elites to convert the range to other uses
(for example, commercial agriculture) often excluding previous and poorer range-users, who may
then become landless and forced into poverty.
In response to the growing criticism of simplistic evolutionary tenure models and evidence of their
adverse and poor performance, property rights analysts (not least at the World Bank) have
readjusted their position such that ITR is now viewed as one of several options for land tenure reform,
a process in which customary forms of tenure may also be retained (e.g. McAuslan, 1998). But strong
suspicion remains that the underlying premises of land reform tenure remain unchanged: that
individual land ownership and the emergence of a land market are seen as the ultimate objective
and as part of the continuing market liberalisation process (Amanor 2001, 8; Peters, 2004, 277;
Shivji 1998, 111).
2.5.2 Communitarian understandings of land tenure

Much of the critique and pressure for re-examining the aptness and validity of the evolutionary land
tenure model has been generated by what Amanor (2001, 9) terms the ‘communitarian’ approach to
land tenure, that draws on anthropological, cultural studies and customary law approaches.
Pioneered and initially developed by anthropologists in the 1940s and 1950s (e.g. Gulliver 1961;
Biebyuck 1963; and Gluckman 1969), the communitarian school posits that customary land tenure
systems are flexible and adaptive, and are embedded in a complex and continually changing matrix
of social, cultural and political relations and meanings (e.g. Shipton and Goheen 1992; Berry 1993;
Shipton 1994; Peters 2004). Thus Toulmin and Quan (2000, 12) state:
Increasingly people have come to recognise that there are considerable merits in
customary systems for land rights management systems since they provide a relatively
secure means for those who are members of the community, at a lower cost than state
run administrative structures. Many arrangements exist within customary systems which
provide flexibility and movement of land between users, through sharecropping,
tenancy, short and long term loans. Also customary tenure systems tend to consider the
needs of poor members and prevent the alienation of land from the group as a whole.
35
While this school provides a far more nuanced approach to understanding land-tenure, it has perhaps
failed, at least in the past, adequately to address the less egalitarian and equitable processes of
social differentiation (c.f. Toulmin and Quan 2001, 12 as above) that may increasingly occur in
contemporary customary tenure systems (Amanor 2001, 11-12). Recent work has shown that national
and local elites often capture the benefits of decentralisation processes aimed at empowering local
and ‘traditional’ leaders (e.g. Carney and Farrington 1998; Ribot 2000; Woodhouse et al. 2000
cited in Peters 2004, 277). Thus while the ‘communitarian’ school has achieved much in successfully
providing alternative and increasingly powerful counter-understandings to the evolutionary land
model, Peters presses for the need to go on beyond current formulations of the social embeddedness
of land. She argues for the need to ask precise questions about the type of social and political
relations in which land is situated, particularly with reference to relations of inequality - class,
ethnicity, gender and age (Peters 2004, 278).
There are two immediate reasons why this need is apparent: the first lies in the point that it is rarely
the case that all (local) people are able equitably to influence and engage in debates and processes
of how and what constitutes the norms of contemporary customary access to land and resources
(Amanor 1999; Peters 2004). Often these debates may be variably over-shadowed by larger
political processes and underlying macro-economic trends such as commoditisation (Bernstein and
Woodhouse 2000, 222). The second point lies in the issue of competition and conflict over access to
land and resources that has long been a part of African society - environment relations. While some
research remains ambivalent about trends in the incidence of land-use conflict (e.g. Hussein 1998),
other research shows that competition and conflict over land is increasing over much of sub-Saharan
Africa (reviewed by Berry 2002; and Peters 2004), and not least in Tanzania (Odgaard 2002), in
part driven by processes of land alienation and privatisation (e.g. Lane 1996; Shivji 1998;
Brockington 2002). As competition for land and resources intensifies, Berry (2002, 640) presses for:
… the importance of situating land (and resource) struggles in specific historical

contexts, taking account of the way multiple interests and categories of people come
into play, and impinge on one another, as people seek to acquire, defend, and
exercise claims on land.
Thus by deconstructing contemporary processes of socio-economic differentiation and inequality in

customary land and resource tenure systems, and through increasing our understanding of the
complexity of contemporary yet historically rooted land conflicts in Africa - which I set out to
accomplish in this thesis - more grounded, sophisticated and representative understandings of these
processes can be developed.
2.6 Mainstreaming new understandings of society and environment in policy
The extended debate over appropriate approaches to land tenure in Africa demonstrates how
alternative understandings generated by extensive research findings may take a long period of time
(in this case well over 20 years) effectively to challenge established policies and the knowledge upon
which they are based. I briefly review why research generating new yet alternative and apparently
unorthodox findings may often not lead to the improved policy outcomes intended for African farmers
36
and herders and the environment. In this regard, Leach and Mearns (1996, 28) argue that
fundamental changes in the relationship between research and development policy-making are
required. This is particularly the case when research increasingly identifies multiple, and potentially
equally valid, complex and/or uncertain outcomes as existing side by side - whether based on
science or on alternative knowledge claims or on both. This leads to two immediate challenges. The
first is that complex plural research outcomes are susceptible to criticism and the disadvantage that
they may be too complicated for the derivation of clear, straight-forward and compelling policy
advocacy (Leach and Mearns 1996, 32). The second is that plural research outcomes may often lie
outside the mainstream and may be contrary to embedded and dominant bodies of knowledge.
Before discussing how these two challenges might be addressed, it is helpful to take a step back and
briefly reflect on the complexity of policy-making processes and the false consensus that may often
underpin established policy (e.g. Cleaver and Franks 2003a cited in Walsh 200415). In this regard, it
is worth quoting from the introduction of a paper by Brock et al. (2001) which explores the dynamics
of the making and shaping of poverty policy. The paper takes as its starting point a critique of linear
versions of policy-making, highlighting the complex interplay of power, knowledge and agency in
poverty policy processes. Thus Brock et al. (2001, iii) view the policy process as involving a complex
configuration of interests between a range of differently positioned actors, whose agency matters,
but whose interactions are shaped by power relations (and structures). I argue here that, within the
context of the following quotation, ‘poverty’ is interchangeable with ‘society and environment’.
Policy… is not shaped simply on the basis of ‘good’ research or information, nor does
it emerge simply from bargaining amongst actors on clearly defined options and
choices. Rather, it is a more complex process through which particular versions of
poverty come to frame what counts as knowledge and whose voices count in policy
deliberations in particular political and institutional contexts. Making sense of
participation in poverty policy processes, then, requires an analysis of the ways in
which power and knowledge define spaces for engagement, privileging certain voices
and versions and excluding others. It also requires an understanding of how particular
ways of thinking about poverty have gained ascendancy, coming to determine the
frame through which poverty is defined, measured and tackled. To do so calls for an
historical perspective, one that situates contemporary poverty policy with regard to
antecedent visions and versions (Brock et al. 2001, iii).
Thus Keeley and Scoones (1999, 32) argue that policy processes need to be understood in discursive
and political context and that the power of political interests and embedded patterns of knowledge
are significant constraints on any policy process.
Returning to look at how to approach the dual policy-advocacy challenge in research of plurality and
unconventionality, Keeley and Scoones (1999, 31) identify two possible modes of engagement. The
first approach is more confrontational and sets out to advocate marginalised interests and excluded
forms of knowledge to counter established policy (Keeley and Scoones 1999, 29). Thus ‘counter-
narratives’ (Roe 1991; 1995) are developed to encompass and integrate the excluded knowledge
and problem definition claims of a wider range of (marginalised) interests in order to counter
15 I acknowledge Dr Walsh for having drawn my attention to this and following papers.
37
established policy paradigms - such as those of prediction and control (Leach and Mearns 1996, 32;
Shackley et al. 1996, 221; Keeley and Scoones 2001, 30). The second approach is more
participatory and consensual (Keeley and Scoones 2001, 31), although it is reliant on the policy
development process recognising the contingency of different knowledge claims and placing more
emphasis on institutions that promote argumentation and deliberation (Keeley and Scoones 2001,
31). The quality of this latter approach is also contingent on the nature of ‘participation’ and
'consensus’ - for example who is included and who is excluded, and on how the ‘policy spaces’ in
which debate occurs are constituted. Thus ‘policy spaces’ and participation may be ‘created from
above’ by powerful institutions and actors, or they can be more autonomous, created ‘from below’
through independent forms of action (Brock et al. 2001, 1).
Nevertheless, increasing recognition of the existence and expression of a broad range of knowledge
and problem definition claims provides a growing opportunity for the ‘democratisation of expertise’
(Funtowicz and Ravetz 1992 cited in Leach and Mearns 1996, 31) such that policy outcomes are
negotiated among an extended range of ‘non-traditional’ actors (Leach and Mearns 1996, 31). The
challenge remains that, whereas ideas move on, institutions and organisations often stand still
(Chambers 1993 cited in Scoones 1996, 53), attached to universal epistemologies and simple ideas
with powerful slogans, even when they are patently inaccurate or inappropriate (Swift 1996, 85;
Shackley et al. 1996, 221).
2.7 The study approach, analytical framework and methodology
Thus far I have reviewed recent and contemporary developments in research on society - environment
relations in the African semi-arid rangelands and discussed and chosen one of two major
methodological approaches that I consider important for better understanding these relations. As a
basis for the presentation and discussion of my research questions (below), I have reviewed two key
themes that I see as most important (with regard to this thesis) in the debate over society and
environment in the African rangelands. And finally, I have provided a brief review of why a growing
body of alternative knowledge about society and environment relations in the African rangelands is
only now beginning to filter through into policy. I now turn to introducing my research questions, the
analytical framework of the thesis and its methodology.
2.7.1 The study context

The approach to this study has been influenced by two main considerations: The first consideration,
following on from the discussion in the preceding sections, is an attempt to contribute to Peter’s (2004,
270) call, for further advancing our understanding of land-use relations and land-based conflict in
Africa. Specifically, Peters urges that scholars strive to uncover who the ‘winners’ and ‘losers’ are in
their studies of contemporary African land relations. She also argues for the need to identify and
include additional important situations and processes, such as commodity production, livelihood
diversification, and the political economy of the state, that might limit or end negotiation and
flexibility for certain social groups and categories.
38
The second consideration is a need to better understand the underlying causes and processes leading
to growing competition and continuing outbreaks of conflict over land and landed resources among
farming and herding groups in Tanzania, briefly outlined in the introduction, and further described in
Chapter Three. Thus the study sets out to investigate the current livelihood trajectories of farmers and
herders in the Idodi rangelands and to consider some of the key processes underlying people’s land-
use relations and outcomes, which are likely to reflect similar processes of competition and conflict for
land in other dryland-wetlands in Tanzania.
2.7.2 The research questions and analytical framework
Research questions
This thesis is centrally concerned with the political ecology of access to rangeland resources. There is
currently strong support for a ‘new’ evolutionary approach to land tenure law and reform which
posits that non-interference in local - or customary - land tenure systems will allow adaptable and
equitable outcomes (Peters 2004, 277; see also Platteau 2000, 72). This approach is predicated on
experience which has shown that direct state intervention in land matters is best minimised because in
the past it has lead to increased insecurity for farmers and herders (Platteau 2000, 71; see Section
2.4). Instead, what is required is a pragmatic and gradualist approach that reinstitutionalises
indigenous land tenure, promotes the adaptability of its existing arrangements, avoids a regimented
tenure model and relies as much as possible on informal procedures at local level (Bruce 1986, 64-
68; Atwood 1990, 667; Migot-Adholla et al. 1991, 170-173 cited in Plateau 2000, 72). As Platteau
(2000, 72) reflects, there are strong merits to this approach.
However, as Peters (2000, 278) argues, it is also increasingly clear that despite its strengths, the new
evolutionary approach to land tenure reform (as, for example, as partially adopted by the new
Tanzanian Village Land Act - see Chapter Three) is not a panacea. If anything, the approach could
be seen as increasingly at odds with the proliferating evidence on competitive and conflictual
relations over land (Peters 2000, 278).
In this thesis, I set out to contribute to a better understanding of Peters’ central concern about
increasingly competitive and conflictual relations over land that, albeit for a specific case study, hold
wider and comparative significance for similar situations across eastern and southern Africa. Thus the
central questions of the thesis - in terms of contributing to the wider debate about inequality and
social conflict over land in Africa - are as follows:
1. What are the major wider factors that have influenced increasing socio-economic differentiation
and unequal access to land and landed resources at local level?
2. When and how does local social flexibility and negotiation over land lead to inclusion or exclusion
at a local level between different resource-users? How does this take place?
3. Scholarship has shown that past and continuing efforts to reinvent, subsume and/or extinguish
customary law within national law throughout sub-Saharan have contributed to growing social
39
differentiation and landlessness. In this context, to what extent does an alternative approach
recognising the legitimacy of customary land tenure practices reinforce the growing trend of social
differentiation and the increasing numbers of landless and resource poor observed across Africa?
Under what circumstances might such an approach limit or reverse the trend?
This thesis sets out not to question the new evolutionary approach to land reform and tenure, but to
examine how it might be better facilitated and supported at local level in relation to past land
reform failures. In relation to my chosen case study of the Idodi rangelands, I document the growing
social differentiation, competition and conflict over access to land through a livelihoods analysis. I
investigate to what extent the increasing competition over access to land and landed resources is a
product of local practices as compared to state interventions and policies. In identifying the key
factors underlying the major outcomes and trends in people’s land use and livelihoods, I investigate
some aspects of current land use relations, focussing on interactions between herders and farmers.
Are local or customary forms of negotiation sufficient for enabling herders to gain sufficient equitable
and legitimate access to landed resources? When and how does local social flexibility and
negotiation over land lead to exclusion or inclusion of particular groups of resource user?
Although I find much evidence in support of Peters’ concern about increasingly competitive and
conflictual relations over local access to land, I argue that this is trend is as much a product of
previous state policy, which has compounded shortages and insecurity over land and landed
resources. I demonstrate - albeit for one set of resource user relations (herder - farmer) - that social
negotiation over landed resources continues to be key for herders maintaining their access to these
resources. At the same time, I demonstrate that informal and socially negotiated land-use agreements
are not sufficient. I argue that locally barely legitimate access to these resources by marginalised
resource users can be made more secure by approaches to land use management that seek to
address marked inequalities in power and facilitate the building of further local legitimacy for
different resource users through local fora.
The underlying rationale of the thesis’ questions lies in the fact that in Tanzania customary tenure
practices and people’s relationships with the land have been long disrupted and marginalised by the
pre- and post-independence state. Thus, people have experienced major disruptions to their
customary resource tenure practices and systems which have been censured, marginalised and
overridden by the state (see Chapter Three). Many rural people continue to contend with the after-
effects of successive state-mediated appropriations and re-organisations of entire landscapes (which
have not ceased) that culminated in the Ujamaa villagisation schemes of the 1970s. I argue that these
and other factors - such as on-going changes in political economy - continue to comprise significant
challenges and constraints for people’s livelihoods and land use relations, particularly for
marginalised categories of resource users.
The analytical framework

As a multi-disciplinary study, the thesis is conceptually divided into three main components - although
linkages should be seen as running through all three. The three components of the thesis in order of
40
appearance are firstly, the political economy of natural resources and land in Tanzania; secondly, the
socio-ecological and socio-economic aspects of people’s rangeland-use, and; thirdly, the socio-
political processes of land-use relations in the Idodi villages.
Since the analysis of livelihoods and land relations is potentially complex, my approach progressively
contextualises the study by firstly reviewing key past and contemporary developments in the political
economy of rangeland management in Tanzania. There are several ways of approaching this
analysis - I focus on two themes that have particularly impacted on people’s land relations in the
Idodi rangelands - state-imposed landscape reorganisation and state control of people’s production
relations and natural resource use. On the basis of these themes and a synopsis of current land
reforms in Tanzania, I explain how cascades of people have moved through Tanzania in search of
land and livelihood. I explain how these developments have had a direct impact on the livelihoods
and land relations of people living in the Idodi rangelands. I then introduce the Idodi rangelands and
provide a review of relevant background information, including a description of current village
administration and land tenure practices.
Having contextualised and situated the field study, I proceed to analyse the socio-ecological and
socio-economic aspects of people’s livelihoods in the Idodi rangelands. Since the Idodi rangelands are
home to a polyethnic community, I disaggregate and review the factors underpinning the different
major in-migrations that have occurred in the rangelands. In order to more easily understand people’s
livelihoods, I define and separately investigate two categories of land-user - herders and farmers -
on the basis of their different social organisation and means of production. In essence, and as will
emerge from the thesis, this categorisation constitutes both an emic and etic labelling of actors which is
further employed to frame the analysis of land-use relations in the third component of the thesis.
Having investigated and described key attributes and trends in people’s livelihoods, the third
component focuses on the socio-political nature of land-use relations between herders and farmers in
the Idodi rangelands. I focus on the varying strategies employed by different herders to secure
access to key landed resources. Much of the analysis lies in examining the negotiability of rules and
processes, and focuses on the investment by herders in social and economic relations with farmers as
an important strategy for maintaining access to key landed resources.
The dryland - wetland rangeland areas of Tanzania, for example those of the Idodi rangelands, are
of particular interest as they provide microcosms and bounded arenas in which:
the contrasting livelihood outcomes and socio-ecological practices of different livelihood groups
can be studied and understood in some detail
the socio-environmental past may be more easily deduced
the range of competing interests that are produced and played out can be examined over a
longer term period - for example, between herders, farmers, government sanctioned wildlife
interests and other groups;
the processes of social differentiation that often occur can be studied;
41
the political economy impacts of past and current policy on different resource users may be
relatively easily traced and analysed;
people’s exposure to increasingly differentiated rights of control and access to land and landed
resources may be understood as part of the interaction of the above factors.
Moreover, these lines of inquiry may be pursuable across a range of different levels of agency and
interest groups - individuals, households, resource user or artisanal groups, corporate groups - within
different and changing institutions and structures.
2.7.3 The study methodology

The multi-disciplinary study methodology was developed in such a way that different lines of inquiry
could be pursued simultaneously. The lines of inquiry were carried out through both formal and
informal data gathering processes. The formal data gathering process consisted of carefully
developed and piloted randomly sampled semi-structured multi-round household survey (for farmers),
and multi-round household-based census (for one group of herders). My approach to investigating
land holding and its impact on people’s livelihoods - in this case farming livelihoods - was carried out
at a household level. It is important to emphasise that this level of analysis may mask underlying
inequalities within the household - but unfortunately it was not possible to sufficiently disaggregate
my data collection and analysis to reveal inequalities in terms of gender, age and kin-status.
Nevertheless, in terms of wider comparisons, the household remains a strongly appropriate and well
recognised level at which analysis can be carried out. Less formal data and material collection
techniques were used to pursue extended case study (e.g. van Velson 1967, 141-149), life history
and other information needs, based on formal interviews, discussions and conversations, participation
at public and other meetings, participant observation and excursions to different places of interest. In
addition, archival work was carried out at village, divisional and district levels and at the Tanzania
National Archives. The rationale for the different interactions and how they were employed to collect
specific data and material are briefly explained as follows:
Multi-round household farmer survey: a random semi-structured multi-round household survey16 was
designed and, after piloting, carried out to gather data from village-based farm households. The
household was the most appropriate sampling level for the farming community in Idodi since it
comprises the domestic group (Fortes 1971, 2) and it is an established level at which data on
people’s livelihoods can be systematically gathered and analysed (e.g. Ellis, 1998). The household
survey was carried out twice during the two year field work period.
16 The survey was administered on the basis of household lists that were generated from asking sub-village
chairpersons to list all the households in their sub-village and cross checking these with the village household
register.
42
Household-based census: a semi-structured household-based census17 for Ilparakuyo herders living in
the Idodi villages was designed and implemented in close consultation with Ilparakuyo informants. The
census was designed to gather detailed data on household composition, herd numbers and structure
(on the basis of emic categories) and farm fields cultivated. Although there are two pastoralist groups
in the Idodi villages, the Ilparakuyo and Barabaig, only the Ilparakuyo were included in household
survey work. It was decided not to survey Barabaig households resident in Idodi as circumstances did
not sufficiently allow for this.
Key informant interviews and conversations: Key informants (men and women) were important for
discussing farming and herding practices, constructing extended case study material, and providing
local oral histories. In addition everyday interactions with people, whether on the road, at a market
place, in a village beer club or at the monthly market, often provided useful background information
or asides about one or other event or issue under investigation.
Participation at public and other meetings: A point was made of attending public meetings. In
addition, semi-quarterly divisional government meetings were attended as these often raised a useful
vignette of some of the major issues ongoing throughout the Idodi villages. Some ‘workshops’ were
attended that were convened by the district (at the behest of the Ministry of Livestock) to examine
herder-farmer conflict, and to look at pastoralist issues.
Archival research: Local village and divisional records (such as they existed) were examined (with
permission), which yielded a limited amount of information for the period between 1970 and 1990.
The national archives in Dar es Salaam were consulted, yielding information from between the mid
1920s to the late 1950s. Further archival work was carried out at Rhodes House Oxford. The aim of
the archive work was to piece together a framework of the key events - from the state’s perspective -
in the Idodi landscape over the latter half of the twentieth century in Idodi.
17 A census was chosen as the most appropriate survey technique as, at the beginning of fieldwork, there were
52 Ilparakuyo homesteads in the field area, and in order to ensure statistical validity during later data analysis,
it was considered appropriate to survey all the homesteads (which was achieved).
43
3 Tanzanian semi-arid rangelands in perspective
3.1 Introduction
The Tanzanian semi-arid rangelands comprise between 65-85 per cent18 of Tanzania’s area. Over
the last half century, they have become increasingly demarcated by the state into different land-use
areas. A large extent of the rangelands has been exclusively set aside for protected areas, and
other parts now comprise large ranches and plantations. The demarcation of the rangelands, together
with other key developments, such as Ujamaa and the marginalisation of customary land tenure
practices and systems, has led to increasing insecurity and shortage of land and landed resources for
herders and farmers in Tanzania. As a result of these developments, many herders and farmers have
been compelled to move in search of new land and landed resources that are still to be found -
albeit decreasingly so - in the dryland-wetland areas of Tanzania’s semi-arid rangelands (see
below). These important rangeland areas can be viewed as rapidly filling ‘frontiers’, which having
become home to polyethnic in-migrant communities of herders and farmers. The dryland-wetland
‘frontiers’ are now increasingly characterised by growing levels of competition and conflict over
access to key resources.
In this Chapter, I therefore provide a summarised account of the developments which I view as having
most impacted on rural livelihoods and land relations in the Idodi rangelands. I initially focus on
Shroeder’s (1999, 360; see Chapter Two) twin problematic of dearth and diversity as being a key
factor in contemporary land-use outcomes. In Tanzania this problematic has developed through the
creation of an extensive protected area network and the extensive resettlement of farming and
herding populations as part of a strategy to capture their production by the State. I then move on to
review other related developments in the political economy of Tanzania over the last 30 years which
have further contributed to the displacement of farmers and herders and the disruption of their
livelihood and land-use practices. Increasing state control emerges as a pervading theme. In recent
years and in contrast to previous policy, the current emphasis on decentralisation and local
government reform is now viewed as a key strategy for reversing some the inequities and deficiencies
of the past, and for improving governance, service delivery and democracy. Yet I demonstrate that in
Tanzania the decentralisation of control (across a wide range of governance and administration
issues) has been resisted in two key areas - the first is the wildlife sector. The second is the
administration of land - which the state has insisted retaining ultimate control of, despite paying
greater deference to local or customary land tenure practices. I argue that both these issues impinge
strongly on farmer and herder livelihoods. Finally, in the light of these developments, I describe how
both herders and farmers have been compelled to move through the landscape in search of key
resources in frontier rangelands, such as those of Idodi.
18 The estimates vary according to different ways of defining a rangeland (see Bourn and Blench 1999, 7).
44
This chapter therefore provides the context for introducing the Idodi rangelands, its peoples and their
livelihoods and land relations in the following chapters. It also provides key evidence for my
argument that previous state interventions and policies are heavily implicated in the increasing levels
of competition and conflict over land that are occurring today in Tanzania’s semi-arid rangelands.
3.1.1 Defining the semi-arid rangelands

Before proceeding, it is appropriate that I clarify what I mean by the semi-arid rangelands. Arid and
semi-arid rangelands cover two thirds of sub-Saharan Africa (Solbrig 1993, Le Houérou 1989 cited
in Homewood 2004, 125). The semi-arid rangelands can be generally defined by the length of
growing period (LGP) for annual plants - areas in which the LGP ranges from 79 - 179 days (FAO
1984 cited in Mortimore 1998, 10). The LGP for arid rangelands is defined as falling between 1 -
79 days (FAO 1984 cited in Mortimore 1998, 12). Land cover in the rangelands includes grazing
land interspersed with cropland mosaic and woodland (Homewood 2004, 125).
The semi-arid rangelands of Tanzania - equivalent to areas falling between the rainfall isohyets of
500-800mm - stretch from the northern plains of the Serengeti, through the central plains and down
to the south-west towards the Usangu plains. To the eastern and western side of this belt lie higher
rainfall areas which receive 800-1000mmyr-1 (see Figure 3.2). Higher rates of precipitation occur in
the highlands and along the coastal strip. The central zone of the semi-arid rangelands is less prone
to tsetse fly and trypanosomaisis (see Figure 4.119) which is endemic to the western and eastern
Brachystegia ‘miombo’ woodlands. There has been a historical long-term flux in the boundaries
between endemic and non-endemic tsetse zones (e.g. Ford 1971). A substantial expansion in the
distribution of the tsetse fly has occurred in the last 50 years, which has been largely attributed to
changes in human land-use and agro-ecological practices. It is thought that at least 60 per cent of
Tanzania is currently varyingly infested with tsetse (Galaty 1988 - but see the higher and more
recent FAO-PAAT estimate as provided in Figure 3.1), adversely impacting livestock keeping and
making human inhabitation difficult20, particularly in the worst affected areas. Thus those parts of the
semi-arid rangelands of Tanzania that remain relatively free of trypanosomaisis are very important
for livestock herders.
A feature of the semi-arid rangelands is that they contain both extensive dryland areas and usually
smaller discrete ‘dryland wetlands’ that may often be part of larger river systems. The ‘dryland
19 The tsetse fly distribution depicted in Figure 4.1 is taken from the UN Food and Agricultural Organisation’s
(FAO) Programme Against African Trypanosomiasis (PAAT) http://www.fao.org/ag/againfo/programmes
/en/paat/maps.html. The map shows the tsetse fly distribution for Glossina morsitans produced by modelling the
‘known’ presence and absence of the flies between 1999 and 2003 (using maps developed by Ford and
Katondo [1977] modified with more recent information collected from national and international agencies and
researchers).
20 Although localised agricultural bush and woodland clearance may significantly reduce local tsetse fly
densities around settlements.
45
Figure 3.1: The topography of Tanzania, location of major wetlands and general tsetse fly distribution
L. Victoria
L.Natron
L.Manyara
L.Eyasi
Nyumba ya
MALAGRASI Mungu
UGALLA
L.Tanganyika
WAMI
L.Mtera
RUKWA
L.Rukwa
KILOMBERO
USANGU
MBWEMKURU
L.Malawi
Legend: Legend:
Tsetse fly distribution
(Glossina species) Wetlands/Swamps
Water Field site

Rift Valley Field site Water
46 Tsetse fly distribution from http://www.fao.org/ag/againfo/programmes/en/paat/
documents/maps/Fusca/fuscagrouppa.gif (map not dated) Accessed on 23-12-04
Figure 3.2: Average rainfall distribution for Tanzania
Average
annual
rainfall for
Tanzania
Field site
location
Source: Institute of Resource

Assessment, University of Dar Es Salaam
wetlands’ are frequently very important for dry season grazing for both (agro)pastoralist livestock,
and also in extensive parts of east and southern Africa, wildlife. In areas with higher rainfall,
particularly for the semi-arid rangelands, ‘dryland wetlands’ are also often farmed by agriculturalists
who productively exploit the rich fertility of these wetlands during the wet season. In turn, the dryland
parts of the semi-arid rangelands are important wet season dispersal areas for livestock and wildlife.
The semi-arid rangelands therefore support a diversity of land-users - agriculturalists, pastoralists,
hunter-gatherers, commercial ranchers and farmers and the wildlife tourism industry. However, the
semi-arid rangelands are becoming an arena for increasing levels of conflict as more people and an
increasingly diverse range of interests depend on them for their livelihoods, competing for access to
their resources.
47
3.2 Some key developments affecting rangeland management in Tanzania
3.2.1 Colonial partitioning and landscape re-organisation

The lie of the contemporary Tanzanian landscape has strong links to the colonial past21. The German
and later the British colonial authorities22 partitioned Tanganyika into administrative districts, and
latterly, provinces and chiefdoms. After the German colonial period, the British set up a system of
indirect rule instituted in 1926. Native authorities were established23 on the basis of colonial
interpretations and inventions of ‘traditional’ chiefdoms, ethnic identity and social organisation (Iliffe
1979, 318-342; Shivji and Maina Peter 2000, 13-15). It was through these native authorities
(abolished at independence) that the British colonial administration collected tax, operated a judicial
system and generally administered Tanganyikans.
The British administration retained most of the protected area estate of forest and game reserves
that it had inherited from the Germans (Neuman, 1998, 100). The first game hunting reserves had
been proclaimed by the German colonial authorities as early as 1896 (Wanitzek and Sippel 1998,
114)24. During the British colonial period, the protected area network was further expanded as the
colonial administration re-ordered the landscape as part of its programme of exerting political and
economic control over Tanganyikans (Neumann, 2001). For much of the colonial period, the
administration waged a campaign to halt the spread of the tsetse fly and improve small-holder
production of agricultural export commodities - particularly during the depression and war years of
the 1930s and 1940s when the colony’s finances were in crisis (e.g. Iliffe 1979, 347-356).
In parts of Tanganyika, people were moved into settlement concentrations as part of the colonial
project of controlling sleeping sickness and wildlife, and capturing and increasing the agricultural
production, labour and tax returns of rural Tanganyikans. A number of these interventions were later
to lead to the creation of new protected areas (Kjekshsus 1995). For example, in south-western
Tanzania, the Selous Game Reserve was much expanded into western Liwale District as part of a
campaign during the 1930s to control elephant in the east of the district. A series of enforced
resettlement and concentration schemes were carried out to depopulate the western district as people
were moved to concentration areas in the east, and as the elephant were driven west by the Game
Department. The impact of the scheme was substantial - 40,000 people left western Liwale District
during the 1930s and early 1940s (Yeager and Miller 1986 cited in Neuman 1998, 147) and some
of the most fertile areas of farmland in the district were lost to wildlife conservation (Neuman 2001,
658). The depopulation of west Liwale District was only brought to an end when it was realised
21 The history of colonial hunting and wildlife protection has been contrastingly documented by Ofcansky
(2002), Bonner (1993), Mackenzie (1988), and (in an edited volume) Anderson and Grove (1987).
22 Tanganyika was the protectorate of the German East Africa Company between 1895-1890, and then part of
Imperial German East Africa from 1895-1919. At the end of World War I it was ceded to Britain in the Treaty
of Versailles as a protectorate on behalf of the League of Nations.
23Under the Native Authorities Ordinance of 1926.
24 Between 1906 and 1914,231 forest reserves and 18 game reserves were proclaimed (Neuman, 1998, 99).
48
during the mid1940s that the remaining human population comprised a useful labour reserve for what
was to become the ill-conceived and infamous ground-nut scheme (Neuman 2001, 658-660; see also
Iliffe 1979, 440-2). In northern Iringa District similar reasons were put forward by the colonial
administration for expanding the Rungwa Game Reserve southwards into the Idodi and Pawaga
rangelands. Specifically, a dispersed rural population living in a huge rangeland was perceived as
presenting substantial administrative problems for the colonial administration, which was also faced
with the challenge of attempting to control the continued spread of the tsetse fly (see Chapter 4).
During the late inter-world war and post-world war years in colonial Tanganyika, as rural human
populations recovered and grew from their depressed levels at the end of the previous century
(Kjekshus 1995), people and wildlife25 came increasingly into conflict. This conflict was exacerbated
in part due to new wildlife rules and regulations now being increasingly enforced by the Game
Preservation Department. Rural populations were increasingly officially barred or regulated26 from
freely hunting and effectively controlling wildlife populations. From the early years of the British
Colonial administration, an uneasy stand-off had existed between some administrators with a more
tolerant attitude allowing local Tanganyikans to continue customarily using wildlife and forest
resources, and others, particularly in the Game Preservation and Forestry Departments, who were
against such practices. The latter argued for further restrictions on customary resource use rights and
an expansion of the protected area estate (Neuman 1998, 102-106). Yet the heavily stretched
Game Department fought a loosing war against crop-raiding elephants27 which raided the fields of
many small-holder farmers over disparate areas of Tanganyika28 (Neumann, 2001).
During the 1940s and 1950s an international wildlife conservation lobby became ever more
concerned that local Tanganyikans were a fundamental threat to nature and wilderness places. An
increasingly powerful European coalition successfully lobbied the colonial office in London for the
expansion and stricter enforcement of the protected area estate in Tanzania (Neumann, 1998, 122-
25 Human-wildlife conflict occurred particularly with elephant which had been unaffected by the rinderpest
pandemic, and which had increasingly begun to recover from the ivory trade of the mid-19th century (Kjekshus
1977).
26 Hunting without the use of precision firearms was initially allowed in Game Reserves but not in National Parks.
27 Although elephants had been heavily hunted in the 19th century as a result of the ivory and slave trade, the
elephant population had much recovered by the 1930s. By this period, elephants were becoming a great
nuisance to farmers (together with other small vermin) and the subject of much complaint by farmers to the
Game Preservation Department. For example, during the 1920s, game scouts were killing 800 elephant per
year in crop depredation control efforts, which consumed 75 percent of the game department’s energies
(Neumann 1998, 648). Neumann also provides an estimate that one quarter to one third of the country’s annual
food crop production was lost to wildlife crop raiding during the 1920s to 1940s. In fact, by the 1940s the
number of elephant shot per year had risen to over 3,000, and between 1931 and 1950, 33,462 elephant had
been culled with no check to the population (Neumann 2001, 659).
28 For example, elephant populations greatly expanded in the Dabaga area of Iringa District in central
Tanzania and in Liwale District in the south east of the country (Neuman, 2001).
49
148). With the passage of the new Game Ordinance of 1940 and National Parks Ordinance of
194829, the wildlife estate began to be further expanded (Kjekshus 1995). For example, in eastern
Tanganyika, the substantial extension of the Selous Game Reserve during the mid 1940s (Neumann
2001); in central Tanganyika, following the creation of the Rungwa Game Reserve in about 1937
(Jennings 1994) the extensive annexation of the new Ruaha section to the Rungwa Game Reserve to
its south in 1954, and; in northern Tanganyika the withdrawal of the Maasai from the previously
created Serengeti National Park in 195830. Thus by the declaration of independence, the
Tanganyikan landscape had become extensively partitioned and a substantial protected wildlife and
forest estate had been created.
3.2.2 Post independence state control: Ujamaa, deconcentration and decentralisation

The process of resettling remote rural populations into settlement concentrations began during the
British colonial administration, was to be carried on by the independent Tanganyikan31 state.
Although in some regards, during the initial years of independence, rural populations began to
achieve a higher level of autonomy from the control of the state, this rapidly changed as the decade
progressed (Hyden 1980).
29 Game ordinances had been passed in both the German colonial period (1896, 1898, 1900, 1903, 1905,
1908 and 1911) as well in the early British colonial period (1921) (Nelson et al. 2003, 8; Wanitzek and Sippel,
1998). Although the Game Preservation Ordinance of 1921 re-gazetted the game reserves created during the
German colonial period, the Ordinance maintained that, ‘… the native should be regarded as having a moral
right to kill a piece of game for food’ (Neumann 1998, 100). This position was to increasingly change to one of
growing restrictions from the 1940s onwards, for example with the passage of the Game Ordinance of 1940,
the National Parks Ordinance of 1948, the Fauna Conservation Ordinance of 1951 (which introduced a game
hunting licensing system for African Tanganyikans in coordination with native authorities) and the current
National Parks Ordinance of 1959, the latter of which remains in force - see Wanitzek and Sippel (1998) for a
detailed review.
30 Throughout the 1940s and 1950s people were increasingly evicted by force from newly created or extended
Game Reserves and National Parks or cajoled to leave - as in the case of the Ndorobo, Ikoma, Sukuma and
Maasai peoples in the Serengeti (see also Nelson et al. 2003, 11-15; Neuman 1998, 129-139; Homewood and
Rodgers 1991, 69-83). In this latter case, the Maasai agreed to withdraw on accepting a government
undertaking that guaranteed them the right in perpetuity to live in the Ngorongoro highlands and crater to the
east. This latter undertaking has for many years increasingly and, sometimes violently, been reneged upon, both
through successive amendments to the law increasingly extinguishing customary rights and also through extra-
legal actions taken by the managing Ngorongoro Crater Conservation Authority. Today the Maasai living in the
Ngorongoro highlands are supplicant to this parastatal authority, which controls - in effect - nearly all natural
resource-based activities within a declared Conservation Area. The parastatal has even gone so far as to
attempt acquisition of land title for the Conservation Area, when title belongs to the Maasai as a right of
customary occupancy (Shivji and Kapinga, 1998).
31 Tanganyika gained independence in 1961. In 1964, after the Zanzibar revolution, Tanganyika and Zanzibar
formed a union to become the United Republic of Tanzania.
50
The early independence years
In the years leading up to independence in 1961, the colonial administrative system of native
authorities was initially supplemented with a system of elected District and Town Councils (Cole and
Denison 1964 cited in Charnley 1994, 55)32. At independence, native authorities were abolished33
being entirely replaced by the District and Town Councils in a strategy to eliminate ‘non-egalitarian’
customary systems of authority. The vesting of power in the District and Town Councils led to an
increasingly politicised and TANU34-orientated administration being established, and the councils soon
became implementing bodies for the central government’s development plans (Havnevik 1993 cited
in Shivji and Maina Peter 2000, 18).
The new District Councils lacked the power to prohibit, restrict or regulate pastoral migration and
agricultural settlement (James and Fimbo, 1973; Charnley, 1994: 58). James and Fimbo (1973 cited
in Charnley, 1994: 58) note that the whole structure of ‘traditional’ hierarchy in relation to land
administration crumbled and a vacuum was left in its stead. Yet throughout the mid to late 1960s the
government passed a series of laws that communalised land35 and exerted greater state control over
herders and farmers, leading up to the Arusha Declaration of 1967 and the subsequent pathway to
Ujamaa and villagisation.:
For example, the Range Development Act of 1964 enabled the creation of Range Development
Commissions and ranching associations36. The Commissions, entirely consisting of appointed
administrators, were given the power to issue bylaws governing the use of land by herders. Ranching
associations were to be set up that would adhere to the Commissions’ bylaws that regulated the
movement and numbers of livestock in an association’s area. All customary rights held within each
association’s area were to be extinguished, including those of the members of the association. Failure
to comply with the bylaws could lead to expulsion without further provision of land elsewhere (Sundet
1997, 18; Wily 1988, 88). A series of range development projects ensued which were externally
conceived and implemented with no appreciable consultation with the targeted ‘beneficiaries’ (Sundet
1997, 19; Hodgson 2001, 208-220). They subsequently failed (e.g. Cliffe and Cunningham 1973;
Jacobs 1980).
The Land Tenure (Village Settlement) Act of 1965 provided the facility for government-sanctioned
Rural Development Commissions to extinguish customary land tenure rights in a settlement area, which
32 District and Town Councils established under the Local Government Ordinance of 1953.
33 Under the African Chiefs Ordinance (Repeal) and Native Authority (Repeal) Acts of 1963.
34 The Tanganyika African Nationalist Union - which was to become the only and ruling political party in
Tanganyika, and latterly, Tanzania. For example, the Local Government Election Act of 1965 decreed that all
District Councillors had to be TANU members (Mwinasa and Shauri 2001, 8).
35 These were the Rural Settlement Commission Act, 1963; the Range Development and Management Act, 1964;
the Land Tenure (Village Settlements) Act, 1965, and; the Land Acquisition Act, 1967.
36 This development followed from recommendations provided by a United States Agency for International
Development (USAID) mission to Tanzania in 1963 (Shivji and Maina Peter 2000, 19).
51
were then to be replaced by ‘derivative’ rights for the allocation of plots of land to ‘entrepreneurial
progressive’ farmers (Sundet 1997, 19 & 26; Shivji and Maina Peter 2000, 18). Thus while
customary rights to land were to be put under administrative arbitration, the Act was in fact little
used, as the ‘transformative37’ approach to national development which had underpinned its creation
was subsequently abandoned (Sundet 1997, 19). However the Land Acquisition Act of 1967 greatly
expanded the Government’s rights to extinguish customary rights for ‘public purposes’ (Sundet 1997,
19).
These combined legislative developments had major implications for rural Tanzanians (Presidential
Commission of Inquiry into Land 1994, Vol. 1 cited in Shivji 1998, 6) 38. Firstly, land tenure was to be
removed from the domain of customary law. Secondly, there was an implied perception of the
evolution of the land tenure system to some form of individual tenure under the firm supervision of the
state (see previous discussion on ITR in Chapter Two). Thirdly, land tenure and land-use systems were
to be administered and managed from above through a series of detailed regulations, rules and
bylaws, with corresponding penalties (e.g. loss of land) and criminal prosecution for the breaking of
these laws. Fourthly, farmers and herders were to have virtually no role in the planning,
administration and management of the land on which they lived.
The Arusha Declaration and the onset of Ujamaa

The development of Tanzania took on a new impetus with the Arusha Declaration of 1967. TANU,
under the leadership of Julius Nyerere, adopted a doctrine of African socialism and self-reliance, and
embarked on a pathway of nationalisation of industry, agriculture and transport. Later in the same
year, a policy document, ‘Socialism and Rural Development’, set out the official blue-print for the
establishment of Ujamaa (Swahili: familyhood) villages (Pratt 1971, 237 cited in Sundet 1997, 24).
All political power was to become consolidated in the party (Hyden 1980)39.
The policy of Ujamaa entailed the restructuring of diffuse rural communities into self-reliant Ujamaa
villages. The aim of creating Ujamaa villages was to transform rural agricultural productivity upon
37 In the early independence era, there were two main approaches to development adopted by the Tanzanian
Government and World Bank (then the IBRD). The first was the ‘transformative’ approach in which large inputs
of capital and technical support would transform national development through resettlement. The second was the
‘improvement’ approach in which resources were to made available to ‘progressive farmers’ who would provide
the country with the fastest economic growth (IRBD 1961, 101-128 cited in Sundet 1997, 15; Shivji and Maina
Peter 2000, 18). Despite continuing debate within TANU and technocrat cadres of government, the
‘transformative approach’ was dropped after the first five year plan in 1966 due to its high cost and failure
(Sundet 1997, 15). Ironically, the Ujamaa and villagisation debacle of the 1970s was an adoption of the very
‘transformative’ approach dropped in the previous decade.
38 Village-based land tenure is discussed further in Chapter Three.
39 This was exemplified during the late 1960s by the party compulsorily taking over the only independent and
voluntary Ujamaa villages that had formed themselves into the Ruvuma Development Association (RDA) (Coulson
1982, 263-271 cited in Sundet 1997, 36).
52
which the party - now synonymous with the state - had placed its hopes of national self-reliance and
economic development. Ujamaa villages would provide the country with agricultural surpluses -
achieved through economy of scales of communalised production (Hyden 1980). Moreover, the
clustering of people together, especially from remote rural areas close to lines of communication
would enable the easier provision and development of health, education and agricultural extension
services.
Between 1968 and 1975, a programme of ‘operations’ was launched to create Ujamaa villages.
Initially the formation and registration of these Ujamaa villages was voluntary, but the overall
registration rate was relatively low, as rural people’s priorities and livelihood objectives differed
from those of the party (Hyden 1980). As might be expected, rural people were more concerned
with securing their livelihoods and they often could not sufficiently relate to the party’s top-down
requirement of nation building and surplus production of targeted crops - despite political
indoctrination processes (Hyden 1980). Moreover, in terms of local realities, even if rural Tanzanians
had supported many aspects of Ujamaa, the nature of their seasonal labour priorities clashed with the
labour demands of communal production. Communal production was therefore frequently less
productive than private smallholder agriculture, despite higher levels of inputs and mechanisation.
Further, as Hyden points out, whereas reciprocity within extended families and kin groups was
common place, the compulsory extension of this practice to a much wider locus such as that applicable
in an Ujamaa village - often made up of people from disparate origins and backgrounds - was
largely a foreign concept and thus resisted (Hyden 1980).
Previously voluntary, Ujamaa was to become a matter of compulsion. Privately held land and
property was nationalised in 1970-197140 (Sundet 1997, 36-39). However, the most significant
development in the Ujamaa period for rural populations was the decree by Nyerere in 197341 that
all people were to compulsorily move and live in designated Ujamaa villages by the end of 1976.
State and party officials moved with alacrity to implement the villagisation decree. The planning of
the Ujamaa villages was unfortunately frequently rushed and inadequately thought through (Shivji
1998, 12). New village sites were often poorly located in relation to the agro-ecological
sustainability of the new settlements (or expansion of established ones) and the simple practicalities of
life (such as the distance to drinking water and fields) were often overlooked. The operation was
substantial with up to five million people being moved to new Ujamaa villages nationwide (Hyden
1980). There is now considerable evidence that force was used arbitrarily as militia and paramilitary
40 The legislation for the nationalisation of property and land (the Freehold Titles [Conversion] and Government
Leases Act of 1963 and the Rights of Occupancy [Development Conditions] Act of 1963) had been enacted by
Parliament in 1962, and went largely un-noticed at the time, as later Nyerere wryly observed in 1971 (Sundet
1997, 17).
41 Supported the Rural Lands (Planning and Utilisation) Act of 1973. Shivji (1998, 13) describes this act as a
thinly veiled attempt to confer open-ended powers on the President and appropriate Ministers to ‘extinguish’
customary ‘rights’ without due process or any legal redress.
53
units were deployed as part of the campaign (Shivji 1998, 12). The Idodi rangelands were no
exception and hamlets were burnt out with little notice (pers com anon) - see Chapter Five.
The Presidential Commission of Inquiry into Land (1994) states of the villagisation campaign:
‘One major feature of the ‘operation’ stands out above all. There was total disregard
of the existing customary land tenure systems as well as the fact that virtually no
thought [was] given to the future land tenure in the newly established villages’ (Vol. 1,
43).
Villagisation opened up possibilities of alienation of village land on a scale greater than even that of
colonial times (Tenga 1987 cited in Shivji 1998, 12). Shivji (1998, 12) observes:
Villagisation had a major impact on land tenure generally and the rights of rural land-
users in particular. In effect, it amounted to major land reform. Yet that was not how it
was conceived, planned or implemented. The result was confusion in tenure and the
total undermining of security for customary landholders.
The farming populations were not the sole target of the villagisation campaign. For example, Maasai
pastoralists in northern Tanzania were forced, as part of operation imparnati (Maa: permanent
habitations), to move into livestock development villages (Arhem 1985a; Ndagala 1985). In each of
these development villages there was to be a central settlement, and wet and dry season grazing
areas (Hoben 1976; Parkipuny, 1979). Although existing land-use and settlement patterns were used
as the basis for the new livestock development villages, a new structure of executive leadership was
imposed on Maasai society based on alien administrative village institutions together with restrictions
upon their stock holdings and movements. Many Maasai were wary of the new villagisation
programme and they considered it just another step taken by the government to subjugate and
control their relationship with the land (Arhem 1985a). This suspicion has since been borne out as
large tracts of land and important natural resources in Maasailand have been corruptly allocated
without consultation by the state to other interests - commercial agriculture, wildlife hunting companies,
and private individuals (e.g. Shivji 1998, 32-39).
The villagisation process was completed in 1976 with the passage of the Villages and Ujamaa
Villages Act of 1975 which provided the enabling legislation for the registration of villages, the
demarcation of their boundaries42, the election of village councils and a village administrative system
right down to 10 household (cell) leaders. Land was to be allocated by District Development
Committees43 that then would be re-allocated to villagers by each village council (Charnley, 1994:
59; Shivji 1998, 14). This latter period perhaps represented the apogee of autocratic state control
and the expansion of state space through decentralisation of state power vested in powerful
42 In very many villages, this process remains unachieved, and long-term inter-village boundary disputes are not
uncommon.
43 The Villages and Ujamaa Villages Act of 1976 replaced District Councils with party-controlled District
Development Councils.
54
Regional Commissioners and their administrations44. A significant impact of the Villages and Ujamaa
Villages Act, in addition to the vague, inconsistent and often contradictory nature of laws pertaining
to land (e.g. Sundet 1997, 47), was that customary land tenure was de jure and de facto heavily
eroded to the point of extinction. For example, the minister responsible for villages (the Prime
Minister) was delegated the authority by the President to issue directions for land-use as he saw fit
through the District Development Councils. In turn the Village Councils were obliged to give effect to
such directions on land-use allocation as issued by the District Development Councils (Sundet 1997,
49; Shivji and Maina Peter 2000, 23-24).
Deconcentration of state power

In addition to the villagisation process of the 1970s, a parallel programme of administrative
‘decentralisation’ was launched in 1972 and swiftly implemented (Shivji and Maina Peter 2000, 20).
The representative local government authorities which had been inherited from the colonial period
were abolished45. They were replaced by larger District Development Councils and Regional
Development Councils (Mwinasa and Shauri 2001, 8; Shivji and Maina Peters 2000, 19-21). Central
government administrative powers were deconcentrated to the regions46 and in turn to the districts,
but the central party and central government retained the real decision-making power (Rweyemamu
1974, 125-126 cited in Sundet 1997, 39). The effect of this decentralisation process was that,
together with villagisation, the government was brought closer to the rural population and direct links
of command were established between the centre and rural areas (Sundet 1997, 52). Thus the twin
programmes of villagisation and ‘decentralisation’ led to extensive control of the people by the state,
in which villages ‘were to be developed’ and regarded as ‘project units’ (URT 1978, 115 cited in
Sundet 1997, 46). In this regard, the decentralisation process that was carried out during this period
can be more accurately described as having been a ‘deconcentration’ process, in which the workload
of central government was shifted outwards, but real power was retained by the central state.
The end of Ujamaa: decentralisation, liberalisation and local government reform

The Ujamaa years drew to a close in the early 1980s when the state and party conceded that macro-
economic reform was necessary if the national economy was not to entirely collapse. In 1986, with the
retirement of President Nyerere as Head of State, a more neo-liberal macro-economic and political
outlook was adopted by the central government when an IMF structural adjustment reform package
44 Up until the mid 1990s, regional administrations were the most powerful locus of local government. However,
with the launch of a new local government reform process in the mid 1990s, and with the passage of the
Regional Administration Act in 1997, substantial administrative powers have since been moved to the district
level and districts are now at the centre of local government.
45 As a result of the recommendations of a management consultancy firm McKinsey and Co (Coulson 1982 and
Mushi 1978 cited in Shivji 1998, 12).
46 The decentralisation process (legislated through the Decentralisation of Government Act of 1972) facilitated
the creation of Regional Integrated Development Plans (RIDEPs) which, coordinated by a newly created Prime
Minister’s Office and supported by donor assistance, were to be implemented by Regional Development
Councils (RDCs) which in turn were to oversee the activities of DDCs.
55
was agreed47. From this point onwards, market liberalisation reforms were implemented, including
major changes in the agricultural sector, such as the ending of input subsides and pan-territorial
pricing (Agriculture Sector Development Strategy [ASDS] 2001, 15; Bryceson 2002, 728; World
Bank 2001, 50)48. In addition, stretched agricultural and livestock extension services were to become
privatised, leading to their collapse, particularly in remoter rural areas.
In parallel with market liberalisation policies, the state embarked upon further reform of local
government. Representative local government had been re-launched in 198349, although it was
marginally democratic, and prone to inefficiency, lack of resources, corruption and overlapping
control by central government (Max 1991 cited in Shivji and Maina Peter 2000, 26; Tax and Hauck
2003, 5). The poor state of local government began to be acknowledged in 1996 with the tabling of
a proposal for a Local Government Reform Programme (LGRP)50. The LGRP was finally launched in
late 1999 (Shivji and Main Peter 2000, 30-32).
The key components of the LGRP - improving democracy and achieving decentralisation via district
councils - are seen by many economists and development experts as not only vital but inevitable for
the continued development of the national economy (for example, Watson and Baek 2001 cited in
Kelsall 2004, 34). Districts are now to be increasingly responsible for directly implementing central
government policy, with central government releasing financial grants and giving guidelines and
advice as to how to best implement policy (e.g. Tax and Hauck 2003, 4-5). In the natural resource
sector, this process has been accompanied by a raft of new policies and legislation, which on paper
set out to achieve varying interpretations of devolved natural resource management (see Section
3.2.3).
However, the LGRP has a long way to go to achieving its goals (e.g. Tax and Hauck 2003, vii-xiii;
Kelsall 2004, 70). The roles and funding mechanisms between central, regional and district
government remain insufficiently clarified (Tax and Hauck 2003, vii-viii), and a case study shows that
institutional capacity remains weak and local government is frequently corrupt, inefficient and
47 See Bigsten and Danielson (2001) for a detailed review of Tanzania’s economy since independence.
48 The termination of input subsidies resulted in farmers being unable to afford or gain sufficient access to
agricultural inputs (e.g. Jambiya 1998 cited in Bryceson 2000, 728). Correspondingly, household agricultural
income is reported to have declined – perhaps as much as 71 per cent between 1979 and 1992 (Mung’ongo’
1998 cited in Bryceson 2000, 728).
49 The District councils were reinstated in 1983, but the Party continued to heavily control them. This remained
the case until prior to the first multi-party elections in 1995, when the state and the Party (CCM) were devolved
from each other (under the Eight Constitutional Amendment Act of 1992) to enable the theoretical possibility of
a political opposition winning the election and taking power. However to this day, the CCM continues to retain
control of the majority of District Councils in Tanzania.
50 Part of the initial reform included transferring the locus of local government power from the regions (Regional
Commissioners and Regional Administrative Secretaries) to the districts (District Executive Directors and District
Councils).
56
insufficiently accountable to its electorate51 (Kelsall 2004). To date, the LGRP has focussed little on
how local government reform is to be achieved at village level, which is the locus at which local
government in Tanzania is theoretically founded and legitimised (Shivji and Peter, 2000, 35).
3.2.3 Control and space: state and nature

The post-independence expansion of state control and space has not been limited to the
administration of people but has also applied to their natural resources. Grievances against the
historical loss of land52 to protected areas, and the colonial state’s control over local people’s
relationship with the environment, had been a source of political capital for the pre-independence
nationalist movement of the 1950s (Gibson 1999). Yet despite the pre-independence nationalist
rhetoric against colonial wildlife policies, the post-independence policies of Ujamaa were to provide
further opportunity for the continuation, if not acceleration, of the wildlife and forestry policies
inherited from the British colonial period.
As many African colonies gained independence, a coalition of international and western conservation
organisations successfully lobbied the newly independent Tanganyikan government to maintain the
status quo in regard to conservation policies inherited from the colonial era. Their lobbying led to a
conference on wildlife conservation in Africa that was held in Arusha, in northern Tanzania in 1961. At
the conference, in a staged53 and now famous speech that became known as the ‘Arusha Manifesto54’,
President Nyerere delivered his landmark commitment to the continued preservation of Tanganyika’s
- and Africa’s - natural heritage. In effect, a de facto deal was arrived at, in which the conservation
lobby would provide the majority of the funding for the protected area network, if the Tanzanian
government would continue to maintain and support the protected area system inherited from the
colonial period and its legislative framework (Nelson et al. 2003, 17).
The expansion of the ‘wild nature’ estate and increasing control over its use continued through Ujamaa
underpinned by, for example, the Wildlife Conservation Act of 1974. The continued expansion of
the protected area estate55, mostly but not exclusively through re-designation of existing protected
51 Admittedly this case study was of Arusha only and looked at wider issues of civil society and local politics.
However experience indicates that Kelsall’s insightful findings are likely to hold for much of Tanzanian local
government.
52 Large areas of land had also been alienated for European and Indian owned plantations and ranches during
the colonial period. The Meru Land case in is perhaps the most well known example (see Spear 1997, 209-235;
Kelsall 2004, 16-19). Between 1949 and 1952, in order to make way for new European farms, the colonial
administration summarily evicted local African farmers from their land at Engare Nanyuki on Mount Meru in
northern Tanzania. The evictions resulted in large protests by Tanganyikans who took a petition to the United
Nations in New York which failed.
53 European conservationists purportedly wrote the speech (Bonner 1993, 11 cited in Nelson et al. 2003, 18).
54 Not to be confused with the totally separate and different Arusha Declaration of 1967.
55 Between 1964 and 1994, the number of game reserves and national parks increased from six to twenty and
from four to twelve respectively (Swai 1996, 51-52 cited in Neslon et al. 2003, 3). Further reserves (for
57
areas to higher levels of protection, further moulded the post Ujamaa political landscape. People
were precluded from returning to previous settlement areas which now had become national parks or
game or forest reserves or, in some cases, barely inhabitable as a result of the spread of the tsetse
fly. Natural resource management control was retained at ministry level with agriculture, natural
resources (fisheries, forestry and wildlife) and livestock offices at regional and district levels.
The re-organisation of local administration in 1982 provided a new legal apparatus for making local
(village) level resource use and management regulations as village councils were given the power to
legislate bylaws subject to the approval of their respective District Council (Charnley, 1994: 59).
However, the ability of village councils to regulate and control natural resource use was heavily
structured within the then current policy and party-controlled environment. Central ministries continued
to exercise tight control (through their officers at regional and district level) over formal rights to
manage and benefit from wildlife and forest resources. Until the early 1990s, very few villages
bordering areas of protected forests and wildlands had been formerly allowed to develop even the
most simple of local management regimes56. Thus despite rural people now being enabled -
nominally through their village councils - to legally exercise some limited control over their land and
natural resources, ultimately real control continued to lie with central government departments (for
example the Forestry and Wildlife Divisions) and the regional administrations.
Today, wild nature, and particularly wildlife, is largely controlled by the state, which derives
substantial revenues57 from a centrally regulated tourism and hunting industry. The ability of rural
people to similarly benefit has never much mattered to the modern state or most of its partner
stakeholders in the wild nature industry. In the late 1970s and through the 1980s, as Tanzania’s
economy faltered and went into crisis, the ability of the Tanzanian government to administer properly
its centralised protected area network of national parks and game reserves collapsed. The outcome
was that by the late 1980s, substantial declines in the elephant population58 had occurred and the
rhino had been effectively extirpated in all but a few refuge areas (Nelson et al. 2003, 24). In
addition, many Tanzanians took advantage of the de facto open access status of wildlife for
bushmeat. By the late 1980s, several conservation and donor organisations, natural resource experts
and a number of senior Tanzanian civil servants decided that the current status quo needed to be
example, the Usangu Game Reserve) and national parks (for example, Sadani National Park) have since been,
created.
56 One exception was a wildlife meat provision scheme introduced in the western Serengeti in the late 1980s as
part of the Serengeti Regional Conservation Strategy which began in 1986. However, the scheme could hardly
be called ‘community-based natural resource management’ since the scheme was run and tightly controlled by
Wildlife Division officials.
57 Annual gross income from both tourist hunting and photographic tourism has grown from about US$9million
and US$60 million respectively in 1990 to US$28million and US$725million in 2001 (World Bank 2002, 21-22
cited in Nelson et al. 2003; Baldus and Cauldwell 2004, 10).
58 In 1970, Tanzania’s elephant population was estimated at 350,000; by 1990 it had declined to 55,000
(WSRTF 1995, 37 cited in Nelson et al 2003, 24).
58
addressed urgently. Firstly, the continued de facto open access needed to be brought under control.
Secondly, the then current protected area system of national parks and game reserves was perceived
as being unlikely to maintain viable wildlife populations in the long term. It was becoming increasingly
clear that the cooperation and support of rural people - herders and farmers - living around
protected areas and in seasonal wildlife dispersal areas was to be key for securing the continued
and long term existence of wildlife. Furthermore, if these herders and farmers were to conserve
wildlife, then they must be enabled to manage it sustainably and, most importantly, derive sufficient,
tangible and continued benefits from it.
Operation ‘Uhai’ (Swahili: life) brought the de facto open access situation to wildlife across Tanzania
rapidly under much greater control. An extended paramilitary and somewhat heavy-handed
operation was designed to stop illegal hunting and to confiscate the weaponry, particularly at the
village level. While the Tanzanian state was swift in its crack down in the closing years of the 1980s,
its track record on re-endowing rural Tanzanians with the right to manage and benefit from wildlife
has been much less successful.
A five-year wildlife sector review process59 (heavily donor driven) was carried out during the first
half of the 1990s. The planning review provided recommendations for a new wildlife policy
advocating a fundamentally different approach to wildlife management in Tanzania (Nelson et al.
2003, 26). A key part of the recommendations was that local communities should be re-endowed with
wildlife management and benefit rights. The new wildlife policy60 was finally passed in 1998 and
explicitly supports this key recommendation. As a product of the wildlife sector review, and reflected
in the new wildlife policy, Wildlife Management Areas (WMAs) were conceived of as being the
protected area category in which local communities would be enabled to manage and benefit from
wildlife. Yet the WMA regulations, when published in 2002, were not what many had envisaged.
The guidelines are heavily bureaucratic and inflexible in their stipulations of how a community can be
granted and allowed to manage a WMA. Moreover the guidelines refrain from setting out what
proportion of the revenues derived from tourist hunting communities can expect to receive, leaving this
to the discretion of the Director of Wildlife. When taken together with the Hunting Regulations of
2000, communities are effectively barred from carrying out any wildlife related enterprise (hunting
or photographic) in any hunting block (see Figure 3.3 for an overview of the areas hunted in
Tanzania61) without the express permission of the Director of Wildlife in Tanzania. Given that hunting
blocks - over which the Wildlife Division has authority - cover a large area of Tanzania, community-
59 The review was carried out by the Planning and Assessment for Wildlife Management (PAWM) project,
funded by USAID, and based within the Wildlife Division in Dar Es Salaam.
60 The Wildlife Policy of Tanzania, 1998.
61 The map appears to erroneously omit the Lunda Mkwambi Game Controlled Area (see Chapter Four and
Figure 4.1). The northern part of the Game Control Area is a tourist hunting block, and the southern part is
reserved for resident hunting.
59
Figure 3.3: Wildlife protected areas in Tanzania in 2004 [not including forest reserves]
(Baldus and Cauldwell 2004)
Western Tanzania 6 Msima GCA 12 Mlele South GCA 18 Wembere OA North

1 Makere Forest 7 Ugunda GCA 13 Lake Rukwa GCA 19 Wembere OA Central
2 Uvinza OA 8 Inyonga West GCA 14 Piti West OA 20 Wembere OA South
3 Gombe GCA 9 Inyonga East GCA 15 Inyonga East 21 Itulu Forest East
4 Luganzo GCA 10 Rungwa River GCA 16 Chunya OA 22 Singida OA
5 Ugalla OA 11 Mlele North GCA 17 Utengule Swamp OA 23 Manyoni OA
Masailand
24 Maswa OA 31 Mto wa Mbu GCA 38 Lolkisale 45 Ruvu Masai
25 Nyichoka OA 32 Monduli Juu 39 Simanjiro West 46 Kitwai North
26 Sibora OA 33 Maswa Makao 40 Simanjiro Kitangare 47 Kitwai Central
27 Loliondo GCA 34 Yaeda Chini OA 41 Simanjiro Naberra 48 Kitwai South
28 Loliondo South GCA 35 Lake Balangida 42 Simanjiro East 49 Masai OA
29 Lake Natron GCA 36 Babati OA 43 Sanya Lelatema 50 Mkungunero
30 Longido GCA 37 Burunge 44 Ruvu Same 51 Kondoa OA
Selous / SE CoastaSouthern
52 Gonabis / Jukumu WMA 56 Kilombero GCA South 60 Tunduru Forest 63 Mahenge OA South
53 Liwale OA North 57 Namtumbo WMA 61 Tapika OA
54 Liwale OA South 58 Tunduru WMA
62 Kilwa OA North, Central & South
55 Kilombero GCA North 59 Sasawara Forest
60
based wildlife management currently remains under the direct centralised control of the state -
contrary to the substance of the new wildlife policy.
The process that led to the development of the new wildlife policy during the 1990s was heavily
mediated by a coalition of donors, yet the WMA regulations which eventually emerged were largely
the product of the Wildlife Division, with relatively little donor involvement. The result was that
Wildlife Division has retained its control over wildlife, particularly in regard to hunting revenues,
unlike the Forestry and Beekeeping Divisions that have devolved extensive control over forest
resources in an innovative way to local communities. In part, the relative revenue generating capacity
of the forest and wildlife sectors underpins the difference. Logging of upland natural watershed
forests was banned in 1977, and while the forest sector remains a high value resource (not least in
terms of fuelwood, bee products and plantation lumber resources used by most Tanzanians), the
wildlife sector constitutes a far more elitist as well as immediate and easily captured form of high
value revenue for the state (particularly for the Wildlife Division).
Land-use planning is a major component of the formal WMA development process. In reality, the
land-use planning process means that large areas of land will continue to be set aside for wildlife -
often exclusively. There is a strong danger, as is already happening in parts of northern Tanzania,
that a large proportion of village members, particularly the poorest who are likely to depend most
on farming, will loose access to land ear-marked for WMAs. Also, farmers cultivating land adjoining
those areas ear-marked for WMAs will increasingly have to contend with crop damage by wildlife.
The decision to participate in a WMA rests with the village council. However, village councils remain
democratically weak and poorly accountable to their constituent village assemblies (see Chapter 3).
Often the wider interests of the village may be disregarded in favour of the interests of the council
and its members. In the absence of careful support and intervention by appropriate organisations, this
may be particularly the case for the WMA creation process. While WMAs may lead to relatively
substantial increases in village council incomes and benefits for communities, the danger exists that this
income will be to the wider detriment of people’s farming and herding livelihoods.
Nelson et al. (2003, 47) aptly sum up the current situation in regard to the large wildlife estate that
has been set aside in Tanzania for the benefit of a few when they state:
The political economy of over 110 years of wildlife law in Tanzania has
overwhelmingly functioned to increase the centralisation of wildlife resources over time,
and any attempts at devolution must confront this legacy. In Tanzania, the partnership
between donors and western conservation NGOs and central government has been
decidedly ineffective in achieving these institutional changes. Overcoming this legacy
and achieving genuine local empowerment for wildlife management is critically
important in terms of both sustainable management of the country’s biodiversity as well
as for local livelihoods and land rights.
Thus, rural herders and farmers continue to make choices about the management of their rangelands
heavily structured and supervised by the state, with insufficient power to negotiate or express their
underlying objectives and livelihood priorities. In most wildlands of Tanzania, they continue to derive
61
little or no benefit from the wild resources around them relative to the substantial revenues received
by the state, while bearing the significant costs of crop loss from wildlife.
3.2.4 New Tanzanian land tenure legislation and its implications for herders and farmers
The nature and legacy of land tenure laws in Tanzania are extremely complex (e.g. Wily 1988)62.
However, a major trend is very clear. From the outset of the colonial period, customary land rights
and practices have, one way or another, been redefined and increasingly overridden and
extinguished by the pre- and post-colonial state. Shivji (1998, 15) concludes that by the end of
villagisation what little was left in the security of deemed (customary) rights derived from the
country’s original land legislation (the colonial Land Ordinance of 1923) had been destroyed. Instead
villagers were apprehensive of loosing even more of their land - a process which the Presidential
Inquiry into Land Matters of 199263 documented in twenty volumes over its two year investigation.
Shivji (1998, 16) describes the legal framework of village land tenure in the late 1980s as utterly
confused64. In fact, days before the Commission submitted its findings to the government, Parliament
passed a law which extinguished all customary land rights for many villages in the Tanzania65 (Lane
1996, 170).
The findings of the Presidential Commission of Inquiry into Land Matters of 1992 were supposed to
have formed the basis for the reform of Tanzania’s land tenure laws. Unfortunately the state was to
disagree with key parts of the Commission’s recommendations on how land could be more equitably
held and democratically administered for the benefit of all Tanzanians. Perhaps most importantly, the
Commission’s recommendations aimed to stop the long-term appropriation of land from ordinary
Tanzanians by the state and other interests through: establishing the legal inalienability of village
land to non-village members; endowing villages with full land rights to manage their lands free of
interference from the state; strengthening the accountability and equitability of land management
processes at village level based on customary land laws and practices; setting up local land circuit
courts to better facilitate village-level land dispute resolution; transferring radical land title66 from
62 Wily (1988) provides a detailed review of the political economy of land tenure in Tanzania between 1891
and 1988.
63 The Presidential Commission of Inquiry into Land Matters of 1992 was set up to review land matters in
Tanzania as the basis for developing new land tenure legislation. The Commission toured the whole of Tanzania
and consulted with a very wide cross-section of Tanzanian society.
64 Shivji makes this description on the basis of the inconsistent and contradictory opinions of the judges who
presided over a major land alienation case brought by the Barabaig against the State (Shivji 1998, 16; see
also Section 2.2.5 and Chapter Six for further details on the appropriation of land from the Barabaig)
65 The law, the Regulation of Land Tenure (Established Villages) Act of 1992, was passed as a reaction by the
state to put an end to the legal proceedings brought by the Barabaig to have their appropriated land returned
(Lane 1996, 169-170).
66 Radical title is taken here as meaning the President of the United Republic of Tanzania holds all land in trust
on behalf of the nation and therefore exercises ultimate ownership and control of all land in Tanzania.
62
the President and diversifying it into national land institutions accountable to parliament, and; making
land a constitutional category.
Perhaps the most significant and important objection on the part of the State was that the
Commission’s recommendations, if implemented, would stop the appropriation of land by the State for
promoting external investment projects - for example, those in the natural resource sector such as
commercial farming, ranching and wildlife enterprise (see Sundet 1997, 109). In essence, as the
State’s reply to the Commission’s report put it, the Commission’s recommendations would make the
state a beggar to the villages and the proposed national land board (Shivji 1998, 81). The State
considered the prospect of having to consult and respect the wishes of villagers and the decisions of a
national land board unacceptable.
A new Land Policy (passed in 1995) was developed by the State with very little further public
consultation in relation to how the Commission’s recommendations could form the basis of the new
policy (Sundet 1997; Shivji 1998). The Land Act (1999) and Village Land Act (1999), together with a
first amendment in 200467, were subsequently developed and enacted68. The development and
consultation process for these acts, given their significance for Tanzanians’ livelihoods and well-being,
was totally inadequate69.
The Acts are long, complex and very bureaucratic70, but their combined thrust can be simplified to the
following (see Shivji 1998, 111-118). In essence the new land laws are designed to ensure that the
on-going neo-liberalisation reforms, advocated by international financial institutions (IFIs) and now
much favoured by the State, proceed regardless of the interests of ordinary Tanzanians - particularly
67 This is the Land (Amendment Act) of 2004 which strengthens the alienability of land and the transfer of
derivative land rights to citizens or non-citizens for investment projects approved under the Tanzania Investment
Act of 1997. The amendment also replaces previous sections of the Land Act governing mortgages. Arguably,
mortgages have failed in Africa for last century and are likely to continue failing (e.g. Shipton and Goheen
1992, 317). Put simply, European and American mortgage models ignore agronomic, economic, social and
cultural realities that make them inappropriate for an African setting (see Shipton and Goheen 1992, 317-318).
68 The land laws came into force in May 2001 with the passage of 16 sets of supporting regulations, which are
not yet widely available.
69 The development process for the new land laws was highly elitist and only included key government officials,
international consultants, and representatives from international financial institutions (IFIs) and bilateral donors
(see Shivji 1998).
70 For example, 50 different forms are to be filled in at one point or another by ‘Village Land Officers’ (see
Sundet’s 2005 useful critique of the Land Acts (especially the Village Land Act). Also, the District Land Officer is
stipulated as having to approve all customary right of occupancy applications which may run into their thousands
for a single district. All villagers who currently occupy land under customary right of occupancy must have their
land claim adjudicated before a certificate of customary title can be awarded them. Given the millions of land
plots in Tanzania and their associated on-going dispositions, what is in effect a land titling process will be a very
long term and cumbersome undertaking. Resting the power of approval with the District Land Officer will lead -
as it has done in the past - to partial decisions influenced by patronage.
63
rural people. Although villages are provided with the right to manage and administer their land
(designated ‘Village Land’), village rights are subordinate and defer to the State. Thus the President
and the State may relatively easily transfer village land to ‘General Land71’ or ‘Reserved Land72’
without the agreement of villagers73. Either of these latter land categories can then, one way or
another, be allocated to an outsider, such as an investor, as the State sees fit. Village land can also
be leased to outsiders74, and the Land Commissioner can intervene if a village council refuses a lease
application. Parastatal and other organisations can obtain customary title to village land, and given
the weak state of village governance in much of Tanzania, this can only lead to the continuation of
land grabbing. In this regard, it is understood that district councils have been encouraged to set-up
‘land-banks’75, in part by identifying village land for potential transfer to general land for the
purposes of commercial investment and enterprise76. Rangeland that is seasonally used by pastoralists
(and therefore seasonally ‘empty’) is particularly at risk, as wetter rangelands are often arable and
potentially viable for commercial agriculture77.
Lastly, the Courts (Land Disputes Settlements) Act 2002 was passed to address the fact that the new
land acts had failed to provide a viable and affordable land dispute resolution mechanism above
that of village councils. The draft strategic plan for the implementation of the land laws (URT 2005)
71 This land category is under the direct control of the state through the Commissioner of Lands.
72 Reserved land comprises wildlife, forest and marine protected areas as well as other land as defined in
section 6 (1) of the Land Act (1999).
73 An important clause in the new land laws is the contestable right of the state to transfer village land to
general land for allocation to outsiders. This arises from a contradiction in the land laws in the definition of
‘General Land’ given in the Village Land Act and that given in the Land Act. (In the former it is defined as, ‘all
public land which is not reserved land’, and in the latter it is defined as, ‘all public land which is not reserved
land or village land and includes unoccupied or unused village land’.) The contradiction strongly lends itself to
the possibility that villages can lose land (Wily, 2003). Furthermore, the state can transfer large amounts of
village land to general land by deeming it in the public’s interest. ‘Public interest’ is defined in sections three and
four of the Land Acquisition Act (1967) as basically anything the President deems, not least large-scale private
investor-driven enterprise. While the village council is ostensibly legally enabled to decide upon land areas
below 250 hectares, the Minister (at the direction of the President) has the right to transfer areas larger than
250 hectares to General Land.
74 As per the Land (Amendment Act) 2004 – see above. When exploited by the wrong hands, this amendment is
likely to lead to expropriation of large tracts of village land with serious lack of transparency.
75 For example, see section 19 of the United States Department of State’s investor guide to Tanzania
http://www.state.gov/e/eb/ifd/2005/42185.htm (accessed 25th February 2005). Also an article in the
Business Times of Tanzania, ‘Land Bank scheme pulls in agro-investments’ on 15th January 2005.
76 This is consistent with the Agricultural Sector Development Strategy (2001, 4) which states in italics, ‘The
Government will work towards creating an enabling environment for medium and large-scale investors to make
use of the abundant land resource in the country’.
77 Loss of dry season grazing land can severely disrupt pastoral rangeland management systems, leading to
unseasonable pressure on the remaining range and increases in livestock morbidity resulting from nutritional
stress. Pastoralist livelihoods suffer as a consequence.
64
although recognising the importance of land-use conflicts and disputes, pays scant attention to
developing the capacity and supporting the development of village land councils and ward tribunals.
Although a cross-sectoral issue, this is a particularly significant inadequacy given the growing
occurrence of land disputes in rural Tanzania. Moreover, the system’s design appears to remain
focussed on formalising land rights and systems of managing resource conflicts within state institutions,
in spite of extensive evidence pointing towards the importance of maintaining an interplay between
‘formal’ and ‘informal’ systems (Maganga 2003, 66; see Chapter Seven).
In summary, there are some strongly beneficial aspects to the new land legislation - for example, the
recognition of local or customary forms of land tenure78 and the move to locate land management in
rural areas at village level. Moreover, the Village Land Act provides for land-sharing agreements
between pastoralists and agriculturalists, although the process through which these land-sharing
agreements are to be developed are heavily and adjudicatory in design. Village land title is to be
vested in the village assembly, and village land managers, accountable to their village council and
assembly, will be given responsibility for village land management. Yet at the same time, the new
laws are heavily overshadowed by their failure to incorporate some of the most important
recommendations of the Presidential Commission of Inquiry into Land Matters.
3.2.5 People cascades: landlessness, insecurity and wanderings

In relation to past and recent developments in Tanzanian land and natural resource policy and law,
there is growing evidence pointing towards a trend of increasingly heavy internal migration in
Tanzania, and growing land insecurity and landlessness (Odgaard, 2002). In this regard, many semi-
arid rangelands and particularly dryland-wetland areas in Tanzania are today home to a polyethnic
assemblage of farmers and herders. Most immigrant people have arrived in these areas seeking new
land and livelihoods. More often than not, they have been compelled to leave their natal areas as a
result of long-term declines in the availability and access to land and natural resources. Growing
landlessness among farmers and herders has been caused or further exacerbated by an increasing
and already extensive protected area system and an expansion in commercial agriculture - such as
the Canadian Wheat Project in Hanang District (Lane 1996; Niamir-Fuller et al. 1994; see below).
Often this land has been critical to the viability of customary resource managements systems - such as
dry season grazing key for pastoralists.
Perhaps the most marginalised groups of rural Tanzanians who have been impacted by these
contemporary land-use developments are pastoralist groups, such as the Ilparakuyo and the
Barabaig. Their societies and production systems have often been viewed, particularly by the state
and international development institutions, as primitive and backward (Coulson 1982, 161). Thus,
apart from the poorly conceived and implemented USAID livestock development project in the 1970s
(Moris 1981; Arhem 1985a; Hodgson 2001; see Section 4.2.2), pastoralism has tended to be
78 On the condition that local land tenure practices do not contradict state law – particularly in regard to the
land rights of marginalised or vulnerable groups in society.
65
disregarded and marginalized. Instead state policies and land-use planning have favoured other
production systems, such as commercial agriculture, the wildlife sector and small-scale farming.
In the last 50 years, there has been an increasing southward movement of herders, particularly
Ilparakuyo79, Barabaig80 and Sukuma peoples in search of new pastures and land. Pastoralists and
agro-pastoralists have had to accept the risks of moving their stock through tsetse infested rangelands
in search of more favourable tsetse-free pasture and water (Galaty 1988, 168). Aided by
techniques such as moving through tsetse-infested areas at night when the tsetse fly rests up, and using
modern anti-trypanosomal81 drugs, pastoralists are able to move their herds with much reduced risk
over long distances (Mtemisika pers com). The reasons underling the movements of these herding
groups are similar but different, and therefore they are individually described as follows.
The migrations of the Ilparakuyo82 have been much longer-term in comparison to other herder
movements. The origins of the Ilparakuyo movements southward can be traced to the Maasai Iloikop
wars of the nineteenth century (Galaty 1993, 172-182). Having lost the wars to their Kisongo Maasai
adversaries, the Ilparakuyo began to move south-eastwards into parts of Pangani, Bagamoyo,
Dodoma83 and Morogoro (e.g. Galaty 1993, 179-182). More recently over the last 50 years, the
Ilparakuyo have continued their movement southwards, entering the Kilombero valley and also settling
to the southwest in the rangelands of lowland Iringa District - in particular, Ruaha Mbuyuni, Ismani,
Pawaga and Idodi (pers com Mgemaa, Mtemisika and Katei; Odgaard 2002). They arrived in
Usangu in Mbeya District in south-western Tanzania in 1953 (Charnley 1994, 85). During Ujamaa, the
Ilparakuyo were subjected to land evictions in some of the Districts in which they live. In Bagamoyo
District, the state alienated about 61,000 acres of land from Ilparakuyo herders and 18,000 acres
from cultivators (Ndagala 1974 & 1986 cited in Rigby 1992, 153). While the farmers were
compensated with other land, the Ilparakuyo were not. In recent years, as farming and herding
populations have grown, the Ilparakuyo have increasingly come into conflict with farmers over land
(e.g. Odgaard 2002), even in areas in which they have resided for many years, as occurred in
December 2000 in Kilosa District, central Tanzania.
While retaining extensive kin-networks over hundreds of kilometres, many Ilparakuyo have become
increasingly sedentary as they have moved into agro-pastoralism and are increasingly tied to their
79 The orthography adopted for Ilparakuyo terms and words follows that of Rigby (1983 & 1985).
80 There are two recognised spellings: – ‘Barabaig’ (in common usage – for example, Huntingford [1953]; Klima
[1970]; Lane [1996]; Wilson [1953]) and ‘Barbayiig’ (after Rekdal and Blystad [1999]). The orthography
adopted for all Barabaig terms follows that used by Lane (1996).
81 ‘Samarine’ is favoured, for example, by many herders (Galaty 1988; pers obs).
82 Rigby (1992, 152) relates the Ilparakuyo migrations more generally to extensive and prolonged land
alienations (particularly of dry season grazing range [Maa: isupuko pl.; osupuko sing.]) suffered by the Maasai
in East Africa during the pre- and post-independence periods.
83 The semi-arid plains of Dodoma - the domain of the Gogo - appear to have been historically frequented by
the Maasai - not least the Ilparakuyo - who raided into this area in the 19th century (Rigby 1985).
66
fields. However, their growing agro-pastoralism does not preclude sporadic migrations to other
distant areas (especially to locations where they have kin) if the local socio-ecological and political
environment becomes unfavourable (pers com Mgemaa, Mtemisika and Katei). Nevertheless, as the
most productive semi-arid rangelands - in terms of pasture and farming - fill up with people,
continued movement is likely to become an increasingly difficult option given the time and investment
required to build the social relations necessary for securing adequate access to land and landed
resources.
The Barabaig migration is of more recent origin. Although there have been long term losses of
Barabaig land in northern Tanzania to their northern Iraqw neighbours, the expansion of the tsetse fly
belt (see Chapter Six) and even historically to the Maasai, it is the recent loss of dry season pastures
which have most impacted on the Barabaig (Lane 1998, 152-4). These muhajega (Barabaig: wet
season forage regime) dry season grazing areas were lost through state-enforced land alienations
for commercial wheat farming84 in the Basotu Plains during the 1970s (Lane 1998, 151). Groups of
Barabaig who could no longer access sufficient grazing moved southwards seeking pasture through
Singida, and then via Tabora towards the Rukwa valley in the far southwest of the country. Others
moved south into central Tanzania via Dodoma, Ismani, Pawaga, Idodi and the Usangu (e.g. Lengisugi
2000). Some Barabaig herders have settled into more localised seasonal range use patterns in
particular localities, for example in the lowlands of northern Iringa. These families have increasingly
integrated themselves socio-economically with locally resident farmers and herders. Others Barabaig
herders remain heavily transhumant moving substantial distances as pasture needs and unit security85
dictate.
For several decades now, Sukuma and Nyamwezi agro-pastoralists have moved through Tanzania
from their homelands in Shinyanga in northwest Tanzania (Brockington 2004, 1). As part of Sukuma
cultural ecology and a longer term ‘expansionism’ (Brandström 1985; Charnley 1994 & 1997;
Galaty 1988), this movement initially began as Sukuma agro-pastoralists entered the rangelands of
Maswa, Shinyanga and Geita (Brandström 1985, 20-21 cited in Galaty 1988, 168) as well into
Nyamwezi86. Cotton production was introduced during the German colonial period, and began to be
expanded from the 1930s onwards, as a result of the colonial drive for increased agricultural
exports (Iliffe 1979, 348-349). The growth in cotton production by the Sukuma continued over the
next 40 years such that by the end of the late1970s, 90-95 per cent of all cotton produced in
Tanzania came from Sukumaland (Hankins 1974 cited in Charnley 1997, 606). While the Sukuma re-
84 The muhajega range areas in Hanang District were identified as being particularly suitable for mechanised
farming of wheat by a Canadian Wheat Project that was subsequently started in the 1970s (Lane 1996, 151).
85 These groups of migrant Barabaig are often the subject of much complaint by locally resident farmers and
herders and conflict is common. In such instances, when the situation becomes untenable, especially after an out
break of conflict, a Barabaig ged (Barabaig: unit) may move out of the area swiftly to avoid further conflict
and/or the intervention of the state.
86 Nyamwezi is the domain of a people with the same language. The Nyamwezi are largely agriculturalists - in
contrast to the agro-pastoralist Sukuma (Galaty 1988, 169).
67
invested their profits in livestock, less and less grazing land remained available for their growing
herds of livestock87, particularly as the Shinyanga rangelands had been increasingly converted to
cotton fields (Birley 1982 and Brandström 1985 cited in Charnley 1997, 607). This process is thought
to have precipitated the southwards expansion of the Sukuma and their herds into Tabora and
Chunya Districts in western Tanzania (Galaty 1988, 169). Today they are successfully established in
many of the semi-arid rangelands of Tanzania - for example Pawaga, the Kilombero valley the
Usangu plains and the Rukwa valley in the southwest of Tanzania.
Over the last 50 years migrations into the semi-arid rangelands of Tanzania have also occurred
among farming populations. The underlying reasons for these migrations are not well understood, but
evidence points towards processes of growing landlessness and possible declines in soil fertility in
some parts of Tanzania. It has been established that the introduction of new perennial cash crops to
local farmers sometimes led to changes in customary land tenure practices and inheritance patterns.
For example, in the Nyakyusa highlands of south-western Tanzania, the introduction of coffee led to
growing landlessness among a younger generation of men as their elders evaded giving up valuable
land earmarked for coffee growing (Gulliver 1958 and Odgaard 1986 cited in Charnley 1997,
598). Landlessness was further compounded during the colonial period in Nyakyusa by the creation
of forest reserves, alienation of land for European farming and a growing population (Charnley
1997, 600). Charnley (1997, 601) posits that young landless men instead chose to move to the
lowlands of the Usangu (Mbeya), where highly fertile and uncultivated soils could be farmed. Similar
processes of land loss or declines in soil infertility in other parts of the southern highlands of Ubena
and Uhehe may have led to the movement of further groups into the lowlands. Thus today’s lowland
farming communities living in the Usangu are polyethnic and at least half of the total population
consists of Hehe, Bena, and Safwa and Wanji immigrants (Charnley 1994, Mwakipesile 1976 and
Walsh 1984 cited in Charnley 1997, 598). A similarly ethnically diverse farming community lives in
the Idodi and Pawaga rangelands of Iringa District (see Chapter Five) 88.
In summary, over the last 50 years, key wetland areas in the semi-arid rangelands have become a
focus of heightened competition for land and water, as farmers and herders converge on these
centres - or ‘frontiers’ - of relatively high fertility and productivity. Most significantly, much of the in-
migration of herders and farmers into the semi-arid rangelands has been caused by the alienation of
large areas for wildlife conservation, commercial agriculture and ranching schemes, and the
expansion of small holder cash-cropping. Today, there are increasing shortages of arable land and
pasture in these dryland-wetland areas. This is leading to growing levels of competition over land
and natural resources, and outbreaks of violent conflict between herders and farmers.
87 Between 1944 and the mid-1960s, the livestock population of central Sukumaland doubled from an
estimated 1,728,400 animal units to 3,360,000 animal units (Charnley 1997, 607).
88 Nahonyo et al. (1998) identify 35 ethnic groups as living in the Idodi and Pawaga area.
68
3.3 Conclusion
The political and socio-ecological control of people initiated during the colonial period was taken to
an extreme by the modern Tanzanian state during the Ujamaa years. The colonial and then the
modern state restructured the semi-arid rangelands through moving people into settlement
concentrations and alienating large areas of land for the creation of a protected area network,
commercial plantations and ranches. The centralising control of the state can be seen as having been
driven by a number of factors. These included its need to: capture the production of rural herders and
farmers in order to maintain and further develop an export-led economy; achieve greater political
and economic control over a dispersed farming and herding population, and; improve its ecological
and political control over the semi-arid rangelands in order to increase the now substantial revenue to
be derived from wildlife tourism. The combined effect of these policies and developments has been
increasing shortages and insecurity in land and landed resources, and increasing competition and
conflict over these resources. The long record of state appropriate of land together with other related
factors has led to substantial movements of herders and farmers across Tanzania, who have been
compelled to search out new fertile land, grazing and water.
From the mid 1980s, with economic collapse threatening, the Tanzanian state was forced to adopt
neo-liberal reforms to its economy and political ideology. In the early 1990s, as it re-introduced
multi-party politics, the state began to slowly modify some of its systems of governance. In the late
1990s it embarked on the reform of local government, in part to be achieved through decentralising
service provision and natural resource management. The decentralisation project is widely viewed as
a key strategy for Tanzania’s development. In natural resource management, decentralisation has
been more fully supported and developed for some sectors than in others. The forestry sector has set
about devolving partial and full rights to villages to manage and benefit from the country’s extensive
protected forest area. In comparison, the wildlife sector has resisted decentralisation, essentially
retaining full control of wildlife, arguably in deference to powerful interests and centralised flows of
substantial revenues to the central state.
A land reform process started in the early 1990s failed to sufficiently build on a number of bold and
highly consultative recommendations of a Presidential Commission of Inquiry into Land Matters.
Amongst other findings, the Commission had found that many Tanzanians had been denied the right
to equitably administer their own land, were chronically land insecure, and that the State together
with other interests continued to appropriate large amounts of land. In not taking the Commission’s
recommendations fully in hand, the new land laws have not sufficiently addressed the insecure land
rights of many Tanzanians. While a new system of village-based land administration has been
created, it is extremely bureaucratic and fails to provide sufficient safeguards for the continuing
inequality in greater access to land by local and other elites at the disadvantage of poorer and
more marginalised groups. To date, little progress has been made at village level in developing the
capacity and systems of accountability required for promoting equitable, efficient and participative
village-based land administration and land-based dispute resolution.
69
While much focus has been given to the reform of government, piloting decentralisation into the
districts (e.g. Shivji and Maina Peter 2000, 35), and finalising land reform legislation, insufficient
consideration has been given to understanding how local people at village level, and in the context
of this thesis, farmers and herders, will be able to respond and adapt to these reforms. In the semi-
arid rangelands, as human populations grow and migrate through the landscape, and as demand
increases for key seasonal rangeland resources (arable land, fertility, water and pasture), the
likelihood for land and natural resource based conflict may be expected to increase. A longstanding
challenge for local government, land and natural resource reform is the need to understand and
support the development of equitable local management systems and natural resource conflict
management practices. This process is likely to remain under-realised and problematic, undermined
by conflicting interests of control, and handicapped by understandings and attitudes within central
and local government that remain entrenched in well established, but now critically discredited
knowledge.
70
4 The dryland-wetland frontier of Idodi and Pawaga
4.1 Introduction
The rangelands of Idodi and Pawaga are located in northern Iringa District in south-central Tanzania.
They were chosen for the site of the field study as they provide an example of a ‘frontier’ rangeland
(see Chapter Three) which has been subject to successive local resettlements and wider in-migrations
of farmers and herders from neighbouring districts, and from further afield. In the past, both Idodi
and Pawaga have also been heavily impacted by colonial land-use management and protected
area creation, as well as more recently by post-independence Ujamaa policies and state wildlife
management practices.
In this chapter I introduce the Idodi rangelands and their people, and provide an overview of their
agro-ecology and socio-economy. I then provide an account, with the aide of selected vignettes, of
the nature and functioning of village government and jural systems (the latter are important in land
dispute resolution). I focus my attention on village government since it has the most impact on people’s
daily lives and also has considerable influence and control on how village lands (see Section 3.4.3)
are managed. This leads into a discussion about current modes of land tenure in Idodi for farmers
and herders, particularly in relation to the recently introduced new national land laws. Finally, I
discuss the position of pastoralists in relation to local government power structures centred on the
farming community.
4.2 The Greater Ruaha Ecosystem: Idodi and Pawaga
The southern extent of the greater Ruaha ecosystem lies in northern Iringa District, central Tanzania. It
covers approximately 30,000km2 of different designated rangeland-use areas - the Ruaha National
Park, the Rungwa, Kisigio and Muhesi Game Reserves, the Lunda-Mkwambi Game Controlled Area89
as well as overlapping village lands (see Figure 4.1). The area is internationally significant in terms of
biodiversity conservation in that it contains the only protected area system covering the transition
between the vegetation communities of the Sudanian Acacia-Commiphora zone of East Africa and the
Brachystegia (miombo) woodlands of southern Africa90. The area is of substantial national economic
importance for a number of reasons: It comprises the greater part of the upper catchment of the
Ruaha, Njombe and Kisigo Rivers which feed the Mtera Dam, one of Tanzania's most important
hydro-electric schemes; the Kisigo and Rungwa Game Reserves together with parts of the Lunda-
Mkwambi Game Controlled Area North (LMGCAN) are prime tourist hunting blocks generating
89 The southern extent of the Game Controlled Area (GCA) may become the MBOMIPA Village Association
WMA (see Section 3.4.1).
90 Bjørnstad (1976) and Nahonyo (1998) provide useful reviews of the vegetation of the Ruaha National Park
and the Lunda-Mkwambi Game Control Area immediately to the south.
71
Figure 4.1: The location of the study site in Tanzania - the Idodi villages, Iringa District
KEY:
LMGCAS
LMGCAN
Ruaha Nat Park
Muhesi GR
Kisigio GR
Rungwa GR
Usangu GR
THE IDODI
VILLAGES
KEY:
RUAHA NATIONAL PARK
1 - Mahuninga
PAWAGA DIVISION
2 - Makifu
3 - Tungamalenga
LUNDA-MKWAMBI GAME 4 - Mapogoro
CONTROLLED AREA 5 - Idodi
LUNDA
6 - Malinzanga
6
7 - Nyamahana
7
4
3 Field site
MKUPULE 5
2
IDODI DIVISION Ruaha National Park
1 Usangu Game Reserve

USANGU GAME RESERVE Lunda-Mkwambi Game Controlled Area North
10 km2 Lunda-Mkwambi Game Controlled Area South
Rungwa Game
Village boundaries
Major roads
Map based on: 1:50,000 topographical maps (YT42 Series; date 1982; UTM Zone 37S; Arc 1960),
Department of Surveys and Mapping, Dar es Salaam;
Boundaries & roads taken from Ruaha Ecosystem & Wildlife Management Project GIS data
Major rivers
72
relatively substantial revenues for the state; the Ruaha National Park is becoming an increasingly
developed and important tourist attraction, and; the Lunda-Mkwambi Game Controlled Area South
(LMGCAS) (see Figure 4.1), while currently of lower wildlife significance, constitutes part of an
important rangeland for pastoralists (mostly around its periphery), an increasingly extensive irrigated
agricultural area for farmers and a pilot area for MBOMIPA, a community-based wildlife
management development.
The LMGCAS, an area of 1,850km2, straddles the northern half of the two administrative divisions of
Idodi and Pawaga in Iringa District (Figures 3.2 and 3.3) and is home to over 70,800 people91
(National Population and Housing Census [NHPC] 2002), comprising of 8 majority ethnic groups92
(Nahonyo et al. 1998). The area lies on a gently undulating plain, stretching from the base of the Rift
Valley escarpment northwards to the Great Ruaha River (see Figure 4.1). To the east of the LMGCAS
lies the confluence of the Great and Little Ruaha Rivers around which lies the natural seasonal
wetland of Pawaga. To the west a range of hills reduces the breadth of the plain in a bottleneck with
the Great Ruaha. During the 1980s, the seasonal wetland in Pawaga was developed for more
intensive irrigated agriculture. Today it is the focus of seasonal and more permanent in-migration of
farmers from elsewhere (many originating from the southern highlands) who cultivate rice paddy
during the seasonal river spate in the wet season. The Pawaga wetland also forms a dry-season
grazing area for Ilparakuyo, Barabaig93 and Sukuma pastoralists who move some of their herds south
across the Great Ruaha once wet season pastures in the National Ruaha Park and LMGCAN become
depleted. The overall land-cover of Idodi and Pawaga consists of relatively open woodland and
bush mosaic in the northern extent of the LMGCAS, and a greater extent of agricultural lands in the
southern periphery, particularly along water courses and in wetland areas (see Figure A2).
4.3 Idodi and Ikwavila valley
Idodi lies in the northwest of Iringa District and, since it borders the Usangu plains of Mbarali district,
it forms a pastoralist migration corridor between the Usangu in the southwest and the Pawaga and
Ismani rangelands to the northwest. Iringa, the District centre, lies some 90km to the east in the
highlands. Idodi division has three wards, Mahuninga, Idodi and Mlowa94, with a combined
population of just under 20,991 people (NHPC 2002). The Ikwavila Valley, which was the focus for
field work, lies to the south western end of the inhabited part of Idodi division (see Figure 4.4). Over
9,000 people live in the Ikwavila valley (see Table 4.1). Over the last 20 years, the population in the
91 The 1978 census recorded the total Idodi and Pawaga populations as being 27,000 people. The 1967
census cannot be used in the time series as administrative areas underwent substantial revision in the intervening
period (Nahonyo et al., 1998:19). In 1996, Iringa District Council estimated the population of Idodi and Pawaga
to be 46,000 people (Iringa District Council 1996), at the time likely to have been a substantial underestimate.
92 Hehe, Nyamwezi, Bena, Kinga, Gogo, Sagara, Sangu and Wanji
93 The events leading to the presence of the Barabaig in Idodi are discussed in Chapter Six.
94 Mlowa ward (consisting of Malinzanga and Nyamahana villages) was not part of the field study, although it
is mentioned occasionally in the following chapters.
73
Figure 4.2: The villages of Idodi, herder homesteads and land-use areas
Idodi
Mapogoro
Tungamalenga
Mt. Chaminya
Makifu
Mahove
Mahuninga
Legend:
Ilparakuyo homesteads Scattered dryland fields
Village settlements Sparse & isolated dryland fields
Mixed wetland & dryland fields LMGCA

Land leased to expatriate farmer Settlement and use disallowed
Main road - unsurfaced Track - unsurfaced
Water course Contour
0 5 10 kms
Scale:
Map adapted from: 1982 Series Y742 Sheet 214/3 Edition 1TSD Surveys and Mapping Division, Ministry of Lands, Housing and
Urban Development, United Republic of Tanzania. Agricultural land cover overlay modified from FAO-AGROMET 2002 data –
Notes: (i) Malinzanga and Nyamahana villages are also in Idodi Division and lie to the north-east lies of the map; (ii)
Land-use areas are approximate and not definitive; (iii) Unshaded areas are a mixture of Acacia-Commiphora
bushland at lower altitudes, Brachystegia woodland at slightly higher altitudes, and treed grassland in other areas;
(iv) Cultivation and livestock are permitted by law in GCAs. In LMGCA, farming and herding have been ‘banned’ for a
large extent of Mkupule and Lunda, and these areas are often referred to as ‘hifadhi’ (Swahili: Reserve) - as depicted
in the map (see also Figure 4.1).
74
Table 4.1: The population of Idodi by village in 2000
Data collected by village governments [Source: Report from Divisional Secretary to District Commissioner, dated
14/12/2000, loose un-filed report]
Data taken from the National Housing and Population Census online (full report remains unpublished) [Source:
http://www.tanzania.go.tz/sensa/districts/iringarural.htm; accessed 29/30/04]
Average Total Total

household population number of
Ward Village People Households
sizea by ward households
by wardb
Idodi 3,598 810 4.44

10,286 1,735
Idodi Mapogoro 3,208 452 7.10 c
(9,205) (2,290)
Tungamalenga 3,480* 473 7.36
Mahuninga 3,638* 750 4.85 5,802 1,008

Mahuninga
Makifu 2,164* 258 8.39 (4,040) (1,085)
Totals 16,088 2,743 5.87 (15,007) (2,743)
* Villages defined in this study as lying in the Ikwavila valley.

a
The reason for the variation in average household sizes is not known - but may relate to variations in how
households were defined by different officials in each village. National census data show the average household size
of Idodi and Mahuninga wards as being 4.0 and 3.7 persons respectively. More disaggregated national census data at
village level are not available.
b
When compared to the census data, total ward population numbers given by village governments for Idodi and
Mahuninga wards are respectively 12 and 44 per cent higher.
c
Data in brackets are from the National census and provided here for comparative purposes.
Table 4.2: Population growth in the Idodi villages between 1976 and 2000
[Sources: Report from Idodi Ward to the Divisional Secretary in 1976 (File
IDO/A40/11); Report from Mahuninga Ward to the Divisional Secretary dated
11/07/1976 (File IDO/A40/11, Ref: H/R 12/2/27)]
People Households Annual

population
Village
1976 2000 1976 2000 growth
rate
Idodi 1,933 3,598 350 810 3.58%
Mapogoro 1,874 3,208 330 452 3.89%
Tungamalenga 2,145 3,480 350 473 4.11%
Mahuninga 1,472 3,638 430 750 2.70%
Makifu 899 2,164 255 258 4.15%
Total 8,323 16,088 1,710 2,743 3.69%
75
Figure 4.3: The Lunda-Mkwambi Game Controlled Area looking west from
Idelemule Mountain in Tungamalenga in the dry season
Figure 4.4: The Ikwavila valley looking south towards Mahuninga from
Tungamalenga in the dry season
76
valley has doubled (see Table 4.2) with an average growth rate of 3.7 per cent. Two of the villages
in the Ikwavila valley - Makifu and Tungamalenga - have relatively the highest growth rates. This is
almost certainly due to continuing in-migration of farmers (see Chapter Five).
4.3.2 The agro-ecology of Idodi and the Ikwavila valley

The rangelands of Idodi have a semi-arid climate with an average annual rainfall of about 388-
960mm per year95. Rainfall is usually unimodal and falls between the months of December and April,
with a short dry period lasting most of February. Very occasionally, rain may fall in September or
October. Local people say that a decade ago and more, rain fell in the form of short showers during
September, but that it no longer does. Rain during September is a useful respite to the dryness and
heat of the mid-dry season and is beneficial for vynungu (Hehe: moist low-lying riverbank gardens)
dry season crops that are irrigated by streams and irrigation furrows. Rainfall can vary considerably
between years. It is often the case that about one year in every four is a poor year of rain, with
another year usually receiving a better than average amount, and one year in 10 being particularly
dry. Such trends are not hard and fast and longer term climatic cycles may impact on inter-annual
rainfall patterns - for example, many local people in Idodi consider that much of the 1990s were
locally drier than the preceding decade. During the dry season, daytime temperatures as well as
evapotranspiration rates are high and thus seasonal crop cultivation outwith irrigated
areas is not possible. The growing season is therefore of limited duration but variable depending on
any one year’s rainfall, ranging from about 60 to 120 days.
The Ikwavila valley is bounded to the south and west by a steep range of hills, and to the east, by
the gently rolling foot-scarps of the highlands (see Figures 3.2 and 3.4). The mouth of the valley is
broad and marked by Idelemule Mountain to the west and by the impressive Chamyina Mountain
(see Figure 4.2) rising nearly 1000 metres from the valley floor in the east. The Ikwavila valley is
watered by seasonal streams draining from the hills to the west and perennial ones to the east from
the highland massif (except in the driest of years). The streams combine to form the Tungamalenga
River96 at Iliango (Hehe: first bridge) and the Mazombe River (a tributary of the Tungamalenga
River), which then flows northwards to meet with the Great Ruaha River some 40km downstream.
95 No long term data is known to exist for rainfall in Idodi – the data quoted here relate to information
collected 30km to the northwest between 1995 and 2000 and include an exceptional year of rainfall during
the El Niño Southern Oscillation (ENSO) of 1998 (960mm of rainfall). If this event is excluded, rainfall ranged
between 388 and 527 millimetres per year during this period. This data is consistent with the national rainfall
map shown in Figure 3.2.
96 The Tungamalenga River is remembered as always having been perennial but, in recent years, it has not
always continued to flow during the latter part of the dry season (although there is a substantial sub-surface
flow). It is suspected that this trend is a result of increasing levels of dry season irrigation and the diversion of
stream water onto cropland cultivated with cash crops such as rice, tomatoes and most recently, paprika.
77
Prior to the in-migrations of people during the second half of the twentieth century, much of the valley
was a mixture of Acacia-Commiphora bushland and Combretum woodland. The area is remembered
as being frequented by large wildlife which declined as the valley became more populated and
cleared for cultivation. Today, although Mahove remains a relatively extensive drier, higher, area of
woodland and bush (see Figure 4.3), much of the valley has been cleared for cultivation. Pastoralists
seasonally graze their livestock in the areas of uncleared woodland and bush, except for parts that
are particularly prone to tsetse fly infestation (in the west). The woodlands above the villages also
contain wildlife, although a lesser amount than the Acacia-Commiphora bushland and Combretum
woodlands stretching out northwards and westwards towards the Great Ruaha River below the
villages.
The intensively farmed valley bottoms are characterised by rich brown soils (supplemented by alluvial
sediments from the highlands) of clays and loams (see MAFS 2002). The valley bottoms form the
mabonde (Swahili: pl. low-lying ground; sing. bonde) in which farmers have constructed a network of
irrigation canals and furrows. Irrigated rice and some maize is grown in the wider, more extensive,
low lying flatland ‘mabonde’ while in the narrower stream margins - the ‘vynungu’ - perennial and
dry-season food crops such as plantain and cassava are cultivated. These eutrophic bonde soils are
regarded as the most fertile and are where the highest value crops are planted. The higher ground,
away from the mabonde, is called ‘nchi kavu’ (Swahili: dryland) and the soil here is generally loamier,
less fertile and with a much higher incidence of heavily sandy and dystrophic soils. Lower value crops
- mainly maize and peanuts - are generally grown in mono-stands on this less fertile land. The two
different bonde and nchi kavu agro-ecological systems underpin substantially different agricultural
production and livelihood strategies for Idodi’s farmers (see Chapter Five). A further, underlying,
character of the land is the heterogeneity of soils and soil conditions even within relatively small
areas, not only in a spatial context, but also in regard to intra- and inter-annual variations in soil
productivity and nutrient status. This heterogeneity is the result of a complex interplay between soil
content, structure, status, nutrient release processes and varying rainfall regimes (Scoones 2001, 22).
This variability underpins the high degree of uncertainty and risk faced by farmers in semi-arid
dryland agriculture (see Chapter Five).
4.3.3 The people and socio-economy of Idodi and the Ikwavila valley
Idodi is today home to a diversity of people of different origins97. There are a small number of Hehe
kingroups who have lived in the valley since before the arrival of the first German colonialists in the
nineteenth century. These Hehe kingroups may be viewed as autochthonous ‘hosts’ who have received
different types of ‘newcomer’ over the years. Many Hehe farmers arrived from the Ruaha River
valley as a result of protected area creation and Ujamaa resettlement during the 1960s and 1970s.
Others arrived from the Hehe highlands in search of new land. These farmers, although in a sense
‘newcomers’, can almost be described as autochthonous as they identified themselves as Hehe and
97 More detailed accounts of the arrival of different people in Idodi over the last 50 or more years are
provided in Chapter Five for farmers and Chapter Six for herders.
78
were relatively easily assimilated by their hosts (see Chapter Five). During the 1970s and 1980s
other non-Hehe ‘newcomer’ strangers (such as Bena, Kinga, Wanji and other farming groups) arrived
from much further afield to farm in Idodi. Over the years, these non-Hehe farming ‘newcomers’ have
become accepted by their local Hehe hosts, although they may be viewed as still - albeit varyingly -
remaining ‘strangers’. While ‘newcomer’ farmers have been either assimilated or accepted by their
Hehe hosts, the experience of arriving herders has been different. Thus although the Ilparakuyo have
been resident in the landscape since the 1950s, their status has remained much more strongly that of
‘stranger’ as compared to non-Hehe farming ‘newcomers’. Despite developing increasingly strong
labour, land and exchange relations with their ‘hosts’ and other farming ‘newcomers’, the Ilparakuyo
have remained excluded from participation in village government, and are far less accepted than
other more recent farming ‘newcomers’. Finally, the ‘newcomer’ Barabaig - as a collective group - are
the least accepted ‘strangers’. To a certain extent, this is because they often remain highly mobile in
the landscape (with exception), and have generally invested the least in socio-economic relations with
the host communities and their ‘newcomer’ associates.
For the purposes of analytical clarity and ease of description, from hereon the term ‘farmer’ is used
to refer collectively to Bantu Hehe, Bena, Wanji and other groups of farmers, despite the fact that
some households in some of these groups may keep a limited amount of livestock. The term ‘herder’ is
used to refer to Nilotic Ilparakuyo and Barabaig pastoralists, both of whom may farm to a greater
(Ilparakuyo) or lesser (Barabaig) extent. The socio-political and socio-ecological factors underpinning
the use of these labels will become increasingly clear in the thesis.
Most farmers today live in villages along Idodi’s major roads. In addition, there are a number of
smaller vitongoji (Swahili: hamlets; sing. kitongoji) in out-lying areas, often long established
settlements that survived or were re-established after the villagisation era. The villagisation
programme of the 1970s substantially disrupted previous settlement patterns as people were evicted
from their hamlets and homesteads and forced to live in collectivised villages. People were
supposedly allocated one acre plots on which to build their houses during resettlement. In practice,
many people found themselves building much closer together leaving little room for household
expansion or for the well-tended gardens envisaged by central TANU party officials. Although not
exclusively, in more recent post-Ujamaa vitongoji, extended family and kin can be found living in
relatively close proximity to one another in loose kin groups, as people have been able to choose
where they live. This is especially the case, for example, for a group of Bena living in Makifu.
Originally from Lupembe in Ubena, these families arrived during and immediately after the
villagisation period to farm Makifu’s fertile wetlands.
Most people live in traditional thatched-roof houses, although increasingly thatch is being replaced
with corrugated galvanised iron sheeting. Wealthier households are today building fired-brick and
mortar houses. Each village has a market place, a primary school, usually a dispensary and at least
one beer club. People’s fields are often some distance away from their homesteads and the village
settlement, and individual households may own a number of different plots of land located in
different parts of the outlying village area. Households may also have a small plot of land farmed as
79
part of the homestead area. In contrast, herders live away from the main village settlements and
agricultural areas, but within relatively short walking distance. Their homesteads are often located in
the woodland periphery and include an enclosure for their livestock. Herders (both men and women)
may spend varying amounts of their time in the village settlements trading, shopping and socialising.
The socio-economy of people living in Idodi is primarily centred on farming, mostly of maize and rice,
and to a lesser extent, peanuts and sorghum (see Chapter Five). Rice is an important cash crop for
those able to access suitable land and afford cultivating it, and it can frequently be very profitable,
particularly depending on when the crop is sold in the agricultural calendar. There are a few large
rice growers, who have become comparatively wealthy, gradually expanding the area of rice they
farm. These larger farmers have additionally expanded their businesses through buying cheap rice
locally from other rice farmers immediately after harvest, often in repayment of cash loans given at
the start of the agricultural calendar. Young men, migrate to Idodi for the rice-growing season from
their homes (often in the highlands), to farm rice by renting fields from locally resident farmers. Rice
growing, as an extremely labour-intensive crop, is generally farmed by households and individuals
able to hire labour, or physically able to carry out the labour themselves. Most of the rice grown in
Idodi is eventually sold in Iringa town to millers as mpunga (Swahili: husked rice) who then may de-
husk the rice and trade it on the national market.
Although rice is the highest value large-scale crop in Idodi, maize is the most important subsistence
crop, in terms of both total acreage and the proportion of households who grow the crop. Most
households will grow at least some maize, as the staple food. Many households need to sell some of
the maize they have grown immediately after its harvest for a cash income, but often the value of
maize at harvest, like rice, is at its lowest in the agricultural price cycle, unless the harvest is early98.
Maize is also used for beer making by women, and is sold locally. Mbege (Hehe: maize beer) is an
important part of recreational and cultural life in Idodi for both farmers and herders, and is
additionally seasonally supplemented by Ulanzi (Hehe: bamboo wine) brought down by vendors from
the highlands during the early dry season months of April to August.
98 This is explained in more detail in Chapter Five – agricultural commodity ‘farm-gate’ prices are generally at
their lowest during harvest, and rise to their highest just before the onset of the next harvest.
80
Table 4.3: Reported livestock numbers in Idodi Division in 1984 and 1999
[Sources: Livestock Census 1984 (File /DC/V.10/1/70 dated 08/05/1987); Report from Divisional Secretary to District
Commissioner, 1999]
Cattle Goats Sheep Donkeys

Village
1984 1999 1984 1999 1984 1999 1984 1999
Idodi 572 1200 174 500 51 150 1 29
Mapogoro 1338 250 302 120 149 90 2 10
Tungamalenga 141 230 127 110 6 70 0 13
Mahuninga 46 550 116 115 116 70 17 8
Makifu 517 320 29 100 29 52 0 7
Total 2614 2550 748 945 351 432 20 67
Note: The data are almost certainly an under-estimate of the livestock present in Idodi as it is likely that livestock
keepers will have evaded enumeration to avoid paying livestock tax. For example, the total Ilparakuyo herd for 2001
was 2253 cattle (see Table 6.3). No field data are available for Barabaig herds which may number perhaps as much
again as the Ilparakuyo herd in Idodi (albeit given relatively large seasonal variations as a result of transhumance).
While farming is central to Idodi’s economy, herding also plays an important, albeit lesser, role.
Ilparakuyo agro-pastoralists and Barabaig pastoralists are the major livestock keepers in Idodi,
although a minority of farmers - particularly wealthier Bena farmers, may also keep some livestock.
Village-based records suggest that livestock numbers in the Idodi villages have remained stable (see
Table 4.3). If this is the case, then per capita livestock holdings over the same period are likely to
have declined with an increasing herder population. In addition to the role that livestock plays in the
socio-economy of individual agro-pastoralist and pastoralist households (see Chapter Six), livestock is
important for district tax revenues and individual herders tend to pay proportionately far more in tax
than farmers do. Both pay district development tax (levied per adult person) but herders also have to
pay livestock tax99. Livestock products - particularly meat - are sold in village markets, especially on
market days. Farmers may additionally benefit from livestock as they rent their harvested fields to
herders during the dry season for crop residue grazing. Although field grazing can be a mutually
beneficial arrangement for both farmer and herder, it is often the cause or catalyst of farmer-herder
tension and conflict, increasingly common in recent years. For example, in the last ten to fifteen years
references to farmer-herder dispute occur quite often in village records, but prior to 1990 there are
very few. Dispute particularly occurs when livestock stray into neighbouring fields. This type of dispute
is becoming increasingly symbolic of socio-political struggles between herders and farmers over
access to land and security of land tenure for herders, discussed in further detail in Chapter Seven.
99 In 2001, district development tax was TShs 4,000 per person. Livestock tax was levied at TShs 500 per head
of large stock (cattle and donkeys), and TShs 200 per head of small stock (goats and sheep). While farmers
may pay agricultural cess taxes, they do so only on the produce that they sell in bulk (e.g. in 80kg sacks) which
is transported out of Idodi.
81
4.4 Village-based administration and land tenure
4.4.1 Village governments in Idodi

The aim of this section is not to provide a comprehensive account of the workings of local and village
government (see Appendix A1 for a short overview of local government structure), but through
vignettes and focussed descriptions, to briefly transect the contemporary nature and functioning of
village government in the Idodi rangelands to illustrate its impact on peoples’ livelihoods and land
relations (discussed in Chapters Five, Six and Seven).
The village council is the key institution of village government. In Idodi, with occasional exception, all
village council members and the village chairman are members of the ruling CCM party who have
been elected through municipal elections, which occur every five years. Council members then elect
village council office holders for the village council’s main committees which are responsible for
coordinating village affairs and services100. In the Idodi villages, although the village councils meet
quite regularly (about once a month and sometimes more), it is the village chairman101 and the village
executive officer (VEO)102, together with the chairs of the three major committees of each village who
tend to hold most of the day to day influence in village administration.
Village government in the Idodi villages is often constrained by low levels of capacity and standards
of governance. For example, may be a tendency for office holders, particularly the village chairman
and VEO, to take advantage of their positions and village councils are often unable or unwilling to
hold them totally accountable. The village assemblies tend to expect a certain amount of impropriety
as the norm, but as long as the affairs of the village run relatively smoothly, improper practices are
tolerated, albeit grudgingly. There may be other, sometimes more important, socio-political reasons
as to why a particular office-holder or chairman holds their position. These may revolve around
kinship, identity, patronage and (factional) village and ward politics. However, in relative terms,
large corruption scandals may result in the dismissal of the village chairman by the village council (this
occurred in Tungamalenga in 2002).
In addition the every-day world of politicking also impacts on village government. In this regard, it is
sometimes the case that one or other village council office holders loose their position not so much
from an abuse of power as due to a political power struggle between factional groups. For example,
a long-standing and effective ward councillor lost his seat, not because he was particularly corrupt,
but because it is thought the local MP wanted to shore-up his political powerbase. The MP sponsored
his uncle in the elections, who was able to use his local identity and his nephew’s influence to
subsequently win the election. Finally, there is a high turn over of council-employed VEOs and WEOs
100 The three main committees at village level usually are the planning and finance committee, the defence and
security committee, and the social services committee.
101 During fieldwork no women held the position of village chairperson in the Ikwavila valley.
102 The VEO is employed by the district council as an administrative clerk for the village.
82
(Ward Executive Officers) as these positions are poorly paid, working and living conditions are at
best basic, and as a result impropriety can be quite common. For example, between 2000 and 2002,
VEOs and WEOs left or resigned in at least three villages and two wards in Idodi.
Village governments in Idodi have low levels of financial capacity -in terms of revenue and the ability
to manage what little revenue they receive for supporting the public needs of the village assembly.
The nature of financial constraint that village governments face on a day to day basis can be
illustrated by the following vignette from Tungamalenga. The water system that supplies the village is
a very well designed and constructed low maintenance gravity-fed system, which has one simple
flaw. During the wet season, occasional large thunderstorms in the highlands cause the river feeding
the village’s water system to flood. When this happens, the water intake often becomes blocked
(about four times a year). Given the importance of clean water for everyone, it might be expected
that it would be a simple matter for the village council to pay the local water fundi (Swahili: handy-
man or technician) the required TShs 2,000103 to cycle two hours to the intake to clean it out.
However, what usually occurred when the intake became blocked, was that the village would often
go for days without water, because the VEO claimed there was no money in the village water
‘account’ to pay the fundi. Instead, the VEO would go about asking the wealthier members of the
village for the required money. They would make their excuses as they had already paid a premium
for piped water to their houses. No money would be forthcoming until the water tanks in one or more
of the wealthier village member’s households had run out, at which point it would be made available.
However, during the days that it took the VEO to find the needed funds, the stand-pipes that most
households in the village depended on for their water would remain dry. Instead women would have
to walk two to three kilometres to the nearby river to draw and use dirty river water. The shortage of
village funds might have been more understandable had it not been for the fact that every year
many of the village’s 473 households pay a water charge of TShs 200. On being asked by some
irate villagers what had happened to the water funds, the VEO was unable or unwilling to explain
where the money was or what it might have been spent on.
This is not to say that the money had been stolen or used up for council ‘allowances’, although this
remains a possibility. Village revenues are limited, and are mostly derived from licensing local
trading, from levying charges - such as the court and water charges, and in some instances collecting
land rent104. The district council is supposed to return ten per cent of the district development tax
annually collected in the villages back to their respective village councils. However this had yet to
happen in the Idodi villages, despite repeated requests by village councils for the money over a
number of years105. Revenue from a local community-based wildlife management project
103 In 2000, one US dollar was equivalent to 800 Tanzania Shillings (TShs).
104 Land rents are more frequently applied in Pawaga than in Idodi. However in the Idodi villages, particularly
in Tungamalenga, village governments are starting to collect rents from non-village members who have leased
land from the village (pers com Chengula).
105 For example, on 17th July 2001, in a heated argument in front of the District Commissioner, a village
chairman angrily complained that the district council was demanding that the village council open a bank
83
(MBOMIPA106) is the single biggest source of village funds and even this does not amount to very
much107. Thus village governments are chronically short of funds, leading to incongruities such as a
water shortage in the middle of the rainy season. This situation is compounded by frequent financial
irregularities and chronically low levels of administrative capability.
In recent years, the Idodi village councils have become much more directly involved in natural
resource management issues, particularly with regard to wildlife. In the mid 1990s, as part of the
‘community-based conservation’ component of the Ruaha Ecosystem and Wildlife Management Project
(REWMP)108, community wildlife management committees were set up in the Idodi villages. The
supposed focus of these committees was subsequently widened when they were re-designated
community natural resource management committees under MBOMIPA. The MBOMIPA committees -
essentially this is what they are - can be seen as fulfilling two roles in the villages. Firstly, they
organise and supply a quota of scouts for joint patrols (with other MBOMIPA villages from Pawaga)
that are carried out in the Lunda and Mkupule sections of the LMGCA. The function of these patrols is
primarily to deter and apprehend illegal village-based subsistence and other hunters109, and to stop
unlicensed resource use such as timber harvesting. Secondly, the MBOMIPA committees sometimes act
as a land-use militia, particularly as the MBOMIPA committee is part of the village defence and
security committee. Thus MBOMIPA scouts are occasionally used to intervene in land-use disputes and
their presence may be used to ensure or sometimes enforce the resettlement of herders from one
particular area of a village to another.
account first before the funds would be made available. The village chairman argued that the village did not
have sufficient money for the minimum opening balance (TShs 50-100,000 depending on the bank), and that the
district’s requirement was just an attempt at evading payment. To be fair, local government rules require that
the money be paid into a bank account for accounting reasons, although the district council could have been less
obstructive and more helpful in enabling poorer villages to open the required bank account.
106 Matumizi Bora ya Malihai ya Idodi na Pawaga (Kiswahili: Sustainable Use of the Wildlife Resources in Idodi
and Pawaga) – a joint Department for International Development (UK) (DfID)/Wildlife Division / Tanzania
National Parks (TANAPA) project which worked towards developing community-based wildlife management in
the Lunda and Mkupule parts of the LMGCA. The project has since closed, but a MBOMIPA Village Association
consisting of 19 villages continues to operate and is trying to complete the application process for re-gazetting
part of LMGCA as a WMA.
107 In 2002, each village received about TShs 438,000 which is equivalent to about TShs 141 (US$ 0.175) per
person for the villages of Tungamalenga, Makifu and Mahuninga.
108 This TANAPA/WD/ODA (Overseas Development Administration - now DfID) project ran between 1993 and
1996. It had two components – strengthening the infrastructure and management of the Ruaha National Park
(partly in response to the wildlife crisis at the end of the 1980s discussed in Chapter 2) and developing
community-based conservation in LMGCA as part of piloting the new wildlife policy. The latter component was
to be further developed by MBOMIPA, the successor project.
109 Arguably, by far the most damaging and unsustainable hunting is carried out by a hunting company from the
Usangu Game Reserve. This hunting company has connections with senior Tanzanian politicians and, amongst
other things, a reputation for irresponsible and highly improper hunting practices. Given this situation, MBOMIPA
scouts have been powerless to intervene.
84
The village council is also responsible for making village bylaws - for example, bylaws about
domestic water usage in the village, or prohibited natural resource uses in certain parts of village
lands. Bylaws can become white elephants, particularly those that are instigated and passed by the
district council, but which have little support at village level. Certain land-use bylaws, such as those
banning livestock grazing of field crop residues rented from farmers by herders, are a case in point
(see Chapter Seven). Also, villagers’ knowledge and observance of district bylaws are often minimal,
except when the bylaws are applied by local district council employees - such as in the case of
agricultural commodity charges, or frequently flouted livestock movement restrictions. Bylaws
declared by the village council may receive popular support from villagers, and after a period of
occasional enforcement by the village militia and prosecution of offenders, become respected by all.
An example is the banning of livestock watering in the upper reaches of local water catchments to
prevent the pollution of the village water system. However, not all bylaws are official in the sense
that they have been approved and passed by the district council according to stipulated procedure.
This is because the approval process for bylaws may take many months, and the district council may
return a set of proposed bylaws to a village council for modification, further delaying the process. In
general, although bylaws do play a role in regulating village affairs and land-use management, they
amount to more of a loose framework upon which, by choice or compulsion, people in a particular
context may or may not structure their socio-political and socio-ecological relations (see Chapter 6).
Finally, the past dictatorial tendencies of local government continue to occur from time to time. In part,
this is a result of the continued existence of autocratic government at divisional level, but district
councils may also be equally implicated. For example in December 2001, it was announced by the
Idodi village government that each household had to farm four acres of land (irrespective of how
much land a household owned or could afford to rent, assuming land was available to rent). Two
acres were to be cultivated with subsistence crops, and two with cash crops. Failure to farm the
stipulated four acres would result in internment in the local village lock-up. Three farmers were
subsequently arrested in Idodi village for ignoring the decree, interned and then subsequently
released. After a while, the order lapsed into ignominy as farmers totally disregarded it and as the
village government gave up pretending that it could enforce it.
4.4.2 Local level courts

Local level courts are an important part of land-use dispute resolution processes in Idodi. I provide a
very short overview of the structure and state of local government and other jural institutions in the
Idodi villages as they exist today110 as background material for a more extended discussion on land-
use relations and land conflict in Chapter Seven.
110 In essence the system of land courts - village land councils, ward tribunals and district land and housing
tribunals - set out in the new land laws are similar to current courts and tribunals already in place.
85
The Baraza la Makahama ya Kijiji (Swahili: village tribunal) is the first court level in which all minor
disputes and crimes (such as minor land or water disputes or avoidance of village levies) are heard.
The ruling of the Baraza la Makahama ya Kijiji may be appealed against and the case referred to
the Baraza la Makahama ya Kata (Swahili: ward tribunal) - which hears more serious cases, for
example those including land-use conflict, protracted domestic disputes, public disorder and
inheritance disputes. The Baraza la Mahkama ya Kata and Baraza la Mahkama ya Kijiji are convened
on a regular basis, and hear most local cases that do not involve serious crime or do not directly
involve the state111. The primary court112, which is the next level of appeal, is located in Idodi village
and barely functions, as the magistrate only visits on a monthly basis. As a result the Baraza la
Mahkama ya Kata may handle cases which the primary court magistrate might ordinarily hear,
particularly since people are reluctant to travel to the primary court at Kalenga where they have
been directed to go instead. This is because having a case referred to the primary court at Kalenga
or, even more inconveniently the district court in Iringa, is an expensive undertaking in terms of travel,
other costs and time - particularly if a case is being heard during the wet season when most people
are heavily occupied farming.
A further aspect of local level legal systems is that there are different systems of customary law in
existence - in addition to the more formal village government jural process. Thus whereas Hehe
customary law is more closely integrated with village government jural process113, Ilparakuyo and
Barabaig customary law remains separate. In particular pastoralists, engage in parallel systems of
jural process - often using their customary law and jural institutions to mediate and resolve dispute
that occurs internally within their own societies, and engaging as necessary with formal government-
based jural process when seeking to resolve disputes with others - for example, with Hehe farmers
(further discussed in Chapter 6).
111 State-prosecuted cases tend to be brought by national parks or the Wildlife Division’s regional anti-
poaching unit for illegal resource use within Ruaha National Park or the LMGCA respectively.
112 Tanzania has five levels of court - the judiciary court of appeals, high courts, resident magistrate courts,
district courts, and primary courts. However, for the great majority of rural Tanzanians, most civil disputes and
petty criminal cases are heard at ward or village level. The Primary Court in Idodi has been without a full time
magistrate since 1987 (when its previous incumbent retired) despite repeated requests for a replacement by the
Divisional Secretary. The court is in a state of disrepair with the roof missing its ridge allowing rain to cascade
in. Hundreds of case files (somewhat insect infested) are stacked in such a way as to avoid the worst of the rain
that pours in.
113 According to Magistrates Court Act of 1984, primary courts are supposed to have not less than two court
assessors who preside over cases together with the magistrate to enable plural jurisprudence. The assessors are
members of the area which the court serves, and their role is to apply interpretations of customary law in cases
that come before the primary court (Maganga 2003, 64). The ethnic identity of the assessors underlines the
balance of different interpretations of customary law and the relative roles of individual agency and institutions
in the application of the law (Maganga 2003, 64).
86
4.4.3 Land tenure categories and practices in Idodi
The contemporary Idodi landscape is divided into variously overlapping land tenure categories, in
terms of rights of occupancy and permitted land-use zones (see Table 4.4). In the Idodi rangelands,
there are two overarching categories of land as set out in the new land legislation: Reserved Land
87
Table 4.4: Land tenure categories in the Idodi villages
Land Category1 Sub-Category Rights and characteristics Example
RESERVED LANDS National parks, marine parks, game Large areas of land set aside by the state. Entry and/or userights are Ruaha National Park, Rungwa Game
(National) reserves, forest reserves, declared mostly extremely restricted. Areas mostly managed entirely by the state Reserve, Usangu Game Reserve
2
watersheds and hazardous lands (jointly managed forest reserves being an exception).
VILLAGE LAND Land owned by Hehe kin groups who have lived in Idodi for a very long time A relatively large piece of land in the
- some families (e.g. Chambulila) for over 100 years. Some land has been centre of Mahuninga village that
4
ª Indigenous customary land given to extended kin as they arrived in Idodi to settle. This land category remains uncultivated but belongs to
continues to be governed by contemporary Hehe customary law and Mzee Chambulila
practice.
Land allocated by the village government, particularly to newly arriving Many of the farm fields in Idodi are
farmers, for example during Ujamaa. Once allocated, this land category is allocated land, owned both by farmers
then governed by contemporary customary law and practice as understood and herders.
ª Allocated land4
Farmland3 by the respective land owner (e.g. Hehe/Bena/Wanji). Village government
very rarely intervenes further (except in case of disputes which may be
resolved through village and ward courts). Also includes group farmland7.
Land (either allocated or indigenous customary land) that has been An increasing number of rice fields are
purchased by a villager or, relatively less often, non-village member. The being bought by wealthier farmers in
ª Purchased land4
purchase may often be recorded on paper and stamped and witnessed by the villages due to their value
the village chairman and/or VEO as a deed of sale / ownership.
Land (in one of the above categories) that is rented for a set fee and Both maize and rice fields are rented
period, or borrowed for a season or more. This practice is quite common and borrowed
ª Rented or borrowed land4
and people from neighbouring and more distant villages may rent land -
particularly rice fields.
88
Table 4.4 cont. Land tenure categories in the Idodi villages
Land that is not perceived as belonging to anyone, but which is Village Most of the lower lying uncultivated
Land. Usually permission must be obtained from the village council before land lying away from the village
the land can be cleared for farming - or settled. Herders (Ilparakuyo and settlements. Herders have been made
ª Open commonage5
Barabaig) rely on this land for their homesteads and livestock grazing, but to move increasingly into the lower
their rights are often no more than those of squatters. This land is often reaches of this land (see Chapter
the most marginal land (i.e. its agricultural potential is limited). Seven).
Village and that has been declared to be off-limits for settlement, The upper catchment areas in Idodi are
agriculture and livestock grazing by the village council. Minor non-timber reserved by the village councils to
3
Bush/Forest forest use often continues - such as firewood and thatch collection. District protect water sources. Also much of the
and other authorities (e.g. TANAPA) may play a substantial role in LMGCA8 in Idodi is de facto ‘reserved
influencing the location and creation of village reserved land. village land’, in which most use has
6
ª Reserved land (Village) been stopped - largely through the
agency of the Wildlife Division and
TANAPA. In recent years, increasingly
supported by village councils due to
income from MBOMIPA with hopes for
further wildlife-derived revenue.
Land that has been formally leased by the council to wealthy outsiders Previously leased land, re-leased to an
(native nationals, nationals of foreign extraction, and expatriates). Leases expatriate farmer for farming to supply
3 4
Farm/Bush/Forest ª Leased land are for varying periods, but not less than 33 years. The village council is his tourist lodge nearby the Park; two
supposed to have agreed to these leases which are often surveyed and campsites leased by Tanzanian
certified by the District Land Officer. nationals; other leased land.
89
Table 4.4 cont. Land tenure categories in the Idodi villages
GENERAL LAND General land Areas of land that are not Reserved Land or Village Land, and which fall No examples in Idodi. Nationally, land
under the direct control of the Commissioner for Lands. that falls between village boundaries,
e.g. in very remote areas.
Notes:
1. These categories correspond to the three major land categories set out in the Land Act (1999) and Village Land Act (1999).
2. These reserved areas are governed by separate sets of legislation e.g. the National Parks Act (1959), Wildlife Conservation Act (1974).
3. The different eco-types of land existing in the village.
4. This land is recognised and governed in the Village Land Act (1999) as ‘individual and family land’. It also includes residential plots of land in village settlements.
5. Open commonage is recognised and governed in the Village Land Act (1999) as ‘communal village land’.
6. Reserved village land is recognised and governed in the Village Land Act (1999) as ‘reserved land’ - not to be confused with reserved land set aside under the Land Act (1999). According to the law,
this land remains as village land and comprises spare land for future or individual use as decided by the village council. Any rights issued to individuals for this land can only be derivative, and cannot be
made permanent (i.e. the land continues to belong to the village assembly).
7. Farmland that was allocated in some of the villages for group/communal production, partly associated with the Ujamaa years. For example, in Tungamalenga village, plots of group farmland are still
owned by a youth group, the local branch of the Tanzania Women’s Union (Swahili: Umoja wa Wanawake wa Tanzania), a parents group, the Lutheran Church and the Roman Catholic Church.
8. A Game Controlled Area (GCA) comprises an area of village (or general) land in which wildlife use is prohibited without a licence from the Wildlife Division or District Wildlife Officers. In large parts
of LMGCA, many of the legal usufruct rights permitted in GCAs under the Wildlife Conservation Act (1974), such as settlement farming and grazing, have been extra-legally suppressed. The President of
Tanzania has the right, according to Section 19 of the Wildlife Conservation Act (1974), to declare specific human land-use practices prohibited in a particular GCA in the interests of wildlife
conservation. No notice to this effect for LMGCA has been published in the Government Gazette - as far as is known.
90
and Village Land (see Table 4.4). Village Land114 is managed by the village council which has the
power to allocate village land to members and non-members of the village115. Village members can
own land in one or more ways: as customary land (mostly inherited), as land allocated by the village
government, or as land purchased from another villager. In addition, there are a number of non-
village members who have leased farmland (and also bush/forest land) on a long-term basis from
the village council (see Table 4.4). Remaining Village Land comprises commonage upon which herders
depend for grazing and land for their homesteads, but for which they have yet to be able to obtain
sufficiently secure rights (see Table 4.4 and Section 4.6). The state has imposed restrictions on the use
of large parts of reserved village lands in the LMGCA (see Table 4.4, note 8).
The allocation of land by village councils may not always be accountable, nor decided in an
equitable manner nor made in consideration of the best interests of all the different village groups
such as pastoralists (see also, for example, Maganga 2003, 66). Land ownership matters are rarely
discussed in public meetings, unless the issue impacts on a significant number of villagers, who may
have lobbied the village council to have the issue discussed more openly. Some of the weaknesses of
village-based land management in the Idodi villages can perhaps be best illustrated through telling
two short stories.
The first story shows that irregular allocation of land is not new in the Idodi villages, and that there
are long-term conflicts both between farmers and also within villages over land ownership. In the late
1970s, Makifu village council leased eleven acres of some of the most fertile land in the Ikwavila
valley to an outsider (pers com Chengula). However, when the land was demarcated, twenty acres
instead of eleven were allegedly allocated to the individual in dubious circumstances. In recent years
this alleged misallocation of land has become an increasingly contentious issue as farmland is now in
short supply, particularly land as fertile as that leased to the outsider. As a result, the village council
has been pressured into holding a number of public meetings to discuss if the nine acres were
corruptly allocated as alleged and whether they should be returned to the village. The individual
remains intent on retaining his lease over the disputed nine acres of farmland as the land is locally
very valuable and productive. As far as is known, the individual has successfully managed to avoid
returning the disputed land back to the village for its redistribution. His claim has remained relatively
strong as some years previously he had foresightedly obtained a supporting lease document from the
district land office
114 While Village Lands are shown as being formally demarcated on district land-use planning maps, as far as
is known the village boundaries have never been formally surveyed. The boundaries remain disputed in several
instances and although such disputes are occasionally brought up in village assembly (public) meetings, most are
long-standing and remain unresolved..
115 The distinction between a villager member and non-member may be defined by their presence or absence in
the village household (tax) register.
91
A second story demonstrates how pastoralists’ lawful entitlement116 to own grazing land is often
cursorily treated by village governments, however long they have been members of the village. In
Tungamalenga Village, two Ilparakuyo families who had lived for seven years in an area about two
kilometres east of the main village settlement were forced to move to a new area. The land which
they had lived on had been designated by the village council, together with experts from a World
Bank irrigation scheme117, as suitable for small-scale irrigation expansion. Unfortunately, the area of
village commonage to which the Ilparakuyo families were sent had a number of drawbacks. Firstly, it
was a very narrow strip of land (only a few hundred meters wide). Secondly the area was on the
edge of thick Acacia-Commiphora bushland prone to tsetse fly that would likely result in an elevated
trypanosomaisis threat for their herds. Thirdly, they were not alone. A thousand acre farm lease (see
Figure 4.3) had been recently given by the village council to an expatriate farmer without the prior
knowledge of the two Ilparakuyo families. Although the farm lease was first established in the 1960s,
it had been long abandoned upon the unfortunate death of the then tenant. The new tenant was able
to persuade the village chairman, VEO, the ward councillor and the village council that he needed the
land to grow produce for a tourist hotel he was constructing. Thus, for an undisclosed sum, the lease
was renewed, resurveyed by the District Land Officer and a new title deed re-issued. It is likely that
the Ilparakuyo families will be forced to move again, particularly if the tenant starts to farm his
leased land118, because the Ilparakuyo are effectively squatters and thus have no rights of ownership
over the commonage that they live in and depend on for their livelihoods.
4.5 Pastoralists and village government in Idodi: Living on the edge or edging in?
Herders are now long established in Idodi and, although a minority, play a noticeable part in village
life and the socio-economy of the villages. Herders and farmers maintain trade and exchange
relations (for livestock and livestock products), engage in client-patron relations (herders may often
hire agricultural labour for their fields), and negotiate access to land and seasonal grazing
(harvested farm fields are often rented from farmers by herders). Herder homesteads - both
Ilparakuyo and Barabaig - are located on the margins of settled lands (see Figure 4.2). While this is
partly pragmatic in terms of minimising the incidence of dispute from livestock incursions into farm
fields, it is also reflective of herders’ status within the villages. Pastoralists are, to a lesser or greater
degree, part of an on-going farmer narrative that identifies them as ‘outsiders’.
116 The Village Land Act (1999) allows for group registration of land holding such that, for example, a family or
association can lawfully secure collective rights of customary occupancy over land. The law ascribes grazing
land equal status as that of farmland and makes it possible for pastoralists and agriculturalists to hold different
rights in the same land through ‘land sharing agreements’. Importantly, the law provides for collective pastoralist
land rights across different villages.
117 This irrigation scheme has since been completed (URT 2002). It covers an area of 350 hectares / 865 acres
(Keenja 2003, 14)
118 I am not suggesting here that the tenant would manoeuvre to have the pastoralist families evicted. Rather a
combination of circumstances will likely cause them to move – since they already consider the location not a
particularly good place to live.
92
The ‘outsider’ status of herders has meant - in part - that they continue to be denied sufficient rights to
land, as the story about the loss of land by two Ilparakuyo families demonstrates. An additional
underlying reason for this situation is that pastoralists are perceived by the Hehe majority as living on
land which is Hehe, land which at some future date may well be required for farming for future
generations - as has been raised in farmer - herder land disputes - (see Chapter Seven). This
distinction by local Hehe society is the focus of a recent paper by Odgaard (2002) who carried out
field work in Ismani - a rangeland that lies immediately east of Pawaga. As Odgaard (2002, 73-74)
describes it, ‘wenyeji’ (Swahili - indigenous, sing. mwenyeji) are people in Ismani who are recognised
by the majority as being ‘indigenous’, in contrast to ‘wageni’ (Swahili - guest, sing. mgeni), who are
‘visitors’ - not least pastoralists - with less than full rights to land, but who may nevertheless have been
resident in the area for many years. When Bena farmers first arrived in Idodi, they experienced
similar perceptions and treatment. However, as their numbers have grown, and as they have
intermarried with the Hehe, and given their much closer cultural and agricultural affinities, they are
far less a part of this distinction today as compared with pastoralists.
A further problem for herders is that their status has meant that their efforts to participate in village
government have been frustrated. For example, the Ilparakuyo l’aiguenoni (Maa: murran spokesman)
has attempted to participate on the Tungamalenga Village Council as a herder representative.
However, he was repeatedly not informed of village council meetings, or his views (on behalf of the
Ilparakuyo) were ignored. Thus despite their long-term established presence in the Idodi villages,
herders have remained on the periphery of village government in terms of their effective
participation in village council and its decisions - particularly those pertaining to land and natural
resource management.
Village governments often regard herders as an inconvenience since their land-use practices are little
understood and may directly interfere with longer term aspirations of farmers (see Chapter Five).
Herders’ seasonal mobility and migration are frequently perceived as a nuisance, and they are often
viewed as ‘backward’, ‘uncooperative’ and ‘belligerent’ towards authority. Moreover, herders often
seek and sometimes may manage to evade taxes for which they are liable.
Herders - and Ilparakuyo in particular - often see themselves as trying to participate in village
government so as to represent their pastoralist interests. They see themselves as longstanding
members of the village who are treated as a collective group as second-class citizens, with rights
commonly inferior to farmers’. Herders often see themselves as being marginalised from village
council meetings. Some herders have remarked that it is more effective, given that they are a
minority, to pursue their own individual networks of influence with village office holders. However,
such strategies may be disadvantageous in the long term.
There is a growing realisation among herders, as a collective group, that pursuing individual
strategies with village governments and individual farmers to secure access to key resources is
leading to their continued marginalisation in the landscape. They have missed opportunities to
participate in collective village land-use and planning decisions that have made way for new
93
developments - such as the rice irrigation scheme that has been extended in several of the Idodi
villages. These decisions have led to their homesteads being pushed further to the village periphery.
Recently herders have begun to form pastoralist committees in their villages - particularly in the last
three to four years, a process catalysed by recent events and almost demanded by local government
at district level. Each committee has appointed office holders, recognised by village and local
government, to better represent herders in village government, and thus more effectively engage in
village politics and administration. The need to form village-based herder committees has been
underpinned by a recent and substantial increase in out-breaks of land-use conflict between some
herders and farmers (see Chapter Seven).
Thus herders are gradually becoming more successful in engaging with village governance processes,
particularly through their newly formed representative committees, although these committees remain
relatively weak. Their office holders often lack the experience and skills needed to successfully carry
out the difficult leadership and negotiation roles required of them, and they may not be able to
always depend on the support of the wider polyethnic herder community.
94
5 The peoples of ‘Kwigongo’: the old and the new
There could hardly be a more marked difference between the valleys

lying either side of Idelemule Mountain and the landscapes that stretch
beyond them. To the west lies the Mudweka valley. Stretching far off into
the haze of the mid-morning blue, one looks on the magnificent panorama
of Mkupule, a brown, gently rolling carpet of woodland, grassland and
wetland from which rise hill range and mountain. Apart from the gentle
play of the wind in the dry tree branches, there is neither sound nor
movement. Eastwards, is an altogether different spectacle. The scarps of
the highlands bound one’s view, and down below in the valley of
Mahuninga, a tight sea of colours and shapes, of metallic glints and lush
greens, lozenged fields and winding paths, pattern the landscape. One
becomes aware of the day’s work going on below as a column of smoke
winds its way upward from the stubble of a harvested field. The faint but
unmistakeable sounds of village life are carried up - the distant clatter of
a mill grinding maize and the faint clanging of a far-off school bell.
5.1 Alienations and catastrophe: the loss of the old way of life
This chapter is the first of two chapters in which I examine people’s livelihoods and land use practices
in Kwigongo (Hehe: a term used for the lowlands of Idodi and Pawaga). In this chapter, I provide a
more detailed account of the current livelihoods and landscape occupancy of the farmers of Idodi,
and in particular, the Ikwavila valley. I begin the chapter by describing how and why the Ikwavila
valley came to be settled by different waves of immigrant farmers. I then move on to describe and
provide an explanation for people’s current farming livelihoods and the increasing socio-economic
differentiation between wetland and dryland farmers. The former represent an older way of farming
the land (in Idodi) while the latter represent a new face to farming, since wetland farming is far more
productive than dryland. Throughout this chapter I will demonstrate that many farming households
remain desperately poor, locked in a cycle of poor soil fertility and dryland crop yields, labour
shortage and poverty, with a minority less constrained by these factors due to their access to wetland
fields, and who have experienced improving fortune. While in theory, there remains plenty of new
dryland for farmers to clear, the area available for expanding cultivation is limited by the
topography of the valley, its vegetation, restrictions on further land clearance and wildlife crop
damage on the periphery of the field area.
5.1.1 The old hamlets and the way of life in early colonial times
Large parts of the Lunda-Mkwambi Game Controlled Area (LMGCA) are today a wildlife preserve
devoid of human settlement. At the turn of the twentieth century this area, which comprises the
95
southern bank of the Great Ruaha River and rangeland stretching southwards to the rift valley scarps,
was home to diverse119 but interrelated peoples who variously farmed, herded stock, hunted and
fished in small scattered settlements. These were mostly set along perennial stream courses with many
sited along the northern and southern banks of the Great Ruaha River (see Figure 5.1). Stretching
east and west were the contiguous areas of Njongomeru, Mdonya, Kayela, and Ilolo which were
bounded by the northern scarps of the Ruaha River Valley. Above the escarpment to the north lay a
number of hamlets including Igula. These northern hamlets formed the northern periphery of Uhehe
having been incorporated into the Hehe chiefdom in the late nineteenth century120.
The Hehe Chiefdom was founded and militarily much expanded from the small chiefdom of
Ng’uluhe121 by the most notable of the Muyinga (Hehe: pl Vayinga) chiefs, Munyigumba, and
subsequently his son Mkwawa, during the latter half of the nineteenth Century (Brown and Hutt 1935;
Redmayne 1964, 1968). Following the rout of the Hehe in 1894 by a German expeditionary force122
(culminating in the death of Mkwawa in 1896), the succession of the Hehe Chiefdom was effectively
suspended, until Mkwawa’s son, Sapi, was successfully reinstated in 1926 by the British colonial
administration under a system of indirect rule. During the intervening years, the Hehe were ruled by
upwards of 70 of vansagila (Hehe: headmen123; sing. munsagila) who were appointed across Uhehe
by the German colonial administration (Brown and Hutt 1935, 49).
119 In this area a number of major tribal groupings have variably overlapped and intermingled in the past
century or more – principally those of the Hehe to the south, the Gogo to the north-east, the Kimbu to the north-
west and the Sangu to the south-west.
120 According to Walsh (pers com) it is likely that the Hehe did not fight these northern - previously Kimbu and
Kosisamba - chiefdoms until 1892/3, when they capitulated. This period corresponds to similar attacks
elsewhere recorded by Nigmann (1908).
121 Ng’uluhe was one of several small chiefdoms that existed during this period in what is now largely central
Uhehe and Iringa District.
122 Redmayne (1968) provides a succinct account of the war campaigns of Mkwawa and, in particular, of the
Hehe-German war that led to the demise of Mkwawa and Hehe military and socio-political hegemony.
123 This is more a corrupted version of the term that was adopted by the German Colonial Administration. A
more correct translation would be that of ‘sub-chief’ each of who had been accountable to the Muyinga
Paramount Chief. The vansagila were the appointed successors of the numerous chiefs (Hehe: mutwa sing; vatwa
pl) who had ruled small chiefdoms that had been subsumed by the then expanding Hehe Chiefdom of the mid to
late 19th century (Brown and Hutt 1935, 59).
96
Figure 5.1: The Ruaha River valley during the British colonial period circa 1950
Ilolo
RUNGWA GAME RESERVE

Kimande
Igula
Kayela PAWAGA
Mdonya
IDODI
Njongumeru Idodi
IRINGA
10Km2
MBEYA
Wasa KIPONZELO
USANGU
Key:
KIPONZELO - Sub-chiefdom boundary M B E Y A - District boundary
Idodi - jumbeate - Old settlement / homesteads - Major river
Key to old hamlets: [Note that the locations are indicative. Refer to Appendix A3 for further detail on each]
1 - Mwa Mlele 15b - Kigoga 31 – Mugaga 46 – Kiganga 62 – Mpanga 79 – Magangamatitu
2 - Chemchem 16 – Mkwata 32 – Nyatele 47 – Matinga 63 – Lungwa 80 – Magoya
ya Magombwa 17 – Makuluga 33 – Ihwani 48 – Msembe 64 – Igula 81 - Sasamambo
3 - Magombwe 18 – Ilala simba 34 – Makutupa 49 – Mdonya 65 – Kangaro 82 – Mloa
4 - Makutupa 19 – Ilusi (S) 35 – Ilolo 50 – Nyamtupa 66 – Mandwa 83 – Msimbi
5 - Chamguite 20 – Nyawagulu 36 – Talangwe 51 – Nyaluhanga 67 – Mtovelisusi 84 – Idodi
6 - Mdeya 21 - Wheluka 37 – Udindamisi 52 – Waga 68 – Makutupa 85 – Nyangano
7 - Igawa 22 – Uwalinzota 38 – Ny’angai 53 – Kitagasa 69 – Ilusi (N) 86 – Kitisi
8 - Mwaliganzi 23 – Kihanga 39 – Mkombilenga 54 – Sanzala 70 – Mukombe 87 – Mapogoro
9 - Nyirenge 24 – Kisoloka 40 – Unynyamala 55 – Kinyangesi 71 – Itunundu 88 – Kidangwe
10 – Ilkelekembe 25 – Cheleganza 41 – Mpangile 56 – Mkupule 72 – Igohungula 89 – Tungamalenga
11 – Uzimoto 26 – Kikulwe 42 – Unylinge 57 – Njongomeru 73 – Mbagi 90 – Isanga
12 – Chauumu 27 – Mgondale 43 – Ibagi 58 – Mpama 75 – Kimande 91 – Makifu
13 – Mwagusi 28 – Kimaling’oko 44 – Mtumbulikwakwa 59 – Lwani 76 – Ndolela 92 – Ikwavila
14 – Ifugulu 29 – Ikorongo 45a – Nyamakala 60 – Wota 77 – Mlengi 93 – Nykapembe
15a – Ididimoto 30 - Ilunda 45b - Fihwawi 61 - Mudweka 78 – Chimamba 94 – Mwitikira
[Note: This map is a work-in-progress. Some of the settlements listed were abandoned prior to 1950,
others had not been established until after this date. The colonial boundaries drawn on the map are
only indicative.]
97
Figure 5.2: The Ruaha River valley circa 2003
PAWAGA
11
10 12 14
Ruaha National Park 15
9 13
16
17 8
18
Lunda–Mkwambi 19
Game Control
Area
6
3
5
IDODI
2 4
1
USANGU 10Km2
GAME RESERVE
Old Usangu GR – LMGCA boundary
KIPONZELO
Key:
- Village settlement - Village boundary K I P O N Z E L O - Division

area
- Major road - Major river

(unsurfaced)
Key to current villages:

1 - Mahuninga 5- Idodi 9 - Isele 13 – Mbuyuni 17 – Ilolo Mpya
2 - Makifu 6- Malinzanga 10 – Kisanga 14 – Mbolimboli 18 – Magozi
3 - Tungamalenga 7- Nyamahana 11 – Kinyika 15 – Itunundu 19 – Luganga
4 - Mapogoro 8- Mafuluto 12 – Kimande 16 – Mkombilenga
Cartographical information for Figures 4.1 & 4.2

Maps adapted with modifications from: 1982 Series Y742 Sheet 214/3 Edition 1TSD, Surveys and Mapping Division, Ministry of Lands,
Housing and Urban Development. United Republic of Tanzania. Map Projection: UTM Zone 37s, Datum Arc 1960
98
The groups of small settlements - comprising well over 60 hamlets124 - along the Ruaha River were
ruled by a number of (then) Vansagila125 and their subordinates during the German period. These
Vansagila are well remembered by the survivors of those generations who once lived in these now
abandoned settlements. Many of the eastern hamlets of Ilolo Jumbeate in the Pawaga sub-
chiefdom126 belonged to families variously identifying themselves as Gogo and Kosisamba127. Those
in the central areas of Igawa Jumbeate128 in Idodi Sub-chiefdom129 claimed their own identity -
variously as Kosisamba, Nyambarazi130 or Hehe, while those hamlets in Njongomeru Jumbeate131
further to the west had greater numbers of people claiming descent from, or identity with, the Kimbu
and Sangu groups.
124 Currently about 65 abandoned settlements (some reportedly only consisting of a few homesteads) within the
Ruaha valley were identified during fieldwork, and others are denoted on maps dating from the late colonial
period.
125 The administrative categories are complicated somewhat by the different systems adopted by German and
subsequent British colonial administrative systems. Essentially, the large number of Vansagila (headmen under the
Germans) appointed by the German Colonial administration were much reduced when the category of sub-chief
was reinstated by the British after World War I. The British acquired Tanganyika as a League of Nations
mandated territory in 1919 as a result of the Treaty of Versailles. Four of the Vansagila in what had now
become Iringa District remained as sub-chiefs, while the remaining majority became majumbe (Swahili: headmen;
sing. jumbe - Hehe: sing Munsagila mudodo; pl Vansagila vadodo) retained under the authority of each of their
respective sub-chiefs. Together with their lineages, they locally retained political hegemony, despite frequent
tampering by the colonial authorities (Brown and Hutt 1935; Redmayne 1964).
126 Pawaga, an area reputedly with at one time a heavier Gogo influence than now, retained a degree of
administrative autonomy from the colonial system of Hehe Vansagila during the early German and British
colonial periods and was only incorporated into the Hehe Native Authority by mutual agreement on 16th June,
1927, the year after Chief Sapi was appointed as paramount chief of the Hehe.
127 From Walsh’s (c2002) summary notes on the Kosisamba, ‘The Kosisamba are a people who are reputed to
have originated from the Sungwa and migrated from the highlands in the south to Pawaga’ (Musso 1968, 46-
47). Musso provides a short account of the group and they are also mentioned by Mnyampala (1954, 11),
Rigby (1969, 15) and Shorter (1972, 205 & 304). However, considerable uncertainty remains as to their
identity and origins. This group has now been fully absorbed by the much larger Hehe and Gogo groups and
that the Kosisamba language is now commonly spoken only by those generations who were born, at the very
least, before Ujamaa (pers obs).
128 A Jumbeate was the smallest administrative area in the British colonial administration, overseen by a Jumbe.
129 Idodi sub-chiefdom was re-created in about 1929 when the Kalenga sub-chiefdom was split in two, the other
sub-chiefdom created being that of Nzombe. However, Idodi had previously had a Munsagila in pre-colonial
times that had been appointed by Mkwawa.
130 ‘Nyambarazi’ is a term collected during fieldwork and refers to a particular cluster of hamlets and their
environs along the Ruaha River in what was the Jumbeate of Igawa. One previous incidence of this name being
recorded has been found in a colonial report (dated 24/09/1935, TNA A/8/8) on the construction of an
emergency landing ground in the area for Imperial Airways during the 1930s.
131 Njongomeru is on the border between the Usangu and Uhehe – and during the British colonial period paid
hut tax to the Sangu chiefdom.
99
These hamlets would have each have consisted of a cluster of homesteads set apart by fields and
uncultivated areas in small communities inter-connected by footpaths through the bush - very different
from the relatively densely packed post-Ujamaa villages of today. People lived in these homesteads
and hamlets often as extended families and members of one or more closely inter-related clans - for
example, the Kayela clan. People farmed subsistence crops of finger millet132, maize133 and peanuts
and, in good years of rain, produced food surpluses that tided them over in some measure during
drought years. While important, farming constituted only part of the rural economy during the first
half of the twentieth century134. The villages to the west especially appear to have had a vibrant
honey collecting economy trading with people from the highlands for grain (e.g. Kjekshsus, 1995, 39)
and sundry items - for example, metal products135. Subsistence hunting and fishing were practised
although hunting appears to have been limited to a much smaller number of specialist hunters across
the hamlets. Many settlements, especially those further east towards Pawaga, had varying amounts
of livestock. Cattle were kept where the disease challenge allowed, especially in the Makuluga and
Igawa areas south of the river and Pawaga to the northeast. Elsewhere small stock were kept - for
example, at Mdonya where the tsetse fly and is recounted as having been endemic for most of the
twentieth century. In the Igawa area, livestock tended to be grazed away from the main Ruaha river
on the higher range both north and, particularly, south of the river during the wet season and then
brought down to graze pastures closer to the river during the dry season. The rangelands to the north
of the Ruaha, immediately east of Ilolo, were important dry season grazing areas for herders in
Pawaga.
There appears to have been a certain degree of fluidity in the movement of people between hamlets
as well as to and from neighbouring areas. Villages in the southwest maintained ties with the Usangu -
for example, people living in Njongomeru had livestock lodging arrangements with relatives and
associates in Mawato and Ulanga villages in the Usangu. Mdonya received a number of immigrants
from the Hehe highlands during the 1930s and 1940s - drawn by news of good harvests and hunting.
Villages to the north-east had close relations with the more central villages of Pawaga, seeking
refuge there on more than one occasion when famine struck.
Oral histories paint a general, if somewhat romanticised, picture of a lightly populated landscape in
which people led lives of sufficiency, with years of plenty and others of vicissitude in which drought
132 There is some anecdotal evidence from oral accounts indicating that millet was grown in drier areas with
maize cultivation occurring in areas of more predictable and higher rainfall or being irrigated from
watercourses close by.
133 As Redmayne (1964, 98) remarks, maize came to increasingly replace finger millet as the crop of choice
during the mid to latter half of the twentieth century (see next footnote).
134 In direct contrast to the post-colonial period when farmers in areas of central Tanzania, which were to
become the grain-heartlands of the country, specialised in producing grain under the state’s pan-territorial
pricing and input subsidies (Mung’ongo’ 1998 cited in Bryceson and Bank 2001, 728).
135 It would appear that no iron making occurred in the area (due to a lack of ore – as opposed to much further
west in Ukimbu where exploitable deposits were utilised).
100
was coped with through a range of strategies - reliance on food surpluses, trading livestock and wild
products for grain and, ultimately, in successive years of drought, falling back on wild collected
famine foods. In the few worst years of famine, people temporarily migrated to the vynungu of
Pawaga and Idodi - areas fed by perennial watercourses in which, even during some of the driest
seasons, grain could often be grown.
5.1.2 Increasing epidemiological and ecological challenges

The onset of the colonial era was marked by a number of remarkable events that contributed to the
collapse of, or major changes in, the rural economy of the Idodi rangelands. Unfortunately, little oral
information with regard to this period can now be gleaned for the Idodi and Pawaga areas136.
It is clear (e.g. from Thompson 1881, 212) that towards the end of the nineteenth century, Uhehe137
was rich in livestock and although not expressly mentioned138, it is likely that the peoples inhabiting
the Ruaha rangelands were largely agro-pastoralists139. In contrast, after the rinderpest pandemic of
1889-1891 the cattle economy was shattered and the country was described as being ‘destroyed
and deserted’ by Adams in 1898140 (cited in Kjekshsus 1995, 40). However, despite the undeniable
disaster of the epidemic, the rinderpest may have, in some cases, paradoxically resulted in the
opening up of new, previously stockless wildlife areas, that had been plagued by tsetse, as wildlife
populations - for example, buffalo (Syncerus caffer) - were equally decimated by the rinderpest141. It
136 A relatively large amount of German material is available which Redmayne (1964) consulted for her thesis
and which, due to language constraints, could not be effectively re-consulted. Further, for similar reasons,
German colonial records archived in Potsdam that exist from the early 1890s up to the outbreak of the First
World War in 1914, were also not consulted. Instead Redmayne’s own diligent reading of these texts, where
appropriate, has had to be relied upon.
137 It would appear, despite Kjekshsus’ inferences (1996, 40), and from a close reading of Thompson’s account,
that he did not pass through, nor ‘visit’, northern lowland Uhehe, having approached the highland plateau of the
Hehe chiefdom directly from the Usungwa lowlands in the east in early to mid August 1878.
138 Thompson (1881) records that the Hehe highlands were largely devoid of cultivation save for small plots.
However, it might have been the case that there was little sign of other cultivation left during the middle of the
dry season - the time of year he passed through the country, especially if his travels coincided with a period of
drought, which he indeed twice implies (p. 213 & 216). Thompson arrived in Uhehe during what appears to
have been a year of very poor harvests and yet, interestingly, there is no mention of people being adversely
affected. This perhaps indicates that they were able to fall back for subsistence on the substantial herds of
cattle that existed during this period.
139 Redmayne (1964; 97) describes the ancestors of the Hehe in the first half of the 19th century as having had
a mixed economy, cultivating and owning herds of sheep, cattle and goats.
140 While this statement likely may not be uniformly applicable to the wider study area, it may have been most
applicable to the eastern-most extent of Pawaga and the Ismani area even further east.
141 In Tour Report No. 5, dated 01/07/1931, the then District Agricultural Officer makes reference to
information he gathered that (part of) the Pawaga area had previously been stockless prior to the rinderpest
pandemic of 1893 due to the presence of large numbers of buffalo and accompanying high levels of tsetse fly
and endemic sleeping sickness. It was only after the end of the rinderpest and the demise of the buffalo
101
It is thought that by the onset of the British colonial period at the end of the first world war, the
inhabitants of the Ruaha river valley had re-built their livestock herds but now came to increasingly
face a different epidemiological challenge - that of the tsetse fly and trypanosomaisis. As has been
well documented (e.g. Ford 1971; Kjekshsus 1995) throughout much of the twentieth century,
substantial areas of the Tanganyikan mainland were invaded by expanding fronts of tsetse fly - due
to the disruption of previous agro-ecological management regimes by the pandemics that swept
through east Africa at the end of the nineteenth century and the beginning of the twentieth century.
The Ruaha area lay in between the eastern and western142 tsetse fly belts143 and a substantial
number of the Ruaha villages - previously stockless parts of Pawaga not withstanding - had to
contend with an increasing level of tsetse exposure and associated higher morbidity levels to their
livestock144.
As already discussed, it is certain that people inhabiting the rangelands of the Ruaha have long had
to cope with periods of more extreme environmental variability. However, it is also probable that the
disruptive epidemiological events (both the rinderpest and smallpox pandemics) of the late nineteenth
century may have substantially challenged people’s capabilities to cope with extreme environmental
events, especially when considered in the light of extended general socio-political unrest during this
period up until the end of the First World War. In addition, throughout the twentieth century people
had to contend with expanding tsetse fly belts and the associated sleeping sickness threat to livestock.
A much incomplete, yet sufficiently illustrative, qualitative overview of ecological and epidemiological
events during the twentieth century is given in Table 5.1. After the rinderpest pandemic of the early
1890s, people’s cattle herds slowly recovered up until the 1930s (Redmayne, 1964; 98). For
example, both colonial145 and oral history sources describe the settlements of Igawa Jumbeate as
being well stocked, with cattle herds of up to 1,000 head146. However from the 1950s onwards,
people who lived in Igawa Jumbeate describe a period in which their herds declined drastically as
they succumbed to sleeping sickness. By the late 1960s far fewer cattle were left. In contrast, in areas
where the tsetse had not arrived, further to the east in Ilolo Jumbeate, even during the 1960s, up to 1
in 3 households are described as then still owning cattle (Petwa, pers com) supporting a vibrant ghee
industry and cattle market. Overall, it would appear that from the end of the nineteenth century
population - and thus an initial decrease in the incidence of tsetse fly and sleeping sickness - that people built up
large herds of cattle by exchanging their goat herds for cattle in Ugogo to the north-east (TNA P4/1/III).
142 These belts were largely characterised by two different tsetse fly species - Glossina morsitans in the western
belt and Glossina pallipides in the eastern belt.
143 The local colonial administration was at pains to note that the two fly belts should not be allowed to meet. In
the late 1940s the administration instigated some experimental bush clearances in the Mlowa area (see Figure
5.1) to prevent this from happening. A report (ref 3/3/263) dated 20/07/1942 by the Department of
Veterinary Science and Animal Husbandry documents the steady advance of the western fly belt some 50
kilometres eastwards during the period 1930 to1947 (TNA 70/1).
144 The spread of the tsetse fly and related developments in colonial policy are discussed further in Chapter Six.
145 For example, Tour Report No. 6, 1938 by P.A.P. Robertson (ADO cadet) (TNA P4/1/III).
146 This number should be seen as indicative and not definitive.
102
Table 5.1: An overview of some significant events in the Idodi and Pawaga rangelands
impacting on people’s livelihoods, 1916-1974
Date Event Date Event

1916- Western tsetse belt reaches Mloa,
Repeated rinderpest outbreaksa 1940
1919 spreading eastwardsc.
Severe drought and shortage of food
1918 Poor harvest, widespread droughta 1943 Cattle moved from Mloa due to sleeping
sickness threatc
‘Njaa ya Ihambwa’
1919 Influenza pandemica 1946 Severe drought and famine; outbreak of
small pox
‘Njaa ya malale’- famine
1922- 3 successive locust plagues, possible Western tsetse belt reaches Pawaga,
1947
1924 further rinderpest outbreak (reports of spreading eastwardsc.
wildlife die-off)
1923 Poor harvest 1948 Rains fail
Poor rain; rinderpest brought in by ‘Njaa ya Foloboto/Ikungu’
1925 1949
Somali stock-tradersa Severe famine
1929 Moderate rainfalla 1950 Sufficient rain
Large rainfall event - floods in Pawaga
1930 Western tsetse belt reaches Idodi, 1954 Large rainfall event - Ruaha river floods
spreading eastwardsc.
1932 Locust plague 1955 Poor rainb; Locust plague
‘Njaa ya Panzize (?)’ - famine
1933 1964 Locust plague
Locust plague - despite good rain
Average rain but continued locust
1934 1965 Floods in Pawaga
problema
Bumper harvest in Pawaga despite locust
1935 1966 Floods in Pawaga
persistence
1936 Good rainfalla 1967 Moderate drought
Average rainfall - rinderpest outbreak in
1937 1969 Heavy drought
Pawaga
1972-
1938 Sufficient rainfall and food Prolonged drought
1974
Sources: a From Iringa Provincial Book, Rhodes House, Oxford; b From File V1/8/III (TNA); c Report 3/3/263 of the
Department of Veterinary and Animal Science Iringa (TNA 70/1); Other events compiled from collected oral histories -
open to some margin of error with regard to the exact year they occurred.
onwards147, people living in the Idodi rangelands had weathered a considerable amount of
environmental adversity. By the beginning of the 1950s a number of settlements had been
abandoned in the tsetse affected areas as ecological conditions forced movement to less challenging
environments.
5.1.3 Depopulation and eviction

The major foloboto famine of 1949 forced many people in the Great Ruaha valley to leave their
homes to take refuge in the more central, better watered, settlements of Idodi and Pawaga Sub-
147 There is little, if any, earlier long-term record similar to that provided in Table 4.1 that would enable a
longer-term qualitative comparison of environmental history of the area to be carried out. Indeed Redmayne
(1964; 89) laments this paucity and the poor quantity and quality of historical sources for this earlier period.
103
chiefdoms, but thereafter many returned. In the years immediately preceding the famine, the colonial
administration had begun to consider the expansion of the Rungwa Game Reserve, and its
gazettement as a National Park148. The move was controversial as there was a debate in the colonial
administration between those who, on the one hand, sought to expand the wildlife estate as well as
simultaneously seek benefits resulting from the consolidation of the ‘African population’ through
eviction and resettlement (also see Chapter Three), and those who, on the other hand, were alarmed
at the continued expansion of the tsetse fly belts. The latter were apprehensive about the subsequent
risks of the tsetse fly expanding into currently un-infested rangeland as result of increasing the area
of wildlife estate.
Undoubtedly the foloboto famine of 1949 - together with the area’s reputation for tax evasion and
illicit wildlife hunting - influenced the district administration’s support of protected area expansion.
Indeed, the famine of 1949 was seen by the district administration as a provident excuse for
beginning the eviction process of the estimated 800 families, ‘if at all possible that year’ (Letter
21/15/7, dated 03/08/1949; TNA G1/1, 7). The Rungwa Game Reserve was finally expanded
with the passage of the Fauna Conservation Ordinance on 1st October 1951. Subsequently, G.W.Y.
Hucks (the District Commissioner) wrote to the Sub-chief of Idodi (Hassan Mwakibumu149) in 1952
informing him that the Jumbeates of Mdonya, Kipera, Igawa and Ifuguru were now designated as
part of ‘Shamba la bibi’ (Swahili: the Queen’s (Elizabeth II’s) estate) (Letter 25/08/1952; TNA G1/1,
507). No weapons (of any kind) were to be permitted. Actions by people against crop-raiding
animals were not to be permitted. If people were to actively guard their crops from the depredations
of wild animals, they would have to farm outwith the Reserve (i.e. in effect move) - ‘with regret’ (sic).
Three years later, the Forestry Department declared a Forest Reserve in 1954 (Mapogoro FR: GN
No. 230/6/8/54) adjacent to an area to which a substantial proportion of the evictees from the
newly extended Rungwa Game Reserve were ultimately to settle. The Forest Reserve was only
degazetted when the District and Native Authorities realised that such an arrangement was
untenable.
All the settlements between the Great Ruaha and Kisigio Rivers lying in the new reserve were
subsequently evicted in the following years. In 1955, Njongomeru appears to have been the first set
of settlements along the Great Ruaha River valley to be evicted, as soon as the rains had abated and
vehicular transport could make it through. An eyewitness describes people as being ordered to move
and they were subsequently burnt out of their homes (Nganylika, pers com). Many of the former
inhabitants of Njongomeru moved to the Usangu, whilst others moved south-eastwards to settlements
in the Mkupule area, especially Kinyangesi and Mkupule. The Mdonya settlement was officially
moved to Msembe in the following year, although a substantial number of people are reported as
having moved to Waga (Kabande, pers com) and Tungamalenga (Mbunde, pers com). Overall, by
148 On 12th May 1949, G.G. Rushby, Senior Game Ranger in Mbeya proposed that Rungwa Game Reserve be
expanded (adding the area to its east between the Msombe and Great Ruaha Rivers) and turned into a
National Park.
149 Sub-chief of Idodi, 1949-1953 (Redmayne, 1964; 410)
104
1960 all settlements along the north bank in the reserve had been abandoned, people either having
moved to the settlements immediately south of the river, or with the persuasion of the native authority,
to the more central villages of Idodi (for example, Kitisi) and Pawaga (for example, Kisanga). The
Native Authority, notably through Munsagila (Sub-chief) Vangisada MwaMakendi150, actively
exhorted people living in the more peripheral hamlets of Idodi sub-chiefdom to move to more central
villages where health and education services could be more easily provided. A similar trend occurred
in Pawaga, where in 1964, people living in the outlying hamlets of Ilolo were compelled to move to
Ilolo where a school and clinic were built. By the time the southern extension of the Rungwa Reserve
was gazetted as the Ruaha National Park in 1964151, all the area north of the Great Ruaha from the
district boundary with Mbeya east to Ilolo had been entirely depopulated. Within 10 years, all the
settlements immediately south of the Great Ruaha River would also be abandoned as people were
evicted as part of the new Tanzanian state’s villagisation programme.
5.2 The Ikwavila Valley
5.2.1 The populating of the Ikwavila valley

The descent of people from the hills into the Ikwavila valley (and beyond) is a theme that repeats
itself continually in the twentieth century. Initially, farmers who seasonally came down to the lowland
Ikwavila valley to farm the rich soils during the growing season, are reported as having returned to
the highlands (notably the Wasa area to the east) after harvest to avoid the oppressive heat of the
dry months. Not only was the Ikwavila valley a fertile area for agriculture, but it, together with
rangelands further west, was also a rich hunting ground for hunters from the highlands152 (at one point
under the patronage of the Muyinga chiefs). Thus it would seem that, on the basis of oral history, the
Ikwavila valley remained lightly populated (see Figure 5.3) until the 1950s when the depopulation of
the Ruaha settlements began. However, during the late 1950s and early 1960s, the Ikwavila valley
became subject to a relatively substantial in-migration from the highlands. The small hamlets or
lilungulu (Hehe: settlement or small cluster of houses - after Brown and Hutt, 1935) spread out along
the river margins in the valley (which can still be traced through the occurrence of mango trees that
were often planted in and around the hamlets and homesteads) gradually grew and were eventually
consolidated during Ujamaa into today’s settlements as shown in Figure 5.4. Although today’s villages
in their present administrative status date back only as far as the Ujamaa period, they remain
representative of the villages that existed before the resettlements of the mid 1970s. The arrival of
people in the Ikwavila valley can thus be traced by reviewing the history of a sample of fields from
each village in the valley.
150 Sub-chief of Idodi, 1953-1962 (Redmayne, 1964; 410); thereafter, the Native Authority was abolished and
replaced with the current administrative system.
151 The Ruaha National Park was created on 7th July 1964 through Special Notice 464 under the National Park
Ordinance (CAP412) of 1959.
152 There is some material from oral histories collected to suggest that the people living in the Wasa highlands –
the ‘Wasavila’ had long had a hunting relationship with the Kimbu to the northwest.
105
Figure 5.3: The Ikwavila valley during the late British colonial era - circa 1950
Legend:
1,600 1,500 1,400 1,300 1,200 1,100 1,000 900 metres asl
5,250 4,921 4,593 4,265 3,937 3,609 3,281 2,953 feet asl
Misufi - Hamlet - Major footpath
Kidale - Hill / mountain - River / stream
0 1 2 km
Projection: UTM Zone 37s, Arc 1960
106
Figure 5.4: The Ikwavila valley today - circa 2003
Legend:
1,600 1,500 1,400 1,300 1,200 1,100 1,000 900 metres asl
- Settlement
3,937 3,609
5,250 4,921 4,593 4,265 3,281 2,953 feet asl
- Permanent agriculture
Mahuninga - Village - Village boundary
- Road / track
Misufi - Hamlet
- Major footpath
Kidale - Hill / mountain
- River / stream
0 1 2 km
Projection: UTM Zone 37s, Arc 1960
107
Figure 5.5: The history of field allocation in the Ikwavila valley in the last 50 years
30.0%
25.0%
20.0%
Percent of total
num ber of fields
15.0%
10.0%
5.0%
Tungamalenga (n=163)
Tungamalenga (n=163)
0.0% Makifu (n=129)
Makif u (n=129)
49 Mahuninga (n=90)
Mahuninga ( n=90)
19 5 4 95 9 6 4 96 9
r e -19 1 -19 74 7 9 98 4
P 950 955 - 60 965 -
1 -19 5 -19 8 9 99 4 t
1 1 9 70 0 -1 -19 en
1 1 19 197
19
8
985 90 -1 pres
1 19 -
95
19
[The histories of individual fields were traced as part of a household-farming survey]
During the stratified random sample multi-round household survey that was carried out in Mahuninga,
Makifu and Tungamalenga, the history of the ownership of fields farmed by members of each
household was traced back to when they were first cultivated. The method adopted is not foolproof,
as field allocation does not necessarily always directly reflect the arrival of new immigrants. Some
households were allocated a second or successive number of fields subsequent to their initial field
allocation. Nevertheless, it is very likely that the overall pattern of field allocation remains a fair
representation of the history of the arrival of people in the Ikwavila Valley. One further aspect to the
method adopted is that it is biased towards the history of men, consistent with the patrilineal nature
of all groups in the valley, since dispositions of land are largely male dominated, even when land is
allocated to ‘households’ (see Bryceson 1995, 61).
Thus, as depicted in Figure 5.4, a substantial proportion of fields surveyed were allocated in the
early 1960s in Mahuninga and Tungamalenga. While Mahuninga was subject to in-migration of
people from the highlands during this period, Tungamalenga received a wave of people who were
persuaded to leave the hamlets on the Ruaha River by Munsagila MwaMakendi.
Although the full circumstances remain to be fully clarified, a combination of factors precipitated the
in-migration of people from the highlands during the late 1950s and early 1960s. The largest
contributory factor was a prolonged period of food shortage, in part as a result of a series of poor
years of rain and drought. The decline in food security was exacerbated by a relatively rapid
108
decline in livestock numbers in parts of the highlands - possibly as a result of East Coast Fever153 from
the late 1950s onwards. Cattle are remembered as being important to the maintenance of fertility of
the long-cultivated soils of the Wasa highlands through a system of organic fertiliser production154. It
is recounted how those people who lost their cattle, or were unable to access manure during this
period, found their crop yields declining drastically. Many people, often the poorest, finding
themselves in this situation, elected to move to the Ikwavila valley where it was said that good yields
could be harvested from fertile soils155. Further, with the extension of the Rungwa Game Reserve and
the subsequent creation of Ruaha National Park, there was a steady in-flow of people from the
Ruaha River valley settlements especially to Tungamalenga during the 1960s. The steady trickle of
immigrants into the valley (for example, see the data for Makifu in Figure 5.5) culminated in the
massive, nation-wide, villagisation programme of 1974 as part of Ujamaa. An indicative overview of
the different origins of households in the Ikwavila valley is provided in Table 5.2.
The data in Table 5.2 are derived from the multi-round household survey that was carried out during
fieldwork. Households were asked where they had moved from, and if the household head (male or
female) was a second generation immigrant, the origin of their patrilineage was traced. Although not
completely representative of all households in the valley156, historical migration patterns are evident.
Table 5.2 shows that Mahuninga received a very high proportion of people - over 70 per cent of
current households - from the Hehe highlands, while Tungamalenga was settled by a more eclectic
range of people, not least those from the Ruaha River valley. Makifu in particular was later settled
(during the late Ujamaa years and into the 1980s - see Figure 5.5) by a number of extended Bena
and Wanji families attracted there by rice growing. Thus is reflected in Table 5.2 which shows that
the village has the highest proportion of Bena and Wanji households in the valley.
153 Although East Coast Fever is endemic to the southern highlands, cattle began to be dipped less and less
during the 1950s and 1960s as government cattle dipping services became more expensive. Livestock that had
been regularly dipped, and once challenged, retained their immunity to the pathogen (T. parva), but immuno-
deficient calves when exposed to the pathogen suffered from much higher morbidity levels without the
protection of dipping.
154 Pits were dug into which the dung of cattle (from their stalls) was deposited. Chopped grass was added to
the manure and the fertiliser was then spread on the fields on an inter-annual rotational basis. However, there is
some anecdotal evidence to suggest that this system of fertility maintenance was, in fact, a result of colonial
agricultural extension, and that the practice was only adopted in the 1940s and 1950s. Nevertheless, cattle
were an important livelihood component and hose people who did not own stock are likely to have been drawn
to the fertile Ikwavila Valley.
155 People also came from further afield - for example, Mufindi in the Hehe highlands to the southeast, an area
in which there were a relatively large number of European farms. People also subsequently arrived from the
Bena highlands to the southwest. From collecting life histories, it is likely that these people were frequently the
poorest who were unable to make a living from the demanding soils in the highlands. A number of Bena who
were labour migrants in the Sagara sisal plantations also heard of the Ikwavila valley’s fertility and came to
farm.
156 Please see table note ‘b’ for Table 5.2.
109
Table 5.2: The origin of heads of households living in the Ikwavila valley
[ (n=222 households; ~20% sample).]
a, b
Mahuninga Makifu a Tungamalenga a
Total
Locale 1.5% 0% 4.2% 2.1%
Hehe highlands 77.6% 34.4% 33.7% 45.3%
Hehe lowlands 3.0% 6.3% 14.7% 8.5%

c
Ex GCA/NP 0% 4.7% 15.8% 7.7%
Bena highlands 11.9% 32.8% 17.9% 19.7%
Wanji d 0% 25.0% 0% 6.8%
Other 4.0% 9.4% 13.7% 14.8%
a
These data exclude Ilparakuyo (and Barabaig) pastoralists who are treated separately in Chapter Six.
b
Due to practical fieldwork difficulties only 2 out of the 5 sub-villages were sampled in Mahuninga - and therefore
the figures are not entirely representative of the status quo - for example, there are a limited number of people
who, formerly living in the Mkupule area, were subsequently evicted, and now live in Mahuninga, but are not
represented in the data.
c
Ex-GCA/NP - people who formerly lived in the Ruaha villages which now lie in what is now the Lunda Mkwambi
Game Controlled Area and the Ruaha National Park.
d
An area to the west of Ubena and south of the Usangu.
During the villagisation campaign of 1974, all the remaining settlements in the Mkupule and Lunda
areas were evicted and people were moved to the designated villages157 of Mahuninga, Makifu,
Tungamalenga and Mapogoro in the Ikwavila valley as well as Idodi and Mlowa further to the east.
A proportion of people did not move voluntarily and once evicted returned to their old hamlets as
soon as they could - only to be subsequently re-evicted. One such settlement was that of Waga,
where there was a protracted battle of wills between the government (both at village and higher
administrative levels) and the people of Waga, who repeatedly returned to, and were evicted from,
their homes after villagisation. Eventually, as the villagisation campaign waned, they were left in
peace to live in Waga. The settlement exists today, and has now been incorporated into the village
administrative structure of Mahuninga having its own Mwenyekiti ya Kitongoji (Swahili: sub-village
chairman) who sits on the village council158.
Today, the Ikwavila valley is inhabited by a diverse range of people, but with noticeable differences
between settlements - see Table 5.2. Whereas a majority of people living in Mahuninga originate
from the Hehe highlands - over 75 per cent of household heads or their parents were born in the
Hehe highlands (mostly in the Kiponzelo division), people living in Tungamalenga come from a
greater range of backgrounds - only 34 per cent claim origins in the Hehe highlands.
157 Some people living in the southern Mkupule villages chose to move to the nearby Usangu instead.
158 A similar example exists in Pawaga for the settlement of what once was Ilolo and what is now officially
called ‘old Ilolo’.
110
Even within the relatively small confines of the Ikwavila valley, there have been substantial changes in
the pattern and nature of human settlement and farming in the last 50 years (compare Figure 5.3
and Figure 5.4). Today, the valley is relatively heavily populated by a diverse range of people and
most arable land has now been cleared for cultivation.
5.3 Livelihoods
Having established how many of the farmers came to live in the Idodi villages, and the Ikwavila
Valley in particular, the rest of this chapter is taken up with a quantitative overview of Hehe/Bena
livelihood strategies and practices. The analysis is based on data gathered from multi-round
household surveys and information derived through interviews and conversations with farmers. The
data are presented in the context of the increasing population density in the Ikwavila valley, the
growing commoditisation of land, a long term fall in crop yields for many poorer farmers, and the
development of a new class of wealthier rice farmers. Thus the data have been gathered with an
emphasis on contrasting people’s different land holding and agricultural production patterns in order
to better understand overall livelihood trajectories in Idodi.
5.3.1 Households
The household level was chosen as the most appropriate level at which to collect data - refer to the
methodology discussion in Chapter Two. Thus in practical terms, a household was taken as being a
family sharing a cooking hearth and living in the same compound or close in proximity to each other.
However, polygynous marriages are quite common in Idodi and usually arranged such that partner
wives live in separate households, often locally distant from each other. In this case, the households of
co-wives were treated as separate and not included in the survey, unless they independently
appeared in the sampling protocol. Households were sampled in Mahuninga, Makifu and
Tungamalenga villages, as these are the main villages of the Ikwavila valley.
Overall household size distribution as derived from the household survey is depicted in Figures 4.6
and 4.7. Figure 5.6 presents data based on the number of people per household. Mean household
size for all households sampled was 4.2 people per household. Figure 5.7 presents the same data but
analysed through conversion to ‘Reference Adults’ (RA)159 index. A RA index is used, as it enables
standardisation of household composition - in terms of the varying numbers of adults, children, infants
and older people that often live together as a domestic unit. Mean household size for all households
sampled was 3.63 reference adults.
159 Reference adults (RAs) or average adult mean equivalents (AAMEs) can be calculated for each household
from household constitution data (Little 1980). Thus an adult male = 1 RA, adult female = 0.86 RA, children 0-5
= 0.52 RA, children 6-10 = 0.85 RA, male child 11-15 = =0.96 RA, female child 11-15 = 0.86 RA.
111
Figure 5.6: The distribution of household sizes in the Ikwavila valley as measured in total
people per household (all adults and children).
25.0%
20.0%
15.0%
Frequency
10.0%
5.0%
0.0%
1 2 3 4 5 6 7 8 9 10 11 12 13
Number of people per household
Figure 5.7: The distribution of household sizes in the Ikwavila valley as measured in
Reference Adults
30.0%
25.0%
20.0%
Frequency 15.0%
10.0%
5.0%
0.0%
9
9
9
9
0
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
11
10
-0
-1
-2
-3
-4
-5
-6
-7
-8
-9
01
00
00
00
00
00
00
00
00
00
0-
0-
.0
.0
0.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10
11
1
Reference Adults per household
Note: The number of households surveyed =234 for Figures 4.6 & 4.7.
112
Table 5.3: The dependency ratio of households analysed by villages
Mean number Mean number Mean number
Mean number Mean
adults per of adults per of children
Village of dependents dependency
household household per household
per household ratio
(15-64 yrs old) (>65 yrs old) (0-14 yrs old)
Mahuninga 1.72 0.58 1.22 1.81 1.05
Makifu 2.19 0.23 1.99 2.21 1.01
Tungamalenga 2.36 0.23 2.09 2.32 0.98
The dependency ratio of households sampled for each village is shown in Table 5.3. The dependency
ratio is defined here as the number of adults in a household above the age of 16 but below 60 years
old, divided by the overall number dependents in the household (children of 16 years or younger and
adults older than 60 years old). Care has to be taken in asserting the demographic validity of this
data, as the dataset is at least an order of magnitude too small for normal statistical demographic
analysis (which was not an objective of the household survey). However, as an indicative measure, the
mean dependency ratio of households sampled in each village was 1.05 in Mahuninga, 1.01 in
Makifu and 0.98 for Tungamalenga. These similar dependency ratios mask possible underlying
differences, as shown in Table 5.3. The Mahuninga sample has more than double the proportion of
adults per household over 65 as compared to Tungamalenga and Makifu. Correspondingly, both the
Makifu and Tungamalenga village samples have nearly double the proportion of children per
household as compared to Mahuninga. This data could be seen to be consistent with the fact that
Mahuninga received the earliest substantial immigration of people from the highlands (in particular)
during the 1950s, reflecting a higher proportion of villagers now over 60 years of age. The higher
proportion of children in the Tungamalenga and Makifu samples may be accounted for by the
continued arrival into the 1990s of immigrant farmers, as shown in Figure 5.5, who are likely to be of
reproductive age.
113
Table 5.4: The dependency ratio of households analysed by derived wealth index
Mean
Mean Mean
number of Mean Mean
number number of
children number of number Mean
Wealth adults per adults per
per dependents of people Dependency
index household household
household per per Ratio
(15-64 yrs (>65 yrs
(0-14 yrs household household
old) old)
old)
1 - high 1.88 0.22 1.22 1.44 3.33 0.76
2 1.88 0.29 1.31 1.60 3.48 0.85
3 2.17 0.38 2.34 2.72 4.89 1.25
4 - low 2.52 0.42 2.31 2.73 4.25 1.04
The household dependency data can also be analysed with regard to a derived wealth index
generated from total household agricultural productivity (see Box 5.1), and as presented in Table
5.4. The data show that, in terms of the dependency ratio for households are evenly spaced across
the wealth index, although there is an anomaly in that the third poorest group has the highest
dependency ratio, for an unknown reason. Wealthier households - as defined by their derived wealth
index - have a substantially lower dependency ratio than poorer households: i.e. the households in the
two poorer derived wealth indexes have a substantially larger proportion of dependents than those
in the two wealthier index categories. The sample size is not sufficiently large (in relation to usual
demographic data sample sizes) to draw any further or statistical inferences from the data, except to
observe that the wealthiest households have the smallest overall household size, and the poorest, the
largest. However, the data are consistent with a tendency for poorer households to have a higher
dependency ratio than wealthier households, and therefore proportionately less household labour
available (in terms of able adults) for agricultural production.
Box 5.1: The methodology used to generate the derived wealth index for households
The derived wealth index was generated by calculating the total agricultural productivity in cash
equivalents (see Figure 5.20) of each household expressed as the total agricultural productivity per
reference adult per year for each household. The entire index dataset was then ranked and each
household given an inter-quartile wealth derived index number between one (the wealthiest) and four
(the poorest). The derived wealth index does not include income from other sources which may be
significant for some households - such as wage labour, beer making, fishing, bee-keeping, small
trading, and firewood collection. The poorest households may especially rely more on seasonal wage
labour for their day to day livelihood security than on farming. That said, it would have been difficult
to collect this diverse data systematically and evenly for the purposes of generating a consistent and
replicable wealth index for all households. Agricultural production is, in the main, the most important
form of livelihood strategy for the great majority of households, and is relatively easily measured in
a consistent and replicable way. Thus agricultural production has been used as the basis for the
derived wealth index employed in this chapter and modified accordingly for use in Chapter Six.
114
The household developmental cycle, although variable, may be characterised as follows: children and
adult children160 tend to live with their parents until such time as they marry. At this point, in the case
of sons they begin their own household as soon as they are able, or if daughters, they move away to
live with their husband’s family. Thus parents whose children have married and moved away may live
by themselves although they frequently will look after, for example, grandchildren or be supported
by any adult children - especially daughters - who may have divorced or temporarily separated
from their husbands or become widowed. Thus a household will frequently consist of two generations,
and sometimes three, all contributing in one way or another to household production161.
There is a noticeable amount of seasonal migration to and from the valley with up to 18 per cent162
of households having one or more members who leave on seasonal labour or farm migration. Out of
these seasonal migrants, young men especially may often leave the valley during the wet season to
farm elsewhere, particularly rice, either in Idodi, Pawaga or, less commonly, Madibira163. During the
dry season, household members may seek casual employment away from the valley - for example
with the national park nearby, or further afield (even as far as the capital, Dar es Salaam). Labour
or farm migration may often be undertaken in order to generate capital. This may be made over to
family commitments, or, in the case of young men, may be saved for the future acquisition of
farmland, a marriage bride wealth and associated marriage costs.
5.3.2 Farming and the farming calendar:

The farming calendar - refer to Table 5.5 - dominates the lives of people for the whole of the rainy
season. From early November onwards, as thunderstorms break out across the highland massif to the
southwest, farmers may begin clearing and burning off what dry vegetation has remained from the
previous dry season. Many of the Ikwavila valley’s soils have a high clay content and are therefore
too hard to cultivate by hand or ox-plough before the rain. Prospective tenants and landlords often
make agreements about field tenancies during the months immediately preceding the onset of the
rains. Once the rain begins to fall - usually any time from mid December onwards - and has
sufficiently penetrated the soils, people begin to till in earnest. The great majority of farmers till their
fields by hand and it is only the richer farmers - farming both nchi kavu (Swahili: dryland) and bonde
160 However, adult sons may leave on a seasonal or longer basis on farm or wage migration – see the next
paragraph in this section.
161 Although the household is treated as the domestic unit in this analysis and that of Chapter 5, Creighton and
Omari (1995, 3) draw attention to a number of assumptions that are inherent in household level analysis. The
differing, and at times antagonistic, interests and priorities of men and women, young and old, within the
domestic group, and the relations of exploitation that may exist between them, are often insufficiently taken into
account. Creighton and Omari also question the discreteness of the household as often members may have
interests beyond the household which influence their individual priorities and allocation of time and resources.
162 This figure is derived from a household survey over a two year period undertaken during field-work
(n=234).
163 Madibira lies to the south across the hills and is part of the Usangu. There are a number of large rice
irrigation schemes that support smallholders. However, access to these schemes depends on having the right
contacts and capital.
115
Figure 5.8: A Wetland ‘bonde’ rice fields with a ‘vynungu’ banana patch in the distance
Figure 5.9: Harvested dryland ‘nchi kavu’ fields with nitrogen fixing
Acacia albida trees
The density of Acacia albida trees in this picture is higher than in other parts of the
Ikwavila valley.
116
Table 5.5: The lowland Hehe farming calendar and Swahili equivalent.
Calendar Swahili Swahili b

Hehe equivalent Associated meaning a
month equivalent seasons
November Minga The rain is coming 11. Novembac Ð

December Mbando The beginning of farming 12. Decemba Masika/kifuku
January Ð Ð 1. Januari Ð
February Kivalila The suspension of rain 2. Februari Ð
March Mudope The final rains 3. Machi Ð
April Ny’ wanule The clearing-up of the weather 4. Aprili Ð
May Very light rain/heavy dew which
Mpubutu creates good track marks in the 5. Mei Kiungazi
dust
June Mlagasa The shedding of the trees’ leaves 6. Juni Ð
July Likoloiana - 7. Julai Ð
August Likololiny’lio - 8. Augosti Ð
September Mutanatwi /
The short rains (if they come) 9. Septemba (Vuli)
Mbosmakomalelo
October Muchamia - 10. Octoba Ð
a
The Hehe calendar has locally increasingly fallen into disuse and the meaning of some of the terms have been
locally forgotten. Instead the Swahili calendar is commonly used.
b
Lexigraphically, the terms can be - and are - applied loosely to local seasonal weather conditions despite their
greater relevance to coastal weather patterns.
c
Both the name of each month derived from the English equivalent as well as the number (in Swahili) of each month
are commonly used.
(Swahili: wet valley) fields that have access to, or can afford paying for, the services of ox-drawn
ploughs. Mechanised cultivation, once much more common in the valley, does occur but is expensive
and hired in only by a minority of the richest on nchi kavu fields.
When cultivating, many farmers are aware of a range of considerations and limitations impacting on
the likely success of their agricultural endeavours. While it is important for a household to cultivate
their fields as quickly as possible in order to make maximum use of the rainy season - once they are
confident that the rains have finally arrived, this may not always be possible or necessarily desirable.
Households, especially the poorer or smaller ones, may face conflicting demands on their labour -
between earning income from casual labouring on other people’s fields to ensure their food security,
participating in social labour-sharing networks (Hehe: Mgowe164) and cultivating their own fields. A
farmer may choose to delay the cultivation of a field in order to allow the herb and grass weed
foliage, which springs to life with the new rain, time to grow sufficiently to be incorporated in the new
tilth. This practice is believed to help improve soil fertility and thus the potential yield of the field. The
different types of soil recognised by farmers and influencing the type of crops they plant in their
fields are shown in Table 5.6.
164 Mgowe are further discussed later on in this section.
117
Table 5.6: The different types of soil recognised by farmers in the Ikwavila valley.
Local soil type name1 Description
Vynyanzi A very fine clay, which may be used for pottery
Kinongo A less fine clay, widespread, especially in lower valley reaches
Kivako A clay able to absorb high amounts of water
Lemunda A loamier clay which is less able to tolerate water-logging
Kihaloga A coarse sandy loam
Tifu-tifu An alluvial and friable clay with small quartz particles visible
Kichanga A sandy soil, with low amounts of clay present.
1
The etymology of these terms is unclear and mixed: some appear to be derived from Swahili (e.g. ‘Vynyanzi’;
‘tifu-tifu’), others, it would appear, come from Hehe (e.g. Lemunda; Kihaloga).
The great majority of nchi kavu fields are tilled in a ridge-furrow (Swahili: pl. matuta; sing. tuta)
system. It is also generally held by farmers that this makes weeding easier although it can make the
soil-bed heavily prone to leaching and rain run off (pers obs). Flat bed (Swahili: sese) cultivation is
locally much rarer although potential yields are identified by the minority of farmers adopting this
method as being higher as there is less leeching of nutrients from the soil. Leaching of nutrients in the
wet season, particularly in the already nutrient-limited nchi kavu soils, appears to be a critical factor
in determining crop yields. Fertilisers, of any kind, out with rice cultivation in the bonde are not
applied and never have been - with the exception of one known farmer165. Crops tend to be grown
in mono-stands and as dictated by soil conditions. On the nchi kavu, maize is mostly planted on the
clayier soils together with sunflower and sorghum. Legumes - groundnuts in particular - are planted on
sandier soils as farmers say that the development of the groundnut pods is not restricted in sandy soils
as it is by clay soils. This pattern of associating and planting particular crops with particular soil types
is widely practised and little deviated from. On the mabonde, wetland rice is predominantly grown
and supplied by a network of irrigation channels dug and maintained by farmers - increasingly
organised into water-user associations. The position of a bonde field in the valley in relation to the
irrigation network may heavily influence the amount of irrigation water a farmer can potentially
secure166 - some bonde fields receive a relatively secure supply of water during the growing season
whereas others, especially those at the extremities of the network, may often not receive enough
water in wet years or even any in drier years.
165 This issue is discussed further in Section 5.3.5.

166 Bonde farmers may often be forced to continually revisit their fields at all hours, particularly during the latter
part of the wet season, to ensure that their irrigation inlets have not been stopped and diverted by another
neighbouring or upstream farmer: such practices often result in dispute. However, as recently found by a local
ward tribunal, usage rights to an irrigation channel may normally reside with those people who constructed it,
and other users are obliged, in theory, to seek the consent of these right holders before taking water.
118
Farmers recount that over the years, crop preferences have changed substantially. Previously,
sorghum was grown more than maize and variably intercropped with castor oil plants167 as a cash
crop. Today, maize is ubiquitously grown, with sorghum and millet far less so. Castor oil is largely no
longer cultivated in the valley. Ground-nuts, a crop that has a long history in the area, form a cash
crop for many of the poorer households (maize also fulfils this role to an extent), with rice being a
popular and potentially highly profitable investment for those households able to afford its cultivation
costs - labour, land, and agricultural inputs.
With most of the crops planted in January168, farmers turn their attention to successive rounds of
weeding (usually twice for each maize crop) which continue throughout February and much of March,
without which the crops would be almost entirely weed-smothered. The availability of labour is
crucial, especially for poorer households’ ability to weed their fields sufficiently. This is particularly so
as a substantial amount of poorer household’s labour may be employed in earning a low wage
income from working on others’ fields, and as a result some fields lie un-weeded. Mgowe may often
take place during field preparation, crop planting and weeding. Mgowe are particularly practised in
Mahuninga (in up to 65 per cent of households), although they occur much less frequently in
Tungamalenga and Makifu169. Today, a Mgowe takes the form of one or more households (or
members therefrom) arranging a day upon which people associated with the Mgowe will meet to
carry out one or more agricultural tasks. Once the work is complete for the day, the participating
members return to the host’s house where beer is usually provided. Umoja170 (Swahili:
together/united) work-shares, a relict of the Ujamaa years, are also practised in Mahuninga, and
consist of farmers arranging to come together and work as a group to complete each others’ field
tasks, so that each participant benefits in turn from the shared work of the group.
By the end of March, the first maize will show signs of ripening and, instead of being immediately
harvested the ripening crop is usually left in the field to dry out for several weeks. New ‘compound’
167 Castor cultivation was promoted as a cash crop by the colonial administration and subsequently by the new
government in the early independence years.
168 The crop cycle described here mostly refers to maize since it is this crop which is most widely grown. For
those households farming rice, this crop is planted in two stages – seedlings are initially raised in a nursery and
subsequently transplanted into paddy fields. Rice tends to be harvested later than maize due to its longer
growing season.
169 This is possibly due to the fact that many people in Mahuninga have wider kin and social networks locally,
since the majority of people originate from particular localities in the highlands – especially villages in
Kiponzelo – for example, Wasa.
170 During the Ujamaa period people were compelled to work the fields of the village cooperative two or more
days a week. The resulting crops were then sold at market and the revenues were supposedly to be returned to
the members of the village cooperative – although many farmers remember that this tended to be the exception
rather than the case. Ultimately, with the decline of the Ujamaa years and as a result of irregularities in the
village cooperatives, most farmers increasingly refused to farm the cooperative fields.
119
seed varieties171 recently introduced have a much shorter growing season than older maize stocks of
about 60-70 days, having been introduced by an agricultural extension project to help promote
greater levels of food security. However, older, noticeably larger and longer growing season (90
days) maize varieties are still planted by many farmers, as it is said that, while these varieties are
slower growing, they produce a good yield in wet years, and the ripening maize can be left in the
field to dry without rotting (as the maize kernel inverts itself). Also, once harvested the maize is less
prone to insect damage. By the end of May, nearly all the maize has been harvested and the rice in
the mabonde too.
As the rains end, the landscape begins to dry out very rapidly. From this point onwards, many of the
harvested fields are rented out to be grazed by the herds of resident Ilparakuyo agro-pastoralists
and, in parts of the valley, by Barabaig pastoralists. Grass fires are often set172, or break out, at this
time of year hastening the transformation of the landscape into a parched dryland.
5.3.3 Farm holdings

Most households practice smallholder subsistence farming on land that is locally recognised as being
owned by people through a form of customary freehold or lungulu (Swahili: land sanctioned by
custom; refer to Chapter Four and also Odgaard [2002, 77]). From the early 1940s onwards, as
increasing numbers of people began to arrive in the Ikwavila valley from the highlands, land was
allocated to these people by resident and ruling lineages - for example, the Chambulilas in
Mahuninga. In Mahuninga, people who were able to claim relatedness to families already resident in
the area were able to relatively easily secure acceptance and land for themselves. In contrast, it is
reported that others, who could not claim ties with known families either locally or in the highlands173,
were treated with a degree of wariness and had to negotiate their access to land. Thus people
arriving who were able to claim closer ties secured land free or for a small gift, but it was prudent
and necessary for those who could not claim ties, to make more generous gifts to the ruling and
influential lineages174 in order to gain acceptance and access to the land they sought. However, with
the depopulation of the Ruaha River valley culminating in the villagisation campaign, the villages in
the Ikwavila valley were forced to quickly accept and allocate the arriving evictees land so that
previous systems of patronage appear to have been overtaken - to an extent - by government (and
171 These varieties – such as Kilima, Cargill ‘41’,’42’, or ‘Limited’ are viewed by some farmers as being suitable
only for cash crop purposes as they claim that the maize varieties have a tendency to rot if left in the field to
dry and if brought in for storage on the cob, they are particularly locally vulnerable to insect borer pests. Thus
this maize may often be sold shortly after harvest.
172 The use of fire and its role in land use conflict is discussed in Chapter Seven.
173 There were, and continue to be, close social and kin networks between the Kiponzelo/Wasa area of the
highlands and the lowland Ikwavila valley.
174 During the colonial period, the ‘karani’ was responsible for allocating land in the villages under his control,
and, apparently, rarely the jumbe. A karani would be appointed by a munsagila mudodo (jumbe) although it
would appear that the jumbe’s choice, at least in the case of Mahuninga, reflected local lineal hegemony.
120
Party175) intervention. With the growing and intensifying network of fields, land boundaries176,
initially unimportant and vague, have become very important and are not infrequently the source of
dispute between neighbouring farmers.
Today, about 70 per cent of households surveyed in the Ikwavila Valley own all the land which they
use, slightly over 10 per cent also farm additional land rented by them from landlords, and just under
15 per cent of households are landless, compelled to rent any fields they farm from other villagers -
as shown in Figure 5.10. With rare exception, nearly all land rented is owned by individuals in the
same village, although some of the larger land owners may not reside in the valley (despite being
native to it) as they pursue business interests away from the valley.
Figure 5.10: Overall land tenure patterns in the Ikwavila valley
2000 2001
14.5% 16.2%
Owned only
13.6%
13.2% Rented/Loaned only
Owned & rented/loaned
71.5% 70.2%
Most households - nearly 75 per cent of those surveyed in Figure 5.11 - have access to 1-3 acres of
land per reference adult with a small minority, no more than about 10 per cent, having access to
much larger areas of land of 5 or more acres of land per reference adult. These land-rich households
often belong to lineages that are among the longest established in the valley, although a number of
larger rice farmers (not represented by those in the household survey) have also acquired substantial
land holdings of 20 or more acres.
175 As discussed in Chapter Two, during this period, the government administration was highly politicised, with
little, if any distinction drawn between the ruling Party and local and village government administration.
176 Land boundary markers are often trees, tree stumps, seemingly insignificant sticks and larger stones.
121
Figure 5.11: The pattern of household access to all arable land in the Ikwavila valley
50.0%
45.0%
40.0%
35.0%
30.0% 2000
Frequency 2001
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
9
9
99
0-
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
.0
9.
10
11
-0
-1
-2
-3
-4
-5
-6
-7
-8
12
-
01
00
00
00
00
00
00
00
00
00
0-
0-
0.
1.
2.
3.
4.
5.
6.
7.
8.
9.
.0
.0
10
11
Number of acres available per Reference Adult per household
Note: (i) Fields were either owned or rented; (ii) Sample size = 234.
The pattern observed of similar land holdings between three quarters of households is reflected in
their relative wealth statuses as measured by their derived wealth index - and as depicted Figure
5.12. Only the richest 25 per cent of households have access to significantly more land than all other
households (e.g. for the year 2000: Annova F=21.770; p= 0.000; df=49). The marked degree of
equitability in overall land holdings for 75 per cent of all households can be attributed, at least in
large part, to the Ujamaa period of the 1970s, when it was government policy to allocate ‘three
acre’ plots to every household to farm. The long established, larger land owning lineages in the
valley were able to retain much of their land holding during this period, as they were the lineages
who held government office or were locally influential. Moreover, during this period, there was
sufficient land available for allocation which largely obviated any conflict of interest in this regard.
Most land, where families were male-headed, was implicitly allocated to men, in line with the norms
of a patrilineal society. In this regard, women appear to have had secondary rights at the discretion
of their husbands who might or might not give over land to them, if it were available, in a usufruct
context. Otherwise, women would share-cultivate their husband’s land. Although not a focus of field
research, some women today may be recognised, in their own right, as owning land, which can be
passed on to their offspring or other kin. Women who own land do so when, for example, female-
headed households have been allocated land by the village government - especially during the
villagisation period, or when their husbands have died and they have inherited, at least with a
usufruct right over, usually part or sometimes all, of their late husband’s land. However, for the most
part, reflecting a heavily patrilineal society, most land continues to be owned by men and is largely
inherited by them with women largely only having transitory usufruct rights in any land that they may
inherit or be gifted.
122
Figure 5.12: The equitability of overall land holding per household by relative inferred
wealth status
3.5
2.5
Acres per 1.5

reference adult
0.5
0
High 1 2 3 4 Low High 1 2 3 4 Low
Total land available - inter-quartile 2000 Total land available - inter-quartile 2001
As the population density has grown, especially in the last 20 years, free previously uncultivated and
arable land has become less easily available. In large measure this has resulted in a new, informal
and growing market in land (which began in the late 1970s). The variation between villages in how
each household has come to be in possession of their land - as an indicator of the trend of an
emerging market in land - is shown in Figure 5.13. Although the data must be treated with a little
circumspection, some trends are nevertheless evident. Overall, Mahuninga has the oldest set of fields,
and thus the highest number that have been inherited from one generation to the next. Because the
village in large part is far less eclectic in origin than the others (see Table 5.2), and the amount of
bonde land in the village area the least extensive, land has generally tended to be inherited within
families, instead of being sold, or even re-allocated by the village government177. Moreover, it is
said that in the late 1950s and early 1960s, the established lineages in Mahuninga allocated
arriving people large tracts of virgin land on the understanding that these areas would be sub-
divided and given to their kin as they arrived from the highlands. In contrast to Mahuninga, land has
continued to be allocated to people in Tungamalenga, as the population grows with a small trickle of
immigrants continuing to arrive, and as the village possesses a large extent of unallocated arable
land to the north of the main village settlement. The market in land has grown particularly in
Tungamalenga and Makifu as these two villages have continued to receive immigrants in
contradistinction to Mahuninga178. The land market in Makifu and Tungamalenga has been stimulated,
at least in part, by a large extent of high value bonde land - relative to that in Mahuninga. The
177 Re-allocation of land is infrequent, but does occur.

178 Although this is statement needs to be qualified by the reality of continual seasonal and incidental movement
between the Wasa highlands and Mahuninga, especially within extended families and kin networks.
123
current pattern of land ownership in Makifu is indicative of the likely future trend in land transfers, as
less un-allocated land remains available, and as increasing numbers of fields are inherited and
Figure 5.13: The origin of land currently owned by households
50.0%
45.0%
40.0%
35.0%
30.0%
Status of fields
sam pled 25.0%
20.0%
15.0%
10.0%
5.0% Tungamalenga (n=196)
0.0% Makifu (n=164)
ovt. Mahuninga (n=103)

ill. G ite d
by V Inhe
r
ate d Gif te
d
Alloc h t
Boug
Note: (i) Land rented is included but categorised according to how each owner acquired the land.
(ii) Land ‘gifted’ has been included as a category in order to represent land which is, for example,
given by a living parent to a daughter, son or other relative to cultivate but which cannot yet be
said to have been definitely inherited - not least due to the fact (and arising implications) that the
inheritance ceremony - which proceeds the death and funeral of a person - has not been carried
out. (iii) Sample size = 234.
perhaps, most noticeably sold. Although difficult to systematically investigate, there is circumstantial
and anecdotal evidence to suggest that the price of land has increased markedly in recent years,
especially bonde land, reflecting the differential in the profitability of farming bonde (wetland) as
opposed to nchi kavu (dryland) land.
Access to different types of land varies considerably between households, especially those falling into
different derived wealth index categories. All households have access - and most long-term tenure -
to lower value and less productive dryland fields with over 75 per cent of households holding
acreages of between 1-3 acres per reference adult - see Figure 5.14. The relatively equitable
pattern of overall access to land is reflected in dryland tenure (the largest land type category in
terms of surface area), with only about 20 per cent of households holding land areas larger than 3
acres per reference adult. The richer half of households possess substantially more dryland than the
two poorer categories - see Figure 5.15.
124
Figure 5.14: The pattern of household access to lower value dryland suitable for wet-
season cultivation only (‘nchi kavu’) in the Ikwavila valley
60.0%
50.0%
40.0%
2000
Frequency
30.0% 2001
20.0%
10.0%
0.0%
9
0- 9
1. .99
2. .99
3. .99
4. .99
5. .99
6. .99
7. .99
8. .99
9. .99
0-
.9
11 10.9
10 -9.9
.0
11
-0
-1
-2
-3
-4
-5
-6
-7
-8
12
01
00
00
00
00
00
00
00
00
00
0-
.0
.0
0.
Notes: (i) Fields were either owned or rented; (ii) Sample size = 234.
Figure 5.15: The average size of dryland holding for households categorised by relative
(inferred) wealth status.
2.5
1.5
Acres/AAME
0.5
0
Dryland available - inter-quartile 2000 Dryland available - inter-quartile 2001
125
Figure 5.16: The pattern of household access to high value garden land suitable for dry-
season cultivation (‘bustani’ / ‘vynungu’) in the Ikwavila valley.
80.0%
70.0%
60.0%
50.0%
2000
Frequency 40.0% 2001
30.0%
20.0%
10.0%
0.0%
9
0
.4
.9
.4
.9
.4
-0
-0
-1
-1
-2
01
50
00
50
00
0.
0.
1.
1.
2.
Notes: (i) Fields were either owned or rented; (ii) Sample size = 234.
Figure 5.17: The average size of riverine garden holding for households categorised by
relative (inferred) wealth status.
0.18
0.16
0.14
0.12
0.1
Acres/AAME
0.08
0.06
0.04
0.02
0
Garden available - inter-quartile 2000 Garden available - inter-quartile 2001
The pattern of access to vynungu (riverine gardens), is different, with the great majority - nearly 70
per cent - of households, entirely lacking access to this land category and with more than 25 per cent
of the remaining households holding up to 0.5 acres per reference adult - see Figure 5.16. Riverine
gardens are in short supply as the number of perennial watercourses is limited and suitable sites must
be found which have not yet been cultivated by others or by rice farmers. While ownership of these
gardens is limited to a minority of households, it is not strongly determined by relative wealth status -
refer to Figure 5.17. To some extent this is because gardens have often been allocated to some
households in order to help facilitate greater levels of food security especially for the poor, since
126
these small gardens may remain productive throughout much of the dry season. During dry years,
local government leaders and the local Member of Parliament have exhorted people to farm riverine
gardens as a food security measure and they have demanded that village governments allocate
plots to those people most needing them. In as much as food aid is a source of patronage, in recent
years the continued need for food handouts has increasingly been seen as an embarrassment
nationally, as was irately expressed by the local Member of Parliament at public meetings in Idodi
(pers obs). Ironically despite statements to the contrary, food aid was arranged and delivered in
Idodi during both years of field work. Finally, in addition to their food-security role, riverine gardens
are often also cultivated for cash crops - such as tomatoes, bananas, sugar cane and greens - which
may be sold in local markets or into regional and national commodity chains.
Figure 5.18: The pattern of household access to high value land suitable for rice
cultivation (‘bonde’) in the Ikwavila valley (n=234). Fields were either owned or rented.
80.0%
70.0%
60.0%
50.0%
2000
2001
Frequency 40.0%
30.0%
20.0%
10.0%
0.0%
9
-
0
00
.9
.9
.9
.9
.9
-0
-1
-2
-3
-4
5.
01
00
00
00
00
0.
1.
2.
3.
4.
Figure 5.19: The average size of irrigated rice field holding for households categorised by
relative (inferred) wealth status.
0.7
0.6
0.5
0.4
Acres/AAME
0.3
0.2
0.1
0
Wetland available - inter-quartile 2000 Wetland available - inter-quartile 2001
127
Access to the bonde (wetlands) suitable for rice growing is substantially different when compared with
all other land categories. As with garden land, the majority of households - 70 per cent - do not hold
any bonde land, with a quarter of households holding up to one acre - see Figure 5.18. However, the
distribution of access to land across households of different wealth categories is such that wealthiest
households have by far the largest holdings of bonde land - on average about five times as much as
other wealth categories - refer to Figure 5.19. Although the data are not independent of each
other179, it is generally, although not exclusively the case that the wealthiest households in the valley
are those that farm larger areas of rice - especially a number of farmers who have bonde holdings in
excess of 20 acres.
Thus, with the exception of the wealthiest 25 per cent of households, overall household holdings in
land appear to be relatively equitable. While access to dryland fields is ubiquitous, access to
wetland bonde is lacking for the majority of households, with the wealthiest 25 per cent of households
having by far the largest wetland holdings. Although less than a third of households have access to
riverine (market) vynungu gardens, access to this land category is much more equitable, not least due
to the important role these gardens play especially for poorer households’ food security.
5.3.4 Cultivation, yields and agricultural productivity
Cultivation
In contrast to the relatively small degree of variability in overall land holding for the majority of
households in the Ikwavila valley, there are much greater differences in agricultural productivity. This
much greater variability is due to the varying productive capacity of different land categories held
(wetland is generally far more productive than dryland180) and also a function of the amount of
agricultural investment (especially in timely inputs of labour) each household is able to make.
Whereas wealthier households are able to invest and hire in labour during periods of peak
agricultural workload, the poorest may find themselves unable to cultivate much of the land available
to them, let alone invest in other subsequently necessary time-dependent agricultural interventions -
see Figure 5.20. Thus most households cultivate about 80 per cent of the land they have access to,
except the poorest 25 percent which are frequently unable to do so. Many of these latter households
are extremely poor and, following a year of very poor yields in 2000, were unable to take
advantage of the better rain of 2001. Many of these households will have been forced, as a result of
their long-exhausted granaries, to engage in wage labour in order to secure their livelihoods and
food security, thereby directly detracting from their capability to farm their own fields.
179 Since rice yields per acre are generally 3-4 times higher than maize, and rice prices up to double those of
maize, households farming rice are likely to be wealthier (in terms of agricultural productivity and cash
equivalents – see Table 4.6) than those only farming maize.
180 In part, this is due to the fact that fertiliser may be applied on wetland, but as previously discussed, with rare
exception, never on dryland.
128
Figure 5.20: The proportion land cultivated in relation to total land held by households
categorised by relative (inferred) wealth status.
100.0%
90.0%
80.0%
70.0%
Percentage of 60.0%
land
available to 50.0%
each household
cultivated 40.0%
30.0%
20.0%
10.0%
0.0%
Cultivation rates - inter-quartile 2000 Cultivation rates - inter-quartile 2001
Yields
Dryland crop yields in the Ikwavila valley are generally very low - up to a tenth the level of what
might be harvested in a similar but large scale dryland commercial context. However, and as
discussed further in section 5.3.5, the smallholder dryland agricultural system in the Ikwavila Valley is
very different from a commercial agricultural system. Smallholder dryland agriculture in the valley is
based on the absence of artificial nutrient inputs and little mechanisation, relying instead on the
regenerative capacity of the soils - which is generally low. Thus a low input - low output system results
in dryland crop yields being consistently low - as shown for inter-annual maize and peanut
(groundnut) yields in Figure 5.21. Wetland rice agriculture is very different to dryland agriculture,
and is more akin to a high input - high output system, with yields on average 3 to 4 times higher than
the dryland system181. Improved rainfall between 2000 and 2001 resulted directly in improved crop
yields for both dryland and wetland agriculture.
181 Although perhaps not strictly comparable since the two systems produce different crops, the comparison is
useful for illustrating the difference in gross agricultural productivity between the systems.
129
Figure 5.21: The inter-annual variation in overall crop yields for individual fields (sample
sizes individually indicated for each data set in parenthesis).
14
Dryland agriculture
12
Wetland
10
8
Yield
(Sacks per acre)
6
n=305 n=60 n=42 n=29 n=63 n=52
0
Maize Peanuts Rice Maize Peanuts Rice
2000 2001
Household agricultural productivity

The variation in agricultural productivity between households in both drier (2000) and wetter years
(2001) as measured in total cash equivalents is shown in Figure 5.22 and Figure 5.23 respectively,
and corresponds to the data presented in Figure 5.19. The production of rice in relation to total
household agricultural production is particularly significant in those households with the highest
agricultural productivity. Households with lower agricultural productivity, with rare exception, all rely
on dryland crop (maize in particular) production, which not only is of lower yield, but also of lower
market value.
Nearly all households are able to benefit from an improved rainfall regimen - a large proportion
increasing the value of their agricultural productivity between drier (2000) and wetter (2001) years
by a third or more - see Table 5.7 and Figure 5.24. As many farmers continually stated, rain is a
major limiting factor in their agricultural livelihoods, a claim supported by the data, although as
discussed in the next section, constraints to agricultural productivity are rather more complex than
these claims and data initially suggest. Nevertheless, the inter-annual variability in agricultural
productivity in a semi-arid dryland production system, even during the two years of field work in
which the difference in inter-annual rainfall was not severe, is extremely evident.
130
Figure 5.22: Total individual ranked household crop production for the year 2000 as
measured in Cash Equivalents per Reference Adult.
400000
350000
300000
Total - Rice
Household 250000 Total - Maize + Others
agricultural
production in Cash 200000
Equivalents
150000
100000
50000
Households (n=234)
Notes: (i) Cash equivalents are derived from reported crop yields (measured in sacks and 20
litre cans) as collected through household-based recall survey. Yields, all converted in sacks or
parts thereof, are then changed into cash equivalents using average local farm-gate unit prices
for each crop class (Dryland crops - Maize, peanuts etc; Wetland crops - Rice); (ii) Cash
equivalents are based on the unit price of each major crop during the middle of the harvest -
when prices are generally at their lowest. Thus frequent and significant inter- and intra- annual
variation in farm-gate prices is controlled for as much as possible by adopting the least elastic
price point in the calendar; (iii) Ranked individual household identifier labels (on the x axis) are
not shown for purposes of diagrammatic clarity.
Figure 5.23: Total individual household crop production for the year 2001 as measured in
Cash Equivalents per Average Adult Male Equivalent.
400000
350000
300000
Rice
Household 250000 Maize + others
agricultural
production in Cash 200000
Equivalents
150000
100000
50000
Households (n=234)
Note: Please refer to the notes for Figure 5.20.
131
Table 5.7: Overall average inter-annual agricultural cultivation and productivity
Total area Dryland Crop Wetland crop

Total area
owned by Total cash value per cash value per Total yield
Year cultivated in
households percentage Reference Adult Reference Adult
sample (cash
(n=234) in sample cultivated (cash (cash equivalents)
(acres)
(acres) equivalents) equivalents)
2000 1252.74 872.86 69.7% 15532 5832 4999144
2001 1271.74 862.08 67.8% 25722 9772 8305687
Figure 5.24: Total individual household crop production for the years 2000 and 2001
compared as measured in Cash Equivalents per Average Adult Male Equivalent.
400 000
350 000
300 000
Household Rice2000
Rice 2000
250 000 Maize
Maize ++Others
others2000
2000
agricultural
Rice
Ric e 2001
2001
production in 200 000
Maizee ++ ot
Maiz others 2001
her s 2001
Cash Equivalents
150 000
100 000
50 000
Households (n=234)
Note: The order of households between the data curves for 2000 and 2001 is
not the same. For other details, please refer to the notes for Figure 5.20.
The substantial differential in agricultural productivity between households depicted in the preceding
figures is further elucidated in Figure 5.25, Figure 5.26, and Figure 5.27. On average, the wealthiest
inter-quartile of households are able to produce proportionately higher yields of dryland crops from
their field areas than less wealthy and poor households - compare Figure 5.15 and Figure 5.25. As
previously discussed, many poorer households are unlikely to be able to invest sufficient amounts of
labour in a timely fashion in their crops. Thus wealthier households are more able to either hire, or
through social exchange networks, share labour, enabling them to better make time-dependent
investments in their crops. Furthermore, many of the poorest households are severely impoverished,
and, often as a result of illness and/or infirmity, are simply unable to effectively farm. Many of these
households are unable to take part in social labour sharing networks and some may often even
132
struggle to find work as labourers on richer farmers’ fields. Instead they may depend on the support
of relatives and networks of patronage for their livelihood and unpredictable food security.
Although the differential between households in dryland crop production is substantial, it is not as
marked as that in rice production. Although, as previously stated the data are not independent of one
another182, there is a much greater differential between households of all wealth categories in rice
production, which is not fully accounted for by the variation in household wetland holdings derived by
derived wealth index category - compare Figure 5.19 and Figure 5.26. Essentially, although some of
the poorest households may have access to wetland, they are simply unable to afford entry into rice
production. Apart from the richest households, others may enter rice production but are far less
successful at rice farming than the wealthiest farmers - as shown in the comparison between Figure
5.19 and Figure 5.26. The differential in agricultural productivity is best explained by the in-
affordability of labour and agricultural inputs - notably fertilisers and herbicides. From anecdotal
evidence, it is likely that many wealthier farmers, and certainly the minority of larger scale farmers
not covered in the household survey data, with longer experience, may be more skilled and are
better able to obtain the best results from using these agricultural inputs. Finally, the wealthiest
farmers are able to capitalise most on better years of rain, for both dryland and particularly
wetland rice crops, producing proportionately the largest inter-annual increment in agricultural
productivity for all derived wealth index categories - as shown in Figure 5.27. This can be accounted
for by their relative lack of resource constraint, in terms of labour and agricultural inputs required, in
order to realise the potential increase in agricultural productivity that a wetter year potentially
provides.
Summary
The wealthiest 25 per cent of households have access to substantially greater areas of land in all
categories than the remaining households that, between them, show little significant variation in
overall land holdings. A minority of households own most of the high value, high yield, wetland rice
fields and the majority of these households are amongst the wealthiest. The ability of households to
gain access and to cultivate bonde rice fields has the largest relative impact on the quantity and the
cash value of total household agricultural productivity - since rice both produces the highest yields
and is of a consistently higher market value. Most households unable to farm rice - as a result of
insufficient capital assets in land, inputs and labour - are restricted to dryland production for their
livelihoods. For many households this consists of concentrating on farming maize and less frequently
groundnuts - sufficient for a subsistence living and a mostly very limited and unpredictable income
from the sale of surpluses. However, there are a limited number of specialist maize farmers, who are
wealthier than others, better able to accomplish timely and sufficient labour investments in their crops
and who skilfully produce relatively far greater dryland crop harvests than others. A smaller minority
of households - the poorest and most destitute - fail to secure a subsistence living from their land.
182 A large measure of confidence can be expressed in the overall pattern of data interpretation as a result of
the trend apparent in dryland crop production being consistent with that of rice production.
133
90000
Figure 5.25: Inter-annual household
dryland crop production measured in 80000
cash equivalents per reference adult

and classified into inter-quartile 70000
ranges.
60000
Note: Household wealth indexes 50000

Cash
remain consistent between years and equivalents
40000
across categories of crop production
30000
20000
10000
0
Maize + others inter-quartile 2000 Maize + others inter-quartile 2001
50000
wetland crop production measured 45000
in cash equivalents per reference 40000

adult and classified into inter-
quartile ranges. 35000
30000
25000
Cash
equivalents
20000
15000
10000
5000
0
Rice inter-quartile 2000 Rice inter-quartile 2001
140000
total crop production measured in
cash equivalents per reference 120000
adult and classified into inter-

quartile ranges. 100000
80000
Cash
60000
equivalents
40000
20000
0
134 Total inter-quartile 2000 Total inter-quartile 2001
Figure 5.28: Women brewing mbege together - by cooking the maize gruel
wort in the early stages of the brew-process
In the background are houses characteristic of the poorer members of the Idodi villages, while in
the foreground is a small homestead plot of long-harvested maize. In the far background is a
baobab tree - these trees are quite common in the valley. The tree still in leaf on the left is an
Acacia that had been planted for shade.
Figure 5.29: A group of men playing ‘bao’ (a popular board game) at

’kilubu’ (beer club) in Mahuninga.
Bao is often played very competitively. This scene is somewhat atypical of this beer club (in
Mahuninga) which is often full of people and a very lively place. An Ilparakuyo woman is standing
in the far back right of the picture - reflecting the fact beer clubs are popular and an important
social focus for trading and recreation for all the different groups of people living in the valley.
135
5.3.5 Dryland agriculture: Rainfall, crop yields and soil fertility
Farmer narratives: changing rainfall patterns

Many people recount that rainfall patterns have changed substantially in the last 40 years and that
these changes have had significant adverse impacts on farming - both in terms of labour and in
yields183. A well and often rehearsed narrative among many local dryland farmers is that the rains
have become much shorter in their duration, and generally overall sparser. Whereas previously the
rains frequently began in early December and ran all the way through to the following May, these
days, they begin later - often in early January - and end much more abruptly - in early April. From
their points of view this has resulted in the window in which planting can be carried out now becoming
much narrower. Formerly, farmers aimed to plant up to three times in the growing season - so as to
make best use of a restricted amount of household labour, as well as to maximise the reliability of
yields. The rains were sufficiently prolonged to achieve this, but today they no longer are. Further,
many dryland farmers say that the progressively changing characteristics of rainfall through the wet
season must be exactly right to generate a good yield. At the onset of the wet season, the heavy
thunderstorms and violent downpours are helpful in making the land workable, but these
thunderstorms should not persist too long. Instead, as the rainy season progresses they should be
increasingly replaced by a pattern of sunny days interspersed with days of light showers - gradually
dying out towards the end of the wet season. It is essential that not too much rain fall just as too little.
Underpinning this narrative is the fact that dryland agriculture, especially where soils that have been
intensively cultivated without rest for many decades, can be extremely risky and requires very
specific climatic conditions in order to be reasonably successful.
Farmer narratives: declining crop yields

Many local dryland farmers say that their crop yields have fallen substantially in their lifetimes. The
yesteryears are recounted as being ones of substantial surpluses184. Today, substantial surpluses for
many farmers are rare if they occur at all, and many see themselves as frequently facing chronic
food deficit challenges. Many identify the decline and change in rainfall as being the proximal cause,
others think that pastoralists are also to blame from their herding of stock on the dryland fields after
the harvest. Other dryland farmers say that they have not suffered from - at least to the same extent
as others - the marked crop yield declines described. Farmers do have strategies to help maximise
crop yields and they focus on the way the soil is tilled. If tilling can be delayed long enough for the
accumulated weeds in the early wet season to be incorporated in to the tilth, then relatively small
beneficial effects for yields will occur in subsequent harvests. Clayier soils should be tilled first as their
183 As discussed in Chapter Three, there is no long-term rainfall data known to be definitely available for the
Idodi area to corroborate this. With hindsight, it would have been useful to have investigated whether any
(incomplete but long term) time series data existed at Msembe in Ruaha National Park.
184 Despite searching through village records from the 1970s and earlier (such as they existed), no reliable crop
production figures could be unearthed. However, circumstantial evidence exists supporting farmers’ recollections,
mostly in passing reference to the existence of large village grain stores which were used to store the maize
shipped out of the valley to Iringa particularly during the late 1970s and early 1980s.
136
early tillage maximises their water absorption and holding capacity. Soils should not be tilled too
deeply, to avoid bringing up infertile lower top/sub-soils. A further strategy followed by a minority
of the more successful dryland farmers is not to allow pastoralists to graze their cattle on the post
harvest residue, which they say helps guard against crop yield declines. These farmers feel that cattle
damage the soil’s structure, as its friable clay component is compacted and/or blown off in the dust
cloud of the grazing cattle. The long-term and cumulative effect of this process is that soils on which
cattle are seasonally grazed become increasingly sandy, and with less clay content, more infertile.
Furthermore, these farmers argue that the manure that is left on the fields by the cattle while they are
grazing is minimal and not significant for improving soil fertility.
Understanding locally held wisdoms

Several decades ago the Ikwavila valley was very fertile and had a local reputation to match that of
the former, well renowned grain basket of Ismani to the east. From around the late 1950s, there was
a maize market in Mahuninga where much of the valley’s surplus maize was sold. Indeed, in a good
harvest year during the height of the Ujamaa period (late 1970s), when provision of sundry supplies
was often unpredictable at best, farmers recount having excess cash derived from maize surpluses
which they had difficulty spending. But gradually over the years, soil fertility has fallen - contributing
in part to the collapse of the local maize market in the late 1980s. Today, the dryland agro-
ecological system is one of low input, low output, relying on the regenerative capacity of the soil as a
result of natural nutrient cycling and mineralization. In the dryland fields, little, if any, soil fertility
intervention takes place, except for the limited number of tilling strategies described above. Although
there is a precedent of active organic soil fertility intervention in the highlands of which many of the
valley’s dryland farmers are aware, a very large majority of these farmers are stockless, and are
thus precluded from practising field manuring. There exists a certain doxa (Bourdieu, 1977) in
practices and attitudes to farming. There is a prevailing and widespread wisdom that any fertilising
the dryland soils is disadvantageous. The wisdom holds that the dryland soils, when fertilised, result in
the maize crop growing extremely tall without bearing grain and, in the event of drought, the crop
being burnt by the fertiliser application. Exhaustive attempts to track the source and empirical
evidence of this wisdom met with little success. One of the few farmers who had, in the past, applied
chemical fertiliser to his dryland fields said that he had been pleased with the results - after having
been shown by a Greek farmer in the highlands how to do so. Instead, Hehe farmers in the Ikwavila
valley view rainfall as the central factor controlling the fertility of the land. Thus people look to rain-
making rituals to bring the right type of rain which they view as necessary for the maintenance of the
fertility of the land.
While there are a minority of successful and innovative dryland farmers, overall dryland agriculture
appears to be characterised by a substantial degree of conservatism about what agricultural
practices are appropriate leading to very low levels of innovation in dryland farming. The underlying
reasons are complex and underpinned by a number of factors.
137
Rainfall and soil fertility: synergistic dynamics?
Scoones (2001) states that the agro-ecology of African soils in the semi-arid zone is extremely
complex and varied - even on a local scale - and this is much in evidence in the Ikwavila valley.
Perhaps most critically, the productivity of these soils is a function of the synergistic interplay between
rainfall and soil fertility status. At any given moment, soil fertility is a product of spatial variation, the
history of its cultivation and the long- and short-term rainfall regime. These factors impact on a soil’s
properties - nutrient cycling, acidity, organic matter levels, water holding capacity, microbial activity
and soil structure. Moreover, within an agricultural season, all these soil properties are likely to
continually change. A farmer has to interpret these prevailing conditions and make strategic choices
as to whether and when to fertilise, when, what, where and how much to plant. Furthermore, the high
levels of unpredictability in rainfall, soil status and thus crop yields may be compounded in situations
where the nutrient and organic stock of the soil is low or has become depleted (see also Scoones and
Toulmin 1999, 21-62). Thus farming remains a high risk and unpredictable undertaking for all
dryland farmers in the Ikwavila valley.
Expensive investments and uncertain returns

Soil fertility intervention for many farmers in the Ikwavila valley is currently not a viable option. Not
only does locally received wisdom weigh against this option, but also the lack of indigenous
experience with chemical and organic fertilisers is a persuasive disincentive, in what is an
unpredictable and high risk environment. Moreover, the economic and labour costs of farming are
challenging for many. Chemical fertilisers are expensive, not easily locally available and are also
required at a time of year at which household cash resources are least available and most in
demand. Organic fertilisers are available from livestock bomas (Swahili: cattle kraal or enclosure) but
these are frequently far removed from the fields and little viable means - or, indeed, precedent185 -
exists for transporting the manure from boma to field. The labour involved in transporting manure
from boma to field would be substantial, particularly during the peak of labour demand and income-
earning pressure, especially for poorer households186. Critically, there is little guarantee that the
substantial investments required in labour and cash to improve soil fertility would yield results in the
short-term. In an environment where the discount factor is extremely high, such a risk is therefore
untenable for the poorest and dissuasive for wealthier households, irrespective of the long-term
potential for agricultural dryland productivity gain that soil investment may represent.
5.3.6 ‘Twililage pe twiwumi!’ - Let us eat while we are still alive!

The agro-ecological and socio-economic constraints hereto described are alone insufficient to provide
adequate explanation of the nature of dryland agricultural production. Socio-cultural norms and
185 Although in the Kiponzelo highlands immediately to the southeast, manuring appears to be increasingly
important in farming systems, with a developing market for manure (Scoones and Toulmin 1999, 109).
186 Cultivation on boma sites, apart from that carried out by herders themselves (see the next Chapter) does not
occur, because these sites are recognised as belonging to the herders since it is they who cleared them from the
bush.
138
expectations play a significant part in how many people in the Ikwavila valley approach and view
farming and their wider livelihoods. Attaining a sophisticated understanding of socio-cultural norms,
given the nature of the fieldwork carried out, was a difficult undertaking, but at the risk of
generalisation and over-simplification, some insights gained are useful in providing further nuance.
Candid and not infrequent reference was made by close informants to wealth-equalisation pressures
and the competing worlds of social and economic reproduction. This can be best illustrated by the
reflection of one Bena farmer who had arrived in the valley in the late 1970s. He recollects how he
and others had heard about the farming potential in the Ikwavila valley and came down to see for
themselves. Having been allocated land they began to farm leaving their families in the highlands.
With time as they established themselves, they became very much aware that it was incumbent upon
them to remit their profits (whether in cash or in kind) back home to their families. Moreover, they
came to understand that it was socio-culturally unacceptable for them to display their new found
wealth to their Hehe neighbours. For many Hehe farmers, land is not only the source food, but it is
also central for social relations mediated through beer brewed from the land’s produce (pers com
Mdindile). Until recently, re-investing the wealth of the land in ways which set one apart from
extended kin and associates (for example, in seemingly relatively ostentatious houses or processions)
remained outwith social norms. It is only over a protracted period of 20 or more years that it has
become more acceptable for the Bena to display their success more openly and to invest more
heavily in wetland fields and material holdings. This demonstrates the degree to which orthodoxy and
the universe of the undisputed (Bourdieu, 1977) prevails in local Hehe society, and how heterodoxy
and the universe of discourse - and therefore innovation in, for example, fertility intervention - have
been resisted and suppressed. There is indeed direct symbolic relevance in the emergence of a new,
competing and entirely different heterodoxy of agricultural and social reproduction - that of wetland
agriculture.
A further factor impacting on people’s attitudes to farming is the historical legacy of the Ujamaa
years and its command economy, where people were forced to grow specific types and quantities of
crops, not only for their own consumption, but especially for the state via village co-operatives. Often
these policies were unpopular and not particularly successful, and are likely to have been
disincentives for agricultural innovation and diversification.
5.3.7 Wetland agriculture: rainfall, crop yields and soil fertility
High inputs and high returns: predictability and profits

Wetland agriculture in the Ikwavila valley presents a very different context to dryland agro-systems
for those farmers able to afford the capital outlay in high labour investments and agricultural inputs.
Whereas agricultural inputs may be largely dispensed with - in the short-term - at the risk of reduced
yields, labour investments cannot, and are thus a pre-requisite for rice farming. Thus only those
households with sufficient access to labour are likely to farm rice. The bonde, especially those fields
with guaranteed/more assured irrigation water flow, are far lower risk environments than the
dryland fields. Since many of the households farming the bonde for rice are among the wealthier,
they are able to afford the high capital outlay in the face of reduced risk and relatively assured -
and high - yields. In stark contrast to dryland agriculture, fertility interventions are frequent as
139
immediate short-term benefit is assured and the long-term costs of not continuing fertility interventions
are high. For example, it is frequently held by rice farmers that once a field has been fertilised with
chemical fertilisers, these fertilisers must be re-applied in successive years or else yields will not only
drop, but drop below the level of yields achieved before the onset of fertiliser application. Wetland
rice cultivation - as demonstrated in Figure 5.19 - is far more productive than dryland maize
production and market prices are consistently, although with exception, higher than maize. Overall,
wetland production is very different to that of dryland as yields, while still dependent on rain, are
much higher and more predictable. Thus innovations and investments in wetland crop husbandry are
much less at risk and far more liable to return a direct and immediate net benefit to the farmer.
5.3.8 Livelihood options
Livelihood diversification
Agriculture remains central to nearly all households’ livelihoods in the Ikwavila valley. Alternative
modes of production are limited and predominately lie in wage labour and in dry season garden
cultivation. In this regard up to 45 and 40 per cent of households surveyed engaged in wage labour
and dry season garden cultivation respectively. Wage labour on wealthier farmers’ fields is most
important during the wet season when labour may be paid for in cash, or less frequently, grain. Dry
season wage labour may involve the clearing of fallow fields or virgin land. A range of other
diversification possibilities exist including beer making, bamboo wine trading, artisanal work
(carpentry, pottery, brick-making and construction), fishing, plant and grass harvesting, honey
gathering, bee-keeping and hunting. However, many of these activities are limited to those with the
knowledge and skills, or they are socially delineated and restricted - for example beer making and
grass collection are culturally in the woman’s domain; hunting, fishing and honey gathering are in the
man’s. In addition, perhaps with the exception of bee-keeping, most activities have a relatively low
return on the amount of labour invested, although their value should not be measured just in terms of
their potential economic potential.
Beer making is especially important for women as a means to benefit monetarily and socially from
the household harvest. Whereas farming may often be nominally a joint endeavour between a wife
and her husband, the disposal and sale of the household’s harvest is likely to be most controlled by
the husband. In this regard, a woman may claim or effect access to part of the household maize crop
for the production of beer, as a means to adding value to the crop and securing a useful income.
Although large amounts of labour are involved in beer-making (maize preparation, wood collection,
beer brewing), women are able to retain at least part of the proceeds from the sale of the beer to
neighbours and villagers, an activity that is socially important and one in which women may invest in
reciprocity and network building.
Debt, market speculation and value adding

While dryland agriculture generally remains important for subsistence and social production, wetland
agriculture represents a greater opportunity to diversify and expand household wealth. However,
despite the relatively poor productivity of dryland production, many households - up to 65 per cent -
may sell a substantial part of their harvest either to obtain a cash income or to pay off the year’s
140
accumulated debts. Debt is a notable feature in the socio-economy of the Ikwavila valley - with
households in long-term debt to patrons who will have loaned them cash or food particularly during
the wet season when all households, not least the poorer, are under the greatest stress. Interest rates
are extremely high. It is not uncommon for a loan of money or food to be repayable within one year
(usually at harvest) with an equal amount again in interest. While small loans and gifts are
commonplace within social and kin networks, larger loans are often the domain of wealthy rice
farmers who operate extremely profitable loan businesses that are often integrated with agricultural
commodity speculation. Thus a small rice farmer will take out a loan from a patron in cash to be
repaid at harvest at twice the value in rice. At harvest, the patron collects the rice to the value of both
the loan and interest accrued from the farmer. He then stores it locally until the early wet season
when the regional and national rice price is at its highest - often up to, or more than, twice that at the
farm gate at harvest - when the patron will sell his large rice stocks and make a considerable profit.
Thus many households are unable to maximise the potential value of their agricultural production since
they face substantial pressures - not least debt and tax obligations - to dispose of their crop
immediately instead of having the freedom to wait for seasonal price improvements187. In contrast,
many of the wealthier households are able to add value to their harvested crops by delaying their
sale on local and regional markets.
5.4 Conclusion
The nature and way in which people occupy and relate to the landscape of the Ruaha River valley
and the Idodi rangelands has changed much in the last 50 and more years. While substantial parts of
the landscape have been depopulated and turned over to the national wildlife estate by pre- and
post-independence governments, the remainder has become increasingly settled by a range of
peoples evicted from their former settlements in the Ruaha valley and by many others from the
highlands seeking new agricultural fertility. The old mixed production and exchange economy that
once existed in the Ruaha valley and which resettled people continue to practice, has gradually
started to be eclipsed by a market economy, based heavily on rice, and dominated by a wealthier
minority of the population, many from the highlands. While dryland-based agriculture has suffered
from the effects of soil exhaustion compounded by the unpredictability of a semi-arid climate regime,
wetland-based cultivation has come to constitute a far more reliable and productive livelihood
component for those in the position to benefit from it.
187 The impact of food-aid may be extremely disadvantageous to farmers who have successfully managed to
store part of their crop in order to benefit from seasonal price increases. Food aid was delivered in both years
of field research and while there was a genuine need for its distribution, due to its poor targeting and
ubiquitous availability, it resulted in the local maize market price collapsing and a number of angry and
dispirited farmers who were compelled to sell their maize at a deflated price far below the opportunity cost
incurred (i.e. the production cost together with the loss in increased potential earnings as well as the loss of use
of the valuable cash tied up in storing the maize for an extended period).
141
Thus a distinction may be drawn between the ‘old’ and the ‘new’. The ‘old’ constitute the majority of
dryland farmers who are having to unsuccessfully contend with long term declines in agricultural
productivity, faced with limited diversification opportunities, and constrained by socio-economic
circumstances and cultural norms. The ‘new’ are a varied group of farmers - largely more recent
immigrants to the valley - who are successfully developing their rice-based agro-economic production
and business interests and expanding their social networks of patronage. Some households are in the
process of trying to enter into the new more market-oriented economy and are investing in wetland
fields and rice production. Many others remain unable to do so, whether as a result of socio-economic
or cultural constraint, or both.
142
6 The Ilparakuyo of Idodi: Contending with change
The peak of Idelemule mountain presents an ideal vantage point from which
to view the mosaic of the landscape stretching out across to the foot of the
highland scarps in the Mahuninga valley several hundred feet below. If one
looks hard enough, through the dry season haze and dust, away from the rich-
green irrigation furrows and streams in the ‘bonde’, casting one’s gaze up on
to the grey leafless woodland slopes, brownish circular clearings scattered
across the landscape will, with some concentration, begin to take shape. All of
these clearings lie a certain distance away from the shiny tin roofs of the more
easily definable farm settlements. Seldom can a clearing be seen sited close
to the lush-green gardens in the stream courses, nearly always instead being
located, at the very least, several stone throws away. Other less-easily
distinguishable clearings emerge from the dryland, shimmering in the heat
waves, only to disappear before one’s eyes to be replaced by other
phantoms of the mid-morning haze.
In this second chapter on people’s land use practices and livelihoods, I present an analysis focussing
on the pastoralist Ilparakuyo together with a short review of the Barabaig in Idodi. I examine how
Ilparakuyo livelihood status and land-use practices have changed in relation to the increasingly
crowded nature of the rangelands in which they find themselves today. I first trace some of the key
events and processes in the Ilparakuyo’s past which led to their arrival in Idodi and were increasingly
impacted upon by the state. I similarly consider the events leading to the arrival of the Barabaig in
Idodi, before presenting an overview of the Barabaig domestic unit, livelihoods, social organisation
and range-use. I then focus on Ilparakuyo herding patterns and farming practices in the Idodi
landscape, leading into a detailed analysis of their households and herds, and their farming and
herding livelihoods. I demonstrate that the Ilparakuyo are increasingly relying on farming for their
livelihoods. I argue that the Ilparakuyo herd has declined as the landscape has become demarcated
and zoned as a result of villagisation and land-use planning. Other factors - such as increases in
livestock disease, a growing cash economy and higher levels of commoditisation - are likely to have
compounded the trend. I argue that for many Ilparakuyo, particularly the poorest, an increased
reliance on farming and decreasing access to livestock has led to growing entrapment in increasing
poverty relative to other Ilparakuyo. While some Ilparakuyo families remain relatively wealthy, many
are now on the verge of falling out of livestock-based livelihoods or already have done so, and thus
Ilparakuyo socio-economic relations with farmer communities are becoming evermore important and
significant.
143
6.1 The establishment of the Ilparakuyo in Idodi
6.1.1 Pre-Ujamaa: new frontiers

The Ilparakuyo, to whom many of these clearings are attributable, are an agro-pastoralist people
who, together with the more recently arrived Barabaig, constitute the two major herder groups that
live in Idodi. The Ilparakuyo are Maasai speaking and forming one of the 22 associated sections the
comprise the wider Maa-speaking peoples of eastern Nilotic origin (Sommer & Vossen 1993, 30) who
live in the rangelands in an area stretching from what is now northern Kenya to south-western
Tanzania. Although distinguishing themselves from other Maa-speaking groups, not least the Kisongo
and Salei Maasai of the plains and highlands of northern Tanzania, the Idodi Ilparakuyo make
reference to themselves as being Maasai188 and as being part of the Maasai socio-cultural diaspora
- especially in regards to the timing and practice of ritual. During the Ilparakuyo-Kisongo (Iloikop)
wars which occurred from the 1820s to the 1880s189 (Galaty 1993, 74), the Ilparakuyo progressively
lost control of their previous territory in Maasailand to the Kisongo. The Ilparakuyo began to
withdraw from their former homeland in what is now central northern Tanzania, moving southeast
towards the central plateau and the coastal lowlands from the mid-nineteenth century onwards. This
dispersion is unlikely to have been their first foray south as Gogo oral histories record Ilparakuyo
raiding in the eighteenth and nineteenth centuries into Ugogo (what is now central Tanzania) (Maddox
1995, 14). The Ilparakuyo were initially resisted and fought against by the Gogo with the help of the
Hehe during the earlier part of the nineteenth century. However, over time the Ilparakuyo increasingly
built close relations with the Gogo through intermarriage and exchange practices. The Gogo adopted
a modified version of the Ilparakuyo age-grade system, and began to speak Maa in addition to their
native CiGogo. Some of the Ilparakuyo families living today in the Idodi rangelands are recognised
by other Ilparakuyo as being of Gogo extraction.
Although the Ilparakuyo were present in southern Ugogo by the end of the nineteenth century, they
only moved more permanently into Pawaga and what is now the periphery of north-eastern Uhehe in
the early part of the twentieth century (Redmayne 1964, 392). The first official reports of Ilparakuyo
pastoralists migrating into the Pawaga area occurred in 1928 and 1934. By 1953 Ilparakuyo
pastoralists had reached Idodi (pers com. Lemu Lebere) and the Usangu (Charnley 1997, 97). It is
likely that the Ilparakuyo initially may only have been seasonal transhumant residents in north-eastern
Uhehe, but by the late 1930s they had become more permanently established in the area. Redmayne
(1964, 396) remarks that the Ilparakuyo were allowed to utilise the rangelands by the resident Hehe
on the condition of refraining from stock raiding, and paying tribute to the Vanzagila190 of Pawaga
(Redmayne 1964, 360). Hehe oral accounts suggest that initially the Ilparakuyo came to be
188 Particularly when interacting with Hehe/Bena farmers or with village or district government and NGOs.
189 The history of the series of battles/wars has been reconstructed by Galaty (1993), Waller (1979) and
Fosbrooke (1948, 1956) by deducing the period in which each age set was extant at the time of each event
described in oral history. Rigby (1992, 66-77 & 106-127) also provides an account of Ilparakuyo history
particularly for the British colonial period in relation to that of the Maasai in general.
190 See previous chapter
144
recognised as occupying - although not necessarily exclusively - particular range areas of what is
now the southern and south-eastern periphery of the Ruaha National Park - in what was previously
known as Ukosisamba191. Singularly pastoralist192, they coexisted and traded with resident Hehe-
affiliated and Gogo agro-pastoralist farmers.
By the mid 1950s, the Ilparakuyo had continued their expansion southwest into the Idodi area,
establishing their presence at particular sites, many of which have been more or less continuously
occupied since. Subsequently expanding to the southwest, groups of Ilparakuyo arrived in the Usangu
plains by the early 1950s (Charnley 1992; Walsh 1983), arrivals that are likely to have been
precipitated, in part, by contemporary developments in the Idodi and Pawaga rangelands.
The Ilparakuyo appear to have largely followed a semi-transhumant lifestyle in the hundred or more
years that they have occupied the wider drylands of Idodi and Pawaga. More recent enkang’ (Maa:
homestead; pl inkang’itie193) site histories have been traced and show, for example, how one or more
families would arrive in a particular area. Staying anything from less than a full cycle of seasons to
several or more years, they would move on, sometimes a substantial distance to another site, perhaps
drawn by news of better conditions. Ecological changes, for example, an outbreak of disease, a
build-up of parasites, and the depletion of grazing and watering conditions due to inter-annual
rainfall variations are likely to have been major considerations in compelling an enkang’ to move to a
different area. However, as is the case today, ecological perturbations alone may have been
insufficient cause for the movement of an entire enkang’. Not least, strategies, such as temporarily
establishing a sub-enkang’ for a family’s herd in a seasonally more favourable area, are likely to
have been more appropriate and convenient. Conversely, different concerns, for example, those of
conflict and security (both physical and magical) or life-cycle events (for example, the dispersal of
family members after the inheritance ceremony of a deceased patriarch) are recounted as having
been cause, perhaps especially if occurring simultaneously with ecological perturbation, to have
resulted in migration to a different area.
The extension of the Rungwa Game Reserve in 1951, and the eviction of the farming peoples living
along the northern bank of the Great Ruaha River in 1954-5 (see Chapter 4), can be seen as the
beginning of government-mediated changes in how the landscape was to be occupied and used by
the Ilparakuyo and their farming neighbours in Idodi. Although extremely difficult to trace with the
Ilparakuyo themselves due to their sporadic movements, oral accounts by Nyambarazi and Kosisamba
191 The Kosisamba are a group who have been assimilated themselves as Hehe in the last 50 years, but who
previously had close links to the Gogo and Kimbu. They formerly lived in what became northern Uhehe, much of
which is now the Ruaha National Park
192 It is very likely that these Ilparakuyo did not farm and they will have purchased their grain requirements
from local Hehe, Gogo and Kosisamba farmers.
193 The terminology used to describe the complex and variable structure of Ilparakuyo homesteads and
households is discussed in Section 6.3.1. Here, an all encompassing term for homestead - enkang’ (or inkang’itie
pl.) is used for simplicity. However this term may not always be technically correct for all homesteads.
145
agro-pastoralist farmers living in the area at the time suggest that significant and adverse changes in
the disease challenge to livestock occurred concurrently with the creation of the new reserve.
People living in Idodi recount a noticeable increase in tsetse fly and the incidence of trypanosomiasis.
Oral histories collected by Jennings in Idodi and Pawaga in the early 1990s concur with those
collected during fieldwork. Colonial records show that the period of the 1940s to 1950s was
certainly characterised by an expanding tsetse front in the area, but local explanations for such a
phenomenon may also have further significance. Outbreaks of other fly-related parasites (which have
not been identified194) are remembered, which, together with tsetse advances, are associated by
people present at the time as occurring simultaneously with the extension of the Rungwa Reserve and
the effective creation of a ‘hard’ boundary between wildlife and livestock. As Ford (1971, 492)
remarks in his seminal work on the role of trypanosomaisis in African ecology:
Equally undesirable, on the other hand, are the activities of conservationists who have
often succeeded in the past in invoking the law to preserve inviolate as natural parks
and wildlife reserves known major foci of (human) trypanosomaisis.’… ‘The real
danger lies in their preservation not of wildlife as such, but of populations of
pathogenic trypanosomes.
Previously, before government intervention, the boundaries between livestock and wildlife had been
gradual and fluid, thereby probably facilitating the modulation and suppression of disease
transmissibility and challenge. The extension of the Rungwa Reserve, and then the creation of the
Ruaha National Park, resulted in the effective hardening of the boundary between wildlife and
livestock. Further, the resettlement of people as a result of protected area creation lead to an
increased concentration of people and livestock on the protected area boundary, in juxtaposition to
higher wildlife densities in the protected area immediately across the Ruaha River195. The relatively
high densities of wildlife and livestock in close juxtaposition to each other, particularly during the dry
season, may be sufficient grounds to explain, to some degree, people’s memories of a substantial
increase in disease occurrence in livestock during the 1950s and early 1960s. If this was the case,
then it is ironic that the then colonial Game Preservation Department, in an agreed policy with the
then District and Provincial colonial administrations, extended the Rungwa Game Reserve as a solution
for minimising human-wildlife conflict, both in terms of physical conflicts, and increasing
epidemiological threats196. A further reason - and probably the underlying and driving motive - for
the extension of the Rungwa Reserve (as discussed in Chapter Five) was the need to respond to the
demand from the then Governor of Tanganyika and his superior - the Colonial Secretary in London -
to create an ‘ adequate’ estate of nature preserves.
194 The description of one set of symptoms which were associated with a parasitic fly are ambiguous and cannot
be definitely used to point towards any particular affliction or other.
195 The Ruaha is, with some minor exceptions, the only perennial source of fresh water in thousands of square
kilometres in what is now the Ruaha National Park. Thus wildlife may be drawn in to water from substantial
distances during the dry season.
196 These were East Coast Fever in the southern Tanganyikan highlands, and in the lowlands, the expansion of
the tsetse fly belt southwards as far as the rift valley scarps which constitute the boundary with the highlands.
146
Thus, whereas previously the Ilparakuyo had been at liberty to move relatively freely in the
landscape, the extension of the Rungwa Game Reserve as a result of the developing land-use
planning policies of the then Colonial administration can be seen as the beginning of a series of
events, that were to precipitate substantial changes in Ilparakuyo livelihood strategies and landscape
occupancy.
6.1.2 Ujamaa and its consequences

The depopulation of greater Idodi through the induced movement197 of people to officially
‘preferred’ settlements during the 1960s had culminated in the abrupt Ujamaa evictions during 1974
and 1975 of the remnants of the farming population in Unyambarazi. In marked contrast to their
agro-pastoralist and farming neighbours, the Ilparakuyo living in the remoter areas of Idodi and
Unyambarazi appear not to have been compelled to move into Ujamaa villages. There are no
recollections of attempts to create the communal Ujamaa pastoralist villages that were created, for
example in Kisongo Maasailand, although pastoralist evictions from the south-eastern periphery of
the Ruaha National Park occurred in the late 1960s (Jennings 1994, 23), and indeed continue to
occur sporadically to the present day (Mtahiko pers com). Aside from their removal from the national
park, it would appear that the Ilparakuyo were overlooked or left largely to their devices, perhaps
since they were a minority group in lowland Uhehe198. A similar process is reflected in Rigby’s account
of the Ilparakuyo in Bagamoyo District (Rigby 1983, 38).
However, the Ilparakuyo who had chosen - some of them 20 years previously - to site their inkang’itie
near the now rapidly expanding Ujamaa villages were to face tenure challenges to the rangelands
they were using. The trickle of immigrant farmers now erupted into a flood of hundreds of evicted
and landless farmers requiring land as part of the villagisation process. The Ilparakuyo recount that
they had to make way for the new farmers and move to more marginal areas, as land was allocated
by the villages to these new farmers. Thus areas of rangeland previously used by the Ilparakuyo for
grazing their stock were converted to farmland, a trend that has since continued.
The depopulation of farmers from the outlying rangelands of Idodi left those Ilparakuyo still living in
these outlying areas without trading opportunities and services provided by the now defunct farming
communities. Most importantly, grain could no longer be purchased. Faced by growing socio-economic
remoteness, and perhaps, likely further increases in the disease threat to their livestock as wildlife re-
colonised the deserted farmland and its outlying areas, the Ilparakuyo began also to gravitate to the
periphery of the recently expanded Ujamaa farmlands. The villages were less remote and provided
trading opportunities, closer livestock markets, easier access to grain and basic but increasingly
popular medical services.
197 Advocated by the late colonial and early independence local administrations – see Chapter Five.
198 Many Hehe were no longer agro-pastoralist, having lost much of their livestock to disease (Jennings, 1994,
23).
147
A further factor contributing towards the Ilparakuyo migrating to the physical periphery of farming
communities was the creation of the Lunda-Mkwambi Game Controlled Area (LMGCA) in 1984199,
and the associated subsequent evictions in the late 1980s and early 1990s of the Ilparakuyo in the
Lunda section200. To the south-west in Idodi, local Wildlife Department officials were persuaded by
wealthy resident hunting interests to extra-legally201 evict pastoralists from the area in the interests of
securing their wildlife hunting prospects. Although the few documentary records (found in village
government reports to the District Game Officer) of these evictions are far from conclusive, oral
accounts by Ilparakuyo individuals nevertheless are. Pastoralists have remained in the Lunda North
section of the Game Controlled Area - however tenuously - more recently accompanied by farming
communities moving back to restart their old settlements post-Ujamaa. Yet in 2002, there was a
further round of evictions (Walsh pers com.).
There is strong cause to consider (as this Chapter attempts to demonstrate) that the depopulation of
the larger part of the Idodi rangelands and the more reason eviction of pastoralists from parts of the
LMGCA has led to substantial changes in the livelihoods and land-use practices of the Ilparakuyo. The
Ilparakuyo have undergone a relatively rapid and, for many, an arguably irreversible transition from
pastoralism to agro-pastoralism. Once leading a more semi-transhumant lifestyle, the Ilparakuyo now
maintain a predominantly sedentary way of life in a tightly bounded landscape. The arrival of the
Barabaig in the early 1980s from northern Tanzania was to further impact on the Ilparakuyo, their
landscape-use practices and their relations with the Idodi Hehe/Bena farming community.
6.2 The Barabaig in Idodi
The Barabaig pastoralists are a minority ethnic group in Idodi. I explain why the Barabaig came to
settle in Idodi as relative newcomers, and provide an overview outlining some of the key reasons why
their use of the landscape and their socio-ecological situation has led to poor relations with other
land-users, an increase in land-use conflict and their continued marginalisation. Although these latter
themes are taken up in more detail in Chapter Seven as part of a discussion about land-use conflict,
some key ethnographical and context-setting data is provided in this section, in relation to that
provided about the Ilparakuyo later in this chapter. Much of the information below, where not specific
199 Prior to the creation of the LMGCA, there had been an ‘Iringa Controlled Area’ (ICA) probably created in
1951 or 1952, under the Fauna Conservation Ordinance of 1951. It is thought that the ICA became defunct
after a number of years. Certainly there is no mention of the ICA during the gazettement process for the Ruaha
National Park in 1964 (pers com Walsh). The LMGCA was created by Government Notice No. 33 published on
1st February 1985 under the ‘Wildlife Conservation (Game Controlled Area) (Declaration) (Lunda Mkwambi)
Order, 1984’.
200 Pastoralist and farmer evictions in the Pawaga Lunda North section of the LMGCA have continued to occur
over the years, but these events have had a less-direct impact on developments in Idodi. More recently, the
authorities have especially targeted Barabaig herders who take advantage of seasonal grazing in the Ruaha
National Park.
201 This issue is also discussed in Section 3.5.
148
to Idodi, is necessarily derived from Lane 1996, as this is the most comprehensive and recent
ethnography available on the human and political ecology of the Barabaig people.
6.2.1 The arrival of the Barabaig in Idodi

The Barabaig are a sub-group of the Tatoga peoples who have lived in Hanang District in northern
Tanzania since the nineteenth century (Lane 1996, 1). The Barabaig first arrived in Idodi in 1984
(pers com Kipilipili) and many have since settled in the area on a permanent basis, becoming residents
of the Idodi villages. The arrival of the Barabaig now living in Idodi is part of a larger long term
trans-migration of Barabaig people from northern Tanzania southwards as far as south-western
Tanzania in search of new pasture and farmland. Lane (1996, 152) explains that the emigration of
the Barabaig from their homeland southward has been due to the long term loss of their land over
the last 50 years and more. Over this period, Iraqw agropastoralists have gradually expanded south
from the Mbulu highlands north of Hanang District, driven southwards by population pressure in the
highlands and attracted by Hanang District’s fertile land (Lane 1996, 152; see also Iliffe 1979, 351).
Thus over time the Iraqw have occupied some of the most fertile soils of the Barabaig plains as they
practise their intensive agriculture, and in surrendering access, the Barabaig have lost some of the
most productive elements of their grazing rotation (Lane, 1996, 153)202. A second factor has been the
expanding and shifting tsetse fly belts, which over the years have reduced the availability of pasture,
in concert with the Iraqw expansion. In response to these developments, Barabaig families began to
leave Hanang District as early as 1957, moving into Singida District in search of new pasture, in what
the Barabaig perceive as being an appropriate relocation response - that of moving effectively and
often (Lane 1996, 153-4).
However, perhaps most significantly during the early 1970s, a large extent of Barabaig customary
land (100,000 acres) on the Bastotu Plains was alienated by the state for developing large scale
commercial wheat farms as part of the Tanzania Canada Wheat Program (TCWP) (Lane 1996, 155).
This appropriation of land comprised almost the entire extent of the muhajega that was so important
for Barabaig pastoral production (Lane 1996, 155). The loss of this muhajega land, together with
constrained access to other grazing resources, comprised a substantial threat to the productivity of
Barabaig herds and led to the emigration of many more Barabaig families southwards.
Today, many of the Barabaig families living in Idodi are environmental refugees who left Hanang
District during the late 1970s and early 1980s, either directly or indirectly as a result of the socio-
environmental impacts of the Tanzania Canada Wheat Program (pers com Gissemoda). For example,
eleven Barabaig families living in the Idodi villages arrived from Hanang District in 1984 and a
further fourteen families arrived during the 1990s, although these latter families lived in other
rangelands, such as Usangu and Pawaga, before settling in Idodi. Many of the Barabaig now living
in the Idodi villages maintain a locally mobile lifestyle, moving their homestead encampments within a
202 Lane (1996, 153) explains that most Barabaig were often willing to move out of an area in the face of an
Iraqw intrusion as the Iraqw and Barabaig had together fought against the Maasai during the nineteenth
century, and also share common lineages.
149
localised area in response to changing grazing conditions. Often these movements follow a regular
seasonal range use pattern. Other Barabaig remain predominately transhumant, moving substantial
distances across Idodi and Pawaga in their search of grazing and water, and negotiating their access
to these resources when compelled to do so with local village governments.
6.2.2 The Barabaig domestic unit, development cycle and household economy
The Barabaig domestic unit can be described as usually consisting of a male household head
(gwarwa), his wives, sons and unmarried daughters living in a ged. Each ged consists of a thorn-brush
enclosure in which there are separate sub-enclosures for people (samod) and livestock (muhaled)203
(Klima 1970, 34; Lane 1996, 122). The ged requires a large amount of wood which is cut from the
surrounding bush or woodland. A firebreak is cut around the ged to prevent catastrophe from wild
fires that pass through during the dry season (Lane 1996, 123). Each wife has her own ga (hut) and
there is a men’s hut, the huland, in which all unmarried males over about the age of six live (Klima
1970, 37; Lane 1996, 3). A single gate (dosht) opens into the samod and mhalend enclosures. If the
household has enough members and sufficient livestock, it may be split across different sites, so as to
enable better access to grazing (Lane 1996, 42). Household work is split strictly along gender lines
(Lane 1996, 43). Women are responsible for milking, food preparation, child rearing, domestic
hygiene, preparation of skins, hut maintenance and collection of water, wood and thatching grass.
Men are responsible for herding, house construction, security and social organisation (see Section
6.2.4).
A ged will expand in size over the period of a development cycle as a man marries several wives to
whom a substantial number of children may be born. On reaching adulthood, and upon his first
marriage, a son and his new wife may initially live in his father’s ged, before moving away at a later
stage to found their own ged and larger polygamous family. On marriage, a woman will leave her
father’s ged and move into that of her new father-in-law’s, bringing her dowry of livestock with her.
These cattle will form part of the matrifocal herd that will be inherited by her sons. Thus different
members of the household may have different rights over the livestock in a ged depending on the
origin of the livestock and the nature in which the livestock became part of the ged herd (for
example, as dowry, bride wealth, inheritance or gifts) (Klima 1970, 41 & 67-71). This means that
although the household head is responsible for managing the herd, he is not entitled to sell his wife’s
cattle without her permission. Male children are given their first cattle as a gift on the eruption of their
milk teeth, and it is from these and subsequent gifts of cattle at different stages of their life that
young men build up a foundation herd. After the death of a lineal head, the ged will usually split up,
each widow either returning to her fathers’ ged or moving away with her remaining children, perhaps
to live with a married son204 (Klima 1970, 107).
203 There may also be a smaller livestock enclosure for sheep, goats and claves (jaboda muhog).
204 Klima (1970, 107) makes reference to leviratic unions as also occurring in order to prevent the dissolution of
the family and cattle herd. It is not known how common this practice is today.
150
Cattle play a central part in the household economy and the social life of the Barabaig - for
example, in their role as providers of milk, as a source of income, and as an important part of social
ceremonies and jural process (Klima 1970; Lane 1996). The structure of the household cattle herd is
managed for the production milk, and herd cows may have a long and productive life, whereas male
stock, other than selected individuals, may be sold or passed out of the herd relatively quickly. Cattle
are usually only consumed as part of a social ceremony or if they are about to die, in which case the
meat may often be shared with neighbours (Klima 1970, 42). Goats, sheep and chickens may also be
kept. As is the case for many pastoralists in Tanzania, some Barabaig are farmers205 as well as
herders, cultivating grain (maize and sorghum) to supplement the production of their herds (Klima
1970, 13; Lane 1996, 74). Lane (1996, 64) describes the Barabaig families that he studied in
Hanang District during the late 1980s as being relatively food self-sufficient, with just twelve per cent
of total household expenditure spent on purchasing food. Although farming may provide the bulk of
food consumed in the household, cattle are the main source of income for the purchase of food, other
commodities (such as honey, veterinary drugs and tobacco), and livestock. Lane (1996, 67) describes
the Barabaig cattle herds in Hanang as likely to be growing only very slowly due to a combination of
high mortality and the need to sell cattle to generate a cash income. Thus it is not surprising that Lane
describes the Barabaig as, ‘…active and astute marketers who invest in livestock and maximise
economic and social returns’ (Lane 1996, 67).
6.2.3 Barabaig herding in Idodi

Barabaig herd zebu short horn cattle (Bos indicus) that are a relatively small but hardy breed. These
cattle are able to tolerate harsh conditions, walk long distances, survive for relatively long periods
without water, and still produce milk, albeit in relatively small quantity, but rich in butter fat (Lane
1996, 45). Although cattle are predominantly herded by men, women may help when there is a
shortage of labour. Children herd calves and often goats and sheep separately close to the ged. The
Barabaig view cattle as ‘grazing themselves’, as the herd is led to a general area and allowed to
graze as they choose (Lane 1996, 45-46). The rangelands of Idodi receive a slightly smaller amount
of rainfall than the Bastotu plains (average 644mmyr-1). As with the rangelands of Hanang, the Idodi
rangelands are prone to both wetter and drier years. Both Hanang and Idodi have some similarities
in the types of vegetation that occur, particularly in regard to two types of woodland. Acacia-
Commiphora and Brachystegia (miombo) woodlands occur on the Barabaig plains and across much of
Idodi. In Hanang, as also in Idodi, the diversity of grazing resources in space and over time requires
the Barabaig to practise transhumance (Lane, 1996, 108).
In Idodi, resident Barabaig families follow a similar pattern of range-use to the Ilparakuyo. Towards
the end of the dry season, as the first rains begin to fall in the hills to the south, the Barabaig take
205 Few Barabaig families living in northern Mbulu District in the Eyasi basin have cultivated much grain for at
least the last 30 years because of the aridity of the area and the need to move livestock large distances to
forage and water (Sieff 1995, Tomikawa 1970 & 1972 cited in Sellen 2003, 534). In comparison, a much
greater proportion of Barabaig living further south tend to farm - for example, in the Basotu area and the
Barabaig plains (Lane 1996, 64).
151
their stock up to the uplands to exploit the new growth. As soon as the rain arrives down in the Idodi
villages of Tungamalenga, Mapogoro and Idodi, they move their stock down usually in early to mid
December depending on when the rain arrives. The herd may then remain in the village commonages
until the head of household decides that the ectoparasite load has reached too high a level, at which
point the herds are moved to Malinzanga village for wet season grazing206. Once the wet season
grazing is depleted, the herds are moved back to Idodi, Mapogoro and Tungamalenga, although
many Barabaig families favour living in proximity to each other particular parts of Idodi village207.
This is likely to be because the neighbourhood is an important unit in the development and sustaining
of community cohesion (Lane 1996, 143). The herds are then grazed on farmers’ field crop residues
and in the Acacia-Commiphora and Brachystegia woodlands around Idodi for the rest of the dry
season. This latter period is frequently a tense time for the Barabaig as they negotiate access (and
sometimes not) to farmers’ fields and their crop residues. Disputes between Barabaig and farmers
often occur, and in recent years have led to outbreaks of violence. Although discussed further in
Chapter Seven, tension between Hehe/Bena farmers and the Barabaig has led to the reinforcement
of stereotypes of the Barabaig as being backward, belligerent and inconsiderate range-users among
villagers and local government officials (see also Lane 1996, 30).
6.2.4 Barabaig and farmer land-use relations

The highly mobile nature of Barabaig land-use has tended to result in a lack of strong social relations
with local farming communities in Idodi. Although Barabaig families return to particular localities in
Idodi in their seasonal use of the Idodi rangelands, they may only be present in these particular
areas for relatively short periods of time. In addition, and in comparison to the Ilparakuyo, there has
been much less need for the Barabaig to enter into client - patron relations with farmers, since their
production systems have remained mobile and largely separate - except for the increasing case of
dry season field grazing. Nevertheless, particular Barabaig personalities are well known in Idodi,
although even some of these resident Barabanda (Barabaig: Barabaig individuals) have often been
at the centre of conflict over grazing practices.
Barabaig families mostly delegate livestock herding responsibilities to their young men (e.g. Lane
1996, 45) who have sometimes driven herds of cattle into farmers’ fields without the consent of
farmers, and sometimes before the fields have been harvested. Rice field irrigation furrows and
standing crops have been damaged or sometimes destroyed. Farmers claim that Barabaig cattle
incursions into their fields have been purposefully and repeatedly perpetrated, despite complaints to
both the Barabaig and village government, often to little avail. Farmers have responded to the cattle
presence by burning their fields immediately after harvest, as well as purposefully setting fire to the
surrounding rangeland areas in the early dry season in an effort to dissuade Barabaig pastoralists
from grazing their stock in the area. This has led to an escalation of dispute and sometimes violent
206 The Korongo and Nyamdawe areas about ten kilometres north east of Malinzanga village are favoured.
207 These are the Kibikimuno and Nyangolo areas of Idodi – which feature further in Chapter Seven.
152
conflict as Barabaig have then had to rely even more on farmers’ fields during the dry season as
sometimes large rangeland areas are burnt.
6.2.5 Social organisation, property relations and jural institutions

Every person born into Barabaig society, is a member of a clan or dosht (Barabaig: gate), that is
unified by male lineage traced to a common founding ancestor (Lane 1996, 2; Klima 1970, 39). All
clans are exogamous (with one exception - the blacksmith clan) and are divided into spiritual
(daremng’ajega) and secular (homtka) clans (Lane 1996, 2; Klima 1970, 81). The clan is the largest
unifying and corporate group, and is an important level at which social, political and jural affairs are
mediated (Klima 1970, 80-87). Genealogical status is an important part of Barabaig everyday
identity (Klima 1970, 83).
Property ownership exists at the private (family), clan and ‘tribal’ level. The ged and its immediate
vicinity (which may include particular shade trees, a small delimited grass reserve for young stock,
and other milking, shade and rest areas) is perceived as being the private property of the household
(Lane 1996, 136). A farm plot that may lie some distance off is also the private property of the
household. Other Barabaig herders are expected to respect these areas and their resources and
refrain from using them. Outside the vicinity of the ged lies a range of other resources which may be
owned by the clan208 - such as particular wells, ritual sites (often associated with funerary monuments
and reserves) and certain trees (Lane 1996, 136-138). These latter features are absent in the Idodi
landscape. Lastly, in their homelands, the Barabaig perceive ownership of their rangelands as a
group, and have in recent years set about to defend their customary rights against the state’s
alienation of land by convening a special committee (see below).
Lane (1996, 2) describes jural authority over land and property as occurring at the ‘tribal’
(Gedahanangwed), clan (dosht) and neighbourhood (gisjeud) level209. Land relations are mediated
208 In former years, clan territories existed.

209 There are five key institutions that are important in Barabaig society. The first is the ‘tribal’ assembly
(Getabaraku – ‘meeting of the wide tree’). All adult males may attend, and although the meeting institution has
no office bearers, hierarchy is determined by skills of oratory, knowledge, wisdom and status from involvement
in the issue being considered. A committee of elders makes the final decision on issues that are debated by all
attending (Lane 1996, 138). The getabaraku is the ultimate authority on matters of common property land rights,
water and trees although its jurisdiction does not extend to authority over clan property (Lane 1996, 139). A
makchamed may be selected from the ranks of elders to meet in camera to consider matters of gravity or
wrong-doing. The secrecy of the deliberation protects makchamed members from personal association with
decisions made, and enables more effective investigations to be made. The oath on which evidence is given
remains particularly effective (Lane 1996, 140). Additionally, a makchamed may be convened by two or more
elders to address a particular issue independently of a getabaraku, as has been the case for the Barabaig’s
long-term campaign to regain alienated land from the state. An important and powerful institution is the
women’s council (girgwageda gadmeg). The council is often concerned with issues at neighbourhood level, but
may also play an important role more widely due to women’s special role in Barabaig spiritual life, for example
153
most strongly at the clan level through clan moots (hulandosht). These are courts made up of all adult
male members of the clan, that deliberate on conflicts over access to clan property, particularly in
relation to matters of marital exogamy and property in livestock (Klima 1970, 84; Lane 1996, 142).
Deliberations are led by the clan head (ashohoda dosht) who is the senior living male descendant of
the clan founder (Klima 1970, 83). There are no jural relations between clans (Klima 1970, 86; Lane
1996, 143).
Disputes of private property are resolved by a ‘council of neighbours’ (Girgwageda gisjeud) convened
by any local elder who proposes the meeting (Lane 1996, 143). The ‘council of neighbours’ plays an
important role in mediating matters of public concern, such as controlling and quarantining a disease
outbreak in a ged, resolving domestic disputes (in private session), sanctioning members of the
community mostly through fines of honey (due to its ritual purity) and regulating the arrival of
newcomers to the area (Lane 1996, 143). There is an understanding that large herd owners should
not live in the same neighbourhood, as this would put stress on the grazing resources (Lane 1996,
143).
6.3 Ilparakuyo landscape occupancy
The Ilparakuyo today comprise a diaspora stretching from north-eastern Tanzania to the south-
western extent of the Usangu wetlands. In Idodi, they live on the peripheries of the Hehe/Bena
farmlands in sites that have come to be recognised by these farming communities as ‘umaasaini’ - the
place(s) of the Maasai.
As discussed in Chapter Four, the Hehe still perceive the land on which the Ilparakuyo live -
‘Umaasaini’ - as collectively belonging to their (Hehe) domain. Thus the Ilparakuyo (and the Barabaig)
ultimately have little de facto defensible tenurial right in the areas in which they live, although they
are able to maintain and defend tenurial rights to farmland through fields which they have bought or
rented. The lack of tenurial security for the Ilparakuyo (and the Barabaig) living in the village
commons that comprise umaasaini is also in part due to their customary practices. As Rigby (1983,
136) comments, customarily the Ilparakuyo do not conceive of land as being owned, and thus until
recently there has been little precedence for claiming tenurial right to rangeland for grazing.
However, as discussed in Chapter Seven this situation is starting to slowly change, as the Ilparakuyo
struggle to secure sufficient access to grazing in Idodi.
The Ilparakuyo have established and repeatedly settled particular enkang’ sites over the years,
usually not far from village settlements. These homestead enkang’ sites have become popular with the
Ilparakuyo, having developed as result of a complex interplay of considerations. Today these may
include access to grazing, the availability of water for the household, the distance to any owned or
rented farm fields, established labour networks, and who else is living in the vicinity (this may be
in mediating socially appropriate access to land and in their jurisdiction in matters involving offences by men
against women (Lane 1996, 141).
154
particularly important in terms of a family’s perception of their magical security in the landscape [pers
com Mtemisika]). Perhaps even more significantly, as discussed in Chapter Four, proximity to farming
settlements has provided the Ilparakuyo with increasingly integrated relations with Hehe/Bena
farmers that have grown out of trade and labour relations and client-patron relationships. To an
extent, both farmer and herder have become increasingly reliant on each other for trade and labour
opportunities and many friendships have arisen out of such client-patron relations.
6.3.1 Pastoral ecology and range use patterns

The rangelands of Idodi, as discussed in Chapter Four, consist of a mosaic of different vegetation
communities comprising of Brachystegia woodland on the higher hill slopes bounded lower down by
relatively narrow bands of Combretum woodlands which quickly shift into grassed Acacia-
Commiphora bushland at the base of the rift valley (Nahonyo et al, 1998). The Ilparakuyo tend to
rely most on the patchy grasslands, woodland glades and herbaceous undergrowth of the Combretum
and Acacia-Commiphora communities for their grazing requirements, and only more occasionally do
they make seasonal use of the higher Brachystegia woodlands.
The Ilparakuyo retain a substantial amount of flexibility in the use of the rangelands - principally
driven by the availability of grazing and water contingent on rainfall. While homesteads remain
more-or-less permanently placed in a locality, the herd, less young calves and a limited number of
milking cows, may be seasonally moved a substantial distance to the homestead of kin or even to a
temporary enkang’, either to take opportunity of seasonally better range conditions, and/or to avoid
increased seasonal disease risks in any particular year.
Dependent on range conditions and if possible, movement of a herd tends to be avoided late in the
dry season as other time and labour demands begin to compete with those of herd movement,
grazing and guarding - activities that away from the homestead locality tend to be seen as the
responsibility of the ilmurrani (Maa: warrior; sing ilmurran) age set. Such other demands are, for
example, arranging the renting of agricultural land, helping to organise labour and carrying out field
ploughing and planting.
The rangelands of Idodi are valued by the Ilparakuyo in regard to the quality of grazing available -
dependent on transient flushes and qualities of grasses, forbs and browse plants, the seasonally
varying parasite and disease risks as well as the ground conditions underfoot. In a year that receives
an average amount of rainfall, a generalised range-use system may be described as follows:
In the late dry season during the months of late October and November, there is a tendency for the
upland Brachystegia woodlands, to receive the first rain of the wet season up to six weeks before the
main rains arrive in the Ikwavila valley and Kwigongo lowlands below. This occurs especially on the
Mguhu, Chamyina and Chamgong’onzi hill ranges together with the uplands to the south including
Ngombaguli. The grass flush that ensues is seen as a critical nutritional opportunity by the Ilparakuyo
for their herds - especially in years when grazing resources in the lowlands have become heavily
depleted and the nutritional value of the grasses and forbs has declined. Ilparakuyo households will
155
cooperate and join together in moving their main herds up onto the upland slopes returning to
temporary enkang’ sites that may have been used the year before or the last time they had occasion
to take advantage of the grass flush to economise on labour requirements.
As soon as the first rains arrive in the lowlands, the herds are moved back down into the lowlands to
avoid increasing seasonal disease risks in the upland areas and to accommodate pressing
commitments in cultivating the year’s grain crop. The lowland grasses, once they begin to flush, are
also perceived as more nutritionally efficacious than the upland. Thus, with the arrival of the rains in
the lowlands, the nutritional security of the herd is reasonably assured for the next 5 months as the
herds take advantage of the rapid grass flushes. As the wet season ends and the dry begins, those
households with larger herds may divide the herd and send some or most of it to the Mlowa lowlands
where the green, more extensive grass flush often remains for longer.
By the end of July, the herds have begun to be brought back to the homesteads to graze on the now
harvested fields of maize ‘mabua’ (Swahili: the residue of harvest crops - usually maize) and other
crop residue - for example legumes such as groundnuts. Access to each field is frequently negotiated
between herder and farmer most often for cash and it is on these fields that the cattle depend for
much of their nutritional requirements for the rest of the dry season. Once there are no longer any
mabua in the fields, the Ilparakuyo are forced to graze their herds in the marginal grazing areas that
are left. The interval between the depletion of the mabua and the onset of the rain in the uplands is
the most nutritionally stressed period of the year for the herds, and their condition continues to
deteriorate until the arrival of the next season’s grass flushes.
There are a range of grass, forb and tree/shrub species that are recognised as being the most
nutritious and beneficial to herd health - see Box 6.1. Many of these species are perceived as having
temporal peak productivities as well as nutritional values and thus the grazing strategies of especially
the more knowledgeable herd mangers may frequently aim to take greatest advantage of these
peaks. Knowledge of the localities of where favoured grasses occur is key to maintaining the
condition of the herd especially towards the end of the wet season when stock need to have
accumulated as much condition as possible before the onset of the seven to eight month dry season.
The annual movement of Ilparakuyo (and Barabaig) herds to Mlowa to take advantage of its late wet
season grass flush is of note in this regard. The condition of female cattle may be critically important
for calf mortality in the latter stages of the dry season when there is a tendency for the many of the
mature female herd to be near term and close to calving - which usually occurs from the middle of the
wet season onwards, tapering off some way into the dry season.
156
Figure 6.1: Grazing farm fields during the early dry season
Figure 6.2: Chamyina mountain and highland scarps looking south from Idodi village rangelands near Kibikimuno
157
Box 6.1: Some of the more important grass, forb and tree/shrub species for Ilparakuyo
livestock in the Idodi rangelands
Important grass species:
Orrmbuliesh - a nutritionally important and popular wet season grass growing in the lowland
grasslands up to a metre high but poorly tolerating dry season grazing, drying out and withering
quickly after the end of the wet season.
Orrarobai - a dry season lowland staple that sprouts in the wet season but has unpalatable barbs
that recede as the grass dries. Grows up to a metre in height.
Orrkerien - a lowland riparian grass, growing along stream banks and especially important for
grazing during the dry season when it may grow up to 2 metres high. It grows especially well in the
wettest areas.
Orratemii - an upland grass important in the early wet season when its new flush is relied upon for
the new season’s grazing.
Other species identified as being of lesser importance include:

Orrkawai - a widespread wet season grass particularly occurring in the lowland grasslands and
swampier areas
Orrngramnatia - particularly occurring in the grasslands, of grazing significance during the wet
season but its flush carrying over into the early dry season
Orrikirian - a grass occurring on the lower slopes of the uplands as well as in the lowlands. A
moderately significant species during the wet season, occurring extensively in the Mlowa area.
Growing to about a metre in height, this grass is a wet-season grass and rapidly dries back during
the dry season.
Orroieti - an uncommon grass but occurring in river flood-recession areas, A creeping grass with a
restricted occurrence, it grows to about a metre tall.
Orrlkijuta orruki - a grass that is grazed in the early wet season while still short but growing to
over 3 metres in height, the grass becomes unpalatable and thus is avoided when mature.
Important tree and shrub species include:

Oljasilalei - Acacia albida - a large tree which bears orange and purple seed pods which when they
fall are grazed by livestock and are perceived to be very nutritious.
Oldabesi - Acacia sieberiana - a medium sized tree with black bark which produces a large crop of
small fruit which are green immature and beige upon ripening.
Olndundulu - Acacia stuhlmannii - a scrub-like species which produces red fruit when ripe early in
the dry season. This species is not as important in terms of cattle fodder as compared to the Acacia
sieberiana and Acacia albida.
Olmakerekala - Premna resinosa - a deciduous tree dropping its leaves during the early-mid dry
season becoming an important source of fodder mainly for small stock.
Embili - Spp unidentified - a short scrub tree the leaves of which are browsed by goats during the
wet season flush
Mbomboi - Spp unidentified - a creeping shrub with white flowers especially important as wet and
early dry season graze for cattle.
Oljogirr - Spp unidentified - a creeper which can grow on trees occurring in the low-lying wetlands
which is browsed in the early dry season.
Descriptions given by Salum Mtemisika Mtango
6.3.2 Livestock disease and range-use patterns

The Ilparakuyo herds perhaps most critically are continually exposed to two diseases of special note -
Trypanosomiasis (Swahili: Ndigana) carried by the tsetse fly (Maa: liipis enado [red/brown Glossina
morsitans?] or lipiis enarok [black Glossina pallidipes?]) and Contagious Bovine Pleuropneumonia
spread directly between animals. The Ilparakuyo herds have been locally exposed to trypanosomiasis
for many years (since at least the 1930s when the tsetse fly belt was recorded as having expanded
south into Idodi and Pawaga). Traditional treatments and prophylaxes are known and, when
practised together with the appropriate application of modern drugs, appear reasonably effective in
reducing herd mortality. However, the herds’ exposure to Contagious Bovine Pleuropneumonia (CBPP)
is much more recent, locally associated with the arrival of the Sukuma who may have brought the
158
disease with their stock from north-central Tanzania. While the symptoms of this disease are widely
recognised by Ilparakuyo, and are treated with the use of antibiotics, the epidemiology of the
disease remains poorly understood by them. The initial arrival of this disease in the early 1980s is
associated with much increased general herd mortality and morbidity210. An annual vaccination
campaign begun in the late 1980s as part of the District’s response to controlling the disease has not
been particularly effective as it has been severely compromised by the somewhat less than total
presentation required of all cattle for inoculation211 necessary for CBPP control. The disease remains
extremely difficult to control with the movement of stock between vaccinated, partially vaccinated
and unvaccinated herds maintaining reservoirs of disease carriers.
Both these diseases are locally largely spatially and seasonally unavoidable and although the
Ilparakuyo avoid use of areas with noticeably higher densities of tsetse - for example, the Mkupule
area - they cannot avoid continual exposure to lower densities of tsetse fly endemic to the Acacia-
Commiphora bushlands. However, seasonal disease problems - such as hoof rot (Maa: orkuluk) 212 -
may be avoided by temporarily moving the herd to a drier area. A further and seasonally varying
impact on livestock morbidity and productivity is tick-borne (Maa: olmeherr sing. ilmeherr pl.) disease
such as East Coast Fever. Currently the Ilparakuyo in Idodi do not have access to dip facilities (and
have not had) although some ilmarei will use back-sprayers (sometimes borrowed from farmers) to
spray cattle with an acaricide when the tick load is perceived as having become acute. Disease
constitutes a substantial factor in continued herd morbidity and mortality, and disease outbreaks
associated with a specific vector or set of climatological conditions, may impact on inter-annual range
use patterns from time to time. Ilparakuyo also report their livestock as suffering from anaplasmosis
(Maa: olodokolak) and being irritated by biting flies (Maa: nemerigesh) and mosquitoes (Maa:
ngojung’wani). Divisional veterinary records report that other diseases such as liver fluke, lumpy skin
disease and brucellosis are present in the area (see Table A3).
210 The authorities have long recognised the significance and importance of CBPP’s spread southwards through
Tanzania by declaring district quarantines as a means to halting the movement of livestock between the most
affected districts. The quarantines have remained difficult to enforce and intermittently broken by itinerant
pastoralists and cattle traders sometimes with the collusion of village and other officials.
211 Largely due to a lack of understanding of the significance of the vaccine but more generally as a result of
suspicion on the part of Ilparakuyo as to the motive for treatment given that the livestock extension staff have
been compelled in the past to act as stock revenue collectors. A bout of adverse immune reactions to the vaccine
immediately after inoculation in 2000 in the herds of some Ilparakuyo was the cause of much complaint and
further suspicion. However this phenomenon has not re-occurred.
212 Hoof rot/ foot rot is a recognised pathological condition caused by various organisms – principally bacterial
but also variably in association with a range of nematode and arthropod parasites. The condition is contagious
and often occurs in hot humid circumstances (generally at the onset of the rains). It is extremely difficult to control
but may be curtailed by shifting infected stock to drier under-hoof conditions (as the Ilparakuyo are very aware)
– see Hall (1977: 119) for further information.
159
6.3.3 Range agreements, range tenure and farm fields
While the ephemeral quality of the range is key to how Ilparakuyo manage their herd movements,
the presence or absence of land-use agreements between the Ilparakuyo, Barabaig and the
Hehe/Bena may also have a substantial impact on range use. While farmland areas are generally
accepted as being proscribed to livestock - a prohibition arising from a variety of rules and decrees
at District, Divisional, Ward and Village government levels, this does not prevent farmland-use during
the dry season for grazing as informally agreed between farmers and herders. Some general village
land-use agreements have been developed between village governments and Ilparakuyo herders
that designate exclusive herding and farming areas - as discussed further in Chapter Four and
Chapter Seven.
While village farmer-herder land-use agreements are starting to modulate herder landscape
occupancy, there is a further dimension to rangeland-use and tenure consisting of the interaction
between Ilparakuyo and Barabaig herders. While both groups of these herders seek to take
advantage of seasonal fluctuations in the quality of the range (characteristic of arid and semi arid
lands), they may frequently compete with each other for grazing resources. As discussed in Chapter
Four, the non-cultivated rangeland areas fall under a number of tenure categories. The village
commonages, which pastoralists depend on, are not perceived by village councils and village
assemblies as formally belonging to any collective group or individual in the village. Currently in
Idodi, since pastoralists have yet to obtain formalised rights over village commonage, the presence of
an enkang’ in a particular locality does not confer de jure precedence over the right to use the
surrounding range. However, the presence of the enkang’ and the fact that a herd is locally resident
confers some de facto degree of tenure over the range, although this remains varyingly non-exclusive
and subject to use by others. It is up to the incumbent herders - Ilparakuyo and Barabaig alike - to
voluntarily co-operate or otherwise in range use and management.
The Ilparakuyo (and the Hehe/Bena) generally view the Barabaig as owning much larger herds.
Many Barabaig continue to follow highly transhumant range use patterns. This enables Barabaig
herders to successfully take advantage of the mosaic of temporal and spatial graze biomass
production moving their herds at short notice substantial distances when range conditions require. This
high reliability strategy (Roe et al. 1998) enables the Barabaig to move into a range area and
graze it thoroughly with a relatively large herd before moving on to another grazing area. The
Ilparakuyo, being much more sedentarised due largely to farming commitments, tend to rely more on
localised range resources and only after they have become depleted will they move their herds to
alternative range areas - if possible. The arrival of an itinerant Barabaig herder in a locality can
severely and adversely impact on an Ilparakuyo herder’s grazing plans when the range identified as
being sufficient for the herd’s requirements is rapidly depleted by the Barabaig herder’s often larger
herd. An Ilparakuyo herder’s knowledge of local grasslands and their performance in annually
varying climatic conditions is critical to the successful maintenance of the overall nutritional security of
the herd. The current lack of range-use agreements between Ilparakuyo and Barabaig herders is
likely to negatively impact on the effectiveness with which the range can be used and managed by
the Ilparakuyo - who are most disadvantaged since they tend to mange their herds at smaller spatial
160
scales than the Barabaig. The insufficient range-tenure security currently experienced as a whole by
the Ilparakuyo is therefore of increasing significance for Ilparakuyo pastoral production and
landscape occupancy strategies. A practical example of the negative impacts of tenure insecurity is
illustrated by the fact that while some Ilparakuyo herders still practice the habit of retaining calf
grazing reserves213 - areas of range close to the enkang’ which are left un-grazed for dry season
grazing by milking cows and calves, the habit is not widespread. A key factor identified by some
Ilparakuyo herders for the lack of grazing reserves was their insecure nature - a result of the high
probability of a reserve being grazed clean by another herder without recourse to any recognised
and enforceable form of rights over the area (pers com. Mtango and Mgemaa).
Farm fields are a key dry season grazing resource, and although access to these fields by herders is
usually negotiated for a fee with their owners, field grazing may frequently be a source of tension
and conflict. In part this arises from accidents where cattle may stray onto other, sometimes
unharvested, fields. When this happens, the field owner will claim compensation from the owner of the
straying cattle. Farmers accuse herders of often wilfully or carelessly letting their cattle graze where
they please, leading to crop and/or field damage. Ilparakuyo herders often claim that such incidents
are relatively uncommon, and when they do occur, farmers try to exploit the situation by demanding
too much compensation. Another source of tension around crop-residue grazing in fields centres on soil
fertility. As discussed in Chapter Five, some farmers believe that livestock can adversely impact the
fertility of their fields due to the clouds of dust raised, which they say, blows away the fertility (i.e.
the clay content of the soil). Some farmers also believe that cattle can destroy the structure of the soil,
compacting and making it more difficult to cultivate. Few farmers think that the manure left by cattle
grazing their fields makes a significant increment to the overall fertility of the soil. Finally, cattle are
less commonly allowed by farmers to graze their harvested wetland fields as the irrigation furrows
are prone to being damaged. In villages where wetland field grazing may be a more common
occurrence, for example in Idodi village, outbreaks of dispute and even violent conflict between
herders (particularly Barabaig graziers) and farmers have been much more frequent (see Chapter
Seven). Many Ilparakuyo herders tend to avoid grazing their livestock in wetland fields, due to the
heightened risk of dispute occurring, which may often involve costly compensation payments to
farmers and also adversely impact on their social relations with the wider farming community (see
Chapter Seven).
6.3.4 Farming
As the Ilparakuyo settled in the farmlands of Idodi, they began to depend increasingly on farming
their own crops for their grain requirements, employing labourers to undertake the work. From the
late 1960s onwards and through the early Ujamaa period, as the Idodi farmlands began to be
increasingly settled by migrant and translocated farmers, a tendency of sowing old enkang’ sites with
maize using employed labour began. The use of old cleared enkang’ sites (Maa: olmuaate sing.;
213 These reserves should not be seen as necessarily constituting a particular permanent area but as a
temporary area that may be identified as being particularly appropriate for a grazing reserve in a particular
year given the extant climatic and tenure-security conditions
161
ilmuaateni pl.) is significant in that the Ildareto214 age set of the elder age class claim a time before
Ujamaa when as ilmurran they did not practice farming to any extent215. It is also probable that
during the years before they began farming, Ilparakuyo families were in the habit of purchasing
rights to the crops of Hehe/Bena farmers’ fields as occurred in Bagamoyo District (Rigby 1983, 187).
As the farmlands became more densely settled and as less transient relations developed between
farmers and more sedentary Ilparakuyo herders, the practical possibility arose for Ilparakuyo herders
to copy farming techniques and access a growing labour market with which to cultivate their fields.
As with the Hehe/Bena farming community, the Ilparakuyo perceive farmland broadly as constituting
either nchi kavu - or bonde. The Ilparakuyo gain access to farmland through a variety of ways. They
may often own plots that may be purchased from Hehe/Bena farmers, or sometimes allocated by the
village government. They may also rent or borrow fields from Ilparakuyo and Hehe/Bena friends and
affiliates. Fields belonging to Ilparakuyo households are indistinguishable in the patchwork of plots
that makes up the lowland farmlands. An exception occurs with crops planted on old enkang’ sites
that are almost exclusively the cultivation of Ilparakuyo. An Ilparakuyo family may have access to a
number of fields that may be spread some way apart, but are most usually clumped around the
enkang’ site and within easy walking distance. Fields that are likely to be further away are those that
have been especially sought after and may often be rice paddy bonde fields. In contrast to
Beidelman’s (1960, 254) observation from the 1950s that the Ilparakuyo did not farm at all216 -
despite traded grain being an integral part of their diet, today the Ilparakuyo not only farm, but
may also sometimes carry out the manual labour required themselves.
Ilparakuyo landscape occupancy strategies are heavily and increasingly bounded by farming
commitments, largely distinguishing them from the strategies of the Barabaig, many of whom do not
farm, although a number of long-term residents increasingly do so. The Ilparakuyo are now effectively
sedentarised and heavily agro-pastoralist. Many Barabaig remain transhumant for part or all of the
year and are able to retain the greatest flexibility in their landscape-use.
6.4 Livelihoods
Having established how the Ilparakuyo (and the Barabaig) arrived in the Idodi rangelands and the
nature of their landscape use practices, the rest of this chapter is taken up with a quantitative
overview of current Ilparakuyo livelihood strategies and practices. The analysis is based on data
gathered through homestead surveys and information derived from a series of informal interviews
214 The Ildareto age set was opened in Idodi and Pawaga in the mid 1950s and closed in 1974. It should be
noted that the opening and closing of age sets in Ilparakuyo society is geographically staggered and is not
entirely synchronous with the age set developments of the Kisongo Maasai in the north. The Kisongo Maasai
usually open and close age sets before their closest Ilparakuyo neighbours who, in turn, have less of a time lag
than those living furthest away in the south and south-west who have a delay of up to several years.
215 In fact, informants were quite unequivocal about the absence of farming in Ilparakuyo before the late 1960s.
216 Beidelman concedes that many historical sources exist which make reference to the Ilparakuyo farming prior
to the period he was describing in the 1950s.
162
and discussions with Ilparakuyo ilmurran and ilpayani (Maa: junior elders; sing: olpayian). The data are
presented in the context of the continued transition of local Ilparakuyo livelihood strategies towards
increasingly agricultural and sedentary modes of production - especially as livestock holdings, while
still important both socio-economically and in terms of their socio-cultural ritual significance, continue
to decline.
6.4.1 Households
Maa terms for household and group household do not directly transliterate into English - for example,
see Bekure et al. (1991) and Coast (2002). In Idodi the structure of Ilparakuyo households is variable
and thus not easy to characterise at a single level. This is because polygamy is the norm in Ilparakuyo
society and their domestic units reflect this. Most domestic units are multi-hearth households - ilmarei
(Maa: family; olmarei sing.), made up of individual houses (Maa: enkaji sing; inkajijik pl.217), each
owned by a woman married to a usually polygamous husband. In some regards the house of a co-
wife, in which she lives together with her children, constitutes what may normally be defined as a
single-hearth household - as for example, in Chapter Five. Since the co-wives of a polygamous multi-
household live in the same compound, they may often share domestic tasks, such as livestock kraaling,
child care and water and firewood collection. However, two or more multi-hearth households - termed
here for clarity as ‘super-households’ - may also live together as a group in a larger homestead
(Maa: enkang’ sing.; inkang’itie pl.) and also varyingly share herding and domestic tasks. For the
purposes of clarity, two or more super-households living together in a single homestead are termed
here as ‘group households’. Super-households living in a larger group-household may be related but
not necessarily so. In Idodi, group-households comprise a third of all homesteads, with most super-
households living separately but in close proximity to others.
In Chapter Five, a household is defined as those people sharing a common hearth, also usually limited
to cohabitation under one roof. However, with higher levels of polygamy in Ilparakuyo society as
compared with Hehe/Bena society, the level of the household becomes less useful analytically when
labour sharing between households is taken into account and also when husbands, brothers and other
affines may not always eat or sleep in the same household each night. Instead, the super-household is
more analytically relevant, and is therefore given greater emphasis in the following description of
Ilparakuyo livelihood strategies. The data cover all Ilparakuyo super-households in Idodi with the
exception of Mlowa and Nyamahana villages. However the data, as collected, does not distinguish
between the different households in each super-household in terms of, for example, the numbers of
livestock allocated to each household, or the fields that one or more households may have cultivated
together as part of the larger super-household.
A domestic unit’s size and composition is dependent upon its position in the developmental cycle. For
example, a junior olmurrani218 (Maa: warrior; ilmurran pl.) may establish an olmarei in his father’s
217 Some of these Ilparakuyo Maa terms differ in form to Kisongo Maa reflecting the variety of dialects in the
Maa language (see Sommer and Vossen 1993).
218 The age grades and age sets of the Idodi Ilparakuyo are presented in Appendix Five.
163
enkang’ upon his marriage. The junior olmurrani will continue to use his father’s cattle byre and the
same gate (Maa: enkishomi sing.; inkishomitie pl.) into the family enkang’ as his father, his father’s
wives, children and other agnatic and matrilateral relatives. Initially, when a woman marries her
husband, for example, a junior olmurrani, she may move into her new mother-in-law’s enkaji until such
time as she has built her own enkaji with the help of the women and, sometimes, men. As the junior
olmurrani passes into senior murranhood through eunoto219, he will open his own enkishomi (gate) into
his father’s enkang’220 and continue to marry further wives. Thus by the time a lineage head has
become a senior elder (Maa: olpayiani sing’; ilpayian pl.), it may often be the case that several
generations may be living together in the enkang’ in one or more of the following combinations: as an
enkang’ in which the lineage head lives together with his wives, sons and their wives and children in
several separate ilmarei - each olmarei comprising one or more households (inkajijik); as a enkang’ of
brothers or sometimes an agnate or age mate living together with their wives and children as well as
perhaps a sister and/or a widowed mother; as a single olmarei with a husband living together with
his wives and children.
Generally, although not always depending on the success of each olmarei, the longer an enkang’ has
been established, the larger it will become. A lineage head may aspire to become a respected elder
or, ‘notable’ (Maa: olkarsis sing.; lkarsisi pl.) in his community221, presiding over large herds (still
possible to an extent in Idodi) and a large kinship group (Rigby, 1983, 145). Yet, as the enkang’
grows, the lineal head’s control over the primary means of production (livestock222) is progressively
weakened and diluted, since wives, sons, affines and matrilateral kinsmen increasingly claim their
rights to, and control of, certain portions of the olmarei herd (Rigby, 1983, 145). When the lineage
head dies, and usually after the inheritance ceremony, the enkang’ will, sooner or later, split up with
the next generation dispersing to found their ilmarei. Women customarily may play a pivotal role in
the inheritance process, as it is they who have the right to allocate livestock in their enkaji to their sons
as they see fit through matrifiliation. Married sons may also leave the enkang’ of their father while he
is still alive to establish their own olmarei223. Thus over the developmental cycle, a domestic unit may
grow from a single enkaji (house) to become an olmarei (multi-hearth household) and eventually a
large enkang’ (a group of multi-hearth households). A varying range of different relations or age
219 A ceremony in which the previous senior ilmurran relinquish control to the junior ilmurran below, prior to the
opening of a new age set some years later.
220 During fieldwork in Idodi, the largest number of gates in any enkang’ was four (in one enkang’ only), with
one third of inkang’itie having two or more gates220 and the rest (two thirds) having a single gate.
221 Women also pass into respected senior eldership (Maa: entasat sing.; intasati pl.). Men and women senior
elders are traditionally keepers and teachers of history , law and cultural unity (Rigby 1992, 69)
222 Although, as Ilparakuyo are now increasingly agro-pastoralist and own farm fields, inheritance patterns and
dynamics are likely to change, with fewer kin benefiting from the inheritance of fewer stock and a small amount
of farmland.
223 For a more detailed description of the physical and symbolic layout of an olmarei or enkang’, in relation to
the developmental cycle and matrifilial patterns of livestock inheritance, refer to Rigby (1983, 140-161) .
164
mates and their wives and children may be brought to live together (as discussed above), before
eventually fission occurs224 after the death of the lineal head of the enkang’.
The compositional variety of inkang’itie in Idodi is shown in Table 6.1. Some Ilparakuyo ilmarei may at
one time or another have one or, more rarely, two unrelated labourers225 living in the enkang’. In
addition to guests, ilaiyok (Maa: uncircumcised herd boys; sing olayioni) or murran (both sometimes but
not necessarily agnatically related to the homestead family) from another homestead may sometimes
also live for temporary or longer periods in the homestead principally helping with herding.
Table 6.1: The varying compositional number of households (inkajijik) in Ilparakuyo

super-households (ilmarei)
Number of households (inkajijik) in each
1 2 3 4 5+
Super-household (olmarei)
Number of super-households (n=52) 8 17 11 10 6
Percentage proportion 15% 33% 21% 19% 12%
The data presented in Tables 5.1, 5.2 and 5.3 are derived from a census survey of all Ilparakuyo
ilmarei and inkang’itie carried out during fieldwork. Average household sizes in all ilmarei are very
similar irrespective of overall olmarei size as shown in Table 6.2, apart from single household ilmarei
which tend to be larger, the difference however not being statistically significant.
Table 6.2: Super-household (olmarei) and household (enkaji) sizes measured in Adult
Equivalents
Number of households in each Super-
1 2 3 4 5+
household (olmarei)
Super-households: average size (RAsa)b 5.6 ±0.70 6.4 ±0.42 9.8 ±0.72 13.5 ±0.74 17.6 ±1.53
a c
Households: average size (RAs ) 5.6 ±0.77 3.2 ±0.21 3.4 ±0.23 3.7 ±0.19 3.0 ±0.29
a
RAs -Adult Equivalents (after ILCA 1981) - calculated for each household from household constitution data (adult
male = 1, adult female = 0.86, children 0-5 = 0.52, children 6-10 = 0.85, male child 11-15 = =0.96, female child 11-15
= 0.86) - total RAs in 100 per cent sample = 501.6; b n=52; c n=150; Total number of people= 660.
However, there are notable differences in the make-up of olmarei in different developmental cycle
stages. Olmarei and olmarei composition in terms of developmental cycle trends, given the fact that a
substantial proportion of ilmarei and their constituent households may be at different development-
cycle stages, can be most easily measured by using the age-set status of the male head of the
olmarei. While not definitive, age set serves to act as a rudimentary indicator of basic trends for each
developmental cycle stage, trends which are likely to be related to variations in livelihood strategies.
This approach is heavily androcentric but necessarily so given the context in which the data were
224 Refer to Rigby (1992, 139-142) for further detail about the fission of domestic units.
225 These labourers are often itinerant moving between homesteads following work, and are frequently Hehe or
Gogo from Pawaga to the northeast and beyond.
165
gathered226. Table 6.3 provides a summary of olmarei characteristics analysed by the age set of the
male head of each olmarei. Although five age sets are currently extant, only three feature since the
oldest age set has nearly completely died out in the area and the youngest has just only been locally
opened. The data show that the ilmarei of the most senior age class of senior elders (currently the
Ilmedoti age set) are the largest. This would be expected given that they have had the longest period
in which to expand as the patriarch continues to marry and take in relatives from discontinued
households out with the homestead.
Table 6.3: Super-household and household characteristics presented by age-set as an

indicator of life-cycle stage
Senior elders Junior elders Warriors
Super-households headed by males of
(Ilmedoti) (Ildareto) (Ilkimunya)
the age seta:
(1956-1970)b (1972-1987)b (1987-2000)b
Number of super-households (n=52) 10 25 17
Percentage proportion 19% 48% 33%
Average number of households per super-

3.7 ±0.54 2.4 ±0.25 3.1 ±0.34227
household
13.4 ±1.78†
Super-households: Average size (RAs) 8.8 ±0.81 8.6 ±0.93
(** df=2)
2.9 ±0.22†
Households: Average size (RAs) 3.7 ±0.25 4.0 ±0.32
(*** df=2)
a
In inkang’itie where different ilmarei are headed by males of different age-sets, the most senior age-set male is
used to classify the enkang;
b
The period in which the age set was ‘open’ i.e. all initiates during this period became members of the open age set -
the dates are specific to Idodi and Pawaga; †Statistically significant difference - Friedman Test & Kendall’s W Test: *
= p<0.05; ** = p<0.01; *** = p<0.001; df = degrees of freedom.
The number of people living in each olmarei and household by age class as measured by Adult
Equivalents (ILCA 1981) is shown in Table 6.3. As would be expected, the ilmarei of the most senior
226 While previous studies on Maasai groups have been justifiably criticised by Hodgson (2000) as being overly
androcentric in their approach, and despite the central but changing roles that women play in Maasai society
(Hodgson, 2001), this analysis is centred around male-informed interpretations of Ilparakuyo society, as there
was unfortunately insufficient appropriate context and opportunity to adequately engage with Ilparakuyo
women in this regard. It is also further recognised that not all Ilparakuyo ilmarei may necessarily be ‘headed’ by
a man and that women may effectively play such a role. However, for the purposes of this study, given that no
cases of women-headed ilmarei were encountered during survey work and bearing in mind the relatively small
sample size, an androcentric household classification system is adopted.
227 A slight ‘anomaly’ is apparent as denoted in Table 5.3 in that the youngest age class (currently the Ilkimunya
age set) appear to have a larger number of households in their ilmarei than the age class immediately above
them. This can be explained, and is the case, for a number of olmarei that, in the last 5 years or sometimes more,
households have been newly formed from recently fissioned ilmarei. The larger numbers of households in these
ilmarei tend to be constituted mainly of widowed mothers and wives (sisters of the male head of the olmarei)
who have moved to live with their sons and brothers respectively.
166
extant age class (currently the Ilmedoti age set) are statistically significantly the largest on average
(χ2=4.90; Kendall’s W coefficient = 0.490) as compared to the younger age classes. Individual
household sizes also show the same trend across age classes with the youngest age class of the
ilmurran (currently the Il kimunya age set) having statistically significantly the smallest households on
average (χ2=14.60; Kendall’s W coefficient = 0.730). Overall the data therefore support what
would be theoretically expected to be the case and which has been described previously for the
Ilparakuyo (Rigby 1985; Beidelman 1968; von Mitzlaff 1988).
6.4.2 Herd structure and management

Describing Ilparakuyo herd structure is a complicated undertaking since herds are nearly always in a
state of flux. In addition to normal herd population dynamics of birth, ageing and death, the herd
may at any time be added to, subtracted from and split for a wide variety of reasons. The livestock
present in any olmarei are likely to be made up of those animals belonging to the olmarei as well as
frequently animals loaned in from affines - usually either for grazing and/or for milking. Livestock
within the emboo (Maa: cattle byre; imbootie pl.) of an olmarei may be allocated by the lineage
head of the olmarei to one of his wives (as part of her matricentral enkaji), or if he is wealthy enough,
kept as residual stock. Although a wife has certain, usually substantive, usufruct rights over each
animal she is allocated for her enkaji (for example, rights to milk, hide and fat), these rights are not
exclusive, as the cow’s milk, blood and perhaps even meat may be used by others from time to time.
This is reflected in an Ilparakuyo (and wider Maasai) adage: ‘Meeta enkiteng’ olopeny’- a cow has no
single owner228 (Rigby, 1983, 142). Yet, despite this adage, women may often (although not always)
retain substantial control over herd products and mediate their consumption. In addition to
matricentral, residual and affines’ livestock, other stock may be present in the emboo (byre) of some
ilmarei which is the property of the limited number of Hehe/Bena farmers in the Idodi villages who
own cattle. These livestock are looked after by the olmarei as part of a reciprocal agreement, which
may include male members of the olmarei training the loaned-in stock to plough, or just looking after
the animals in return for, for example, a share of their milk production. Conversely, the olmarei may
have lent stock out to other olmarei for grazing and milking and sometimes to avoid disease risks.
Some of these arrangements are temporary, lasting only a period of a few months, while other loans
are much longer term and last for years.
Herding, while often seen as primarily falling to the olayioni (herd boy) age grade and also, closer to
the olmarei, not infrequently carried out by the ‘natoyie’ (Maa: young girls; sing intoyie), is closely
supervised by head of the olmarei. It is largely they who ultimately decide, often in consultation with
the women, where the herds are to be grazed, when they are to be moved and where to. Women
tend to more closely oversee the herding arrangements of small stock and calves, and men the large
stock, although such divisions of work are far from exclusive. Calves and young small stock are left
228 As Rigby (1983, 142) points out, this saying is synonymous with the fact that ownership of things in
Ilparakuyo society tends towards being inclusive rather than exclusive, although certain things, such as a woman’s
house, are recognised as being her private property - access to which is governed by strict convention.
167
Figure 6.3: Tending small stock in the enkang’ in the early morning
Figure 6.4: Ilaiyok with their favourite oxen
168
behind in or close to the olmarei when the main herd is out grazing and only join them once they are
old enough to be able to move with the herd without trouble. In a typical day, the herd, once it has
been milked, will be taken out to graze one to two hours after dawn. Returning by late morning, they
will rest up in the heat of the day before going out again in the early afternoon, returning again
towards dusk. In contrast to Rigby’s (1983, 137) description of inkang’itie suffering from a chronic
shortage of labour due to their relatively small size - as compared with inkang’itie in the Kisongo
Maasailand, it is more likely the case that herding labour shortages are increasingly seasonal for
Ilparakuyo ilmarei in Idodi. This is because there are now nearly two and a half times fewer livestock
per person for the Ilparakuyo in Idodi than there were in Bagamoyo for the Ilparakuyo Rigby
described during the 1970s. Thus Ilparakuyo labour shortages are worst during the farming season,
as all Ilparakuyo ilmarei now farm, and now also pertain to both herding and farming.
At the onset of the dry season, male heads of ilmarei may especially and increasingly spend
considerable amounts of time in the ‘kilabu’ (local beer club) drinking with Hehe/Bena farmers.
Drinking relationships are cultivated by the Il Parakuyo which often enable them to negotiate
(preferential) access to fields of crop residue in the dry season which is critical to the nutritional
security of the herd. Access to the residue in the fields may frequently be agreed several weeks to
sometimes months in advance, to ensure a greater chance of cattle being able to graze the fields.
Analytical considerations
There are two ways of examining stock holding at the olmarei level: either by including all animals
loaned out and not animals loaned in for each analysis or, by ignoring all animals loaned out and
including stock loaned in. Whereas the former approach provides an absolute count of the overall
livestock wealth of each olmarei, the latter provides an over-view of the availability of livestock in
each olmarei in contributing to daily livelihoods. Further, for many of the poorer ilmarei, the stock
present in their imbootie (cattle byres) may consist largely or entirely of animals loaned in from
relatives. Thus absolute counts of stock holdings owned would provide a misleading picture of the
availability of cattle in an olmarei, especially for the poorest ilmarei.
AS discussed in Chapter Two, data on livestock holdings were collected in the form of an annual
census annual over two consecutive years (2000 & 2001) during the months of September-December
of each year after the majority of the year’s calving had occurred229 and when livestock tended to
have been returned from grazing seasonal pastures and were then grazing local fields of mabua
(Swahili: crop residue). It should be noted here that data collection and therefore this analysis was
carried out at the olmarei level since acquiring accurate stock census data at the household (enkaji)
level would have been difficult and compromised by inaccuracy and ambiguity230. It is therefore
229 For details about how the data was collected, refer to Chapter Two.
230 It is certainly the case that some or all female cattle may be effectively allocated by the lineal head of the
olmarei (super-household) to each enkaji (household). However, it was concluded during fieldwork that analysis
at the super-household level, particularly given the cross-comparative nature and focus of the study between
farming and herding groups, would avoid the following potential ambiguities: (i) The fact that frequently not all
169
necessarily assumed that stock were generally allocated evenly amongst all the households of each
olmarei and that the overall livestock status of an olmarei was representative of individual household
(enkaji) statuses within each olmarei. Studies (citations) have shown that this assumption is often not the
case, and that the livelihood status of enkaji may vary considerably within ilmarei. Unfortunately, the
dataset as collected is insufficiently disaggregated to enable analysis of livelihood status variation
within ilmarei.
An overview and explanation of herd structure is provided before the nature of major herd
transactions is discussed.
stock present in each emboo is the property of the olmarei and that these stock may be short or long term loans;
(ii) All such stock in each olmarei is not necessarily always allocated to a specific household, and; (iii) Households
may often share or borrow milk and other stock-derived products from each other.
170
Table 6.4: Ilparakuyo large stock (cattle - Bos indicus) herd composition and structure for 2000 and 2001
Female Male
Engahe kikainoti
Engiteng lepong
Engehe olarami
Ol bu'ngai kinyi
Engiteng olupii
Emerimeruwa
Engahe kinyie
Olaheoeleki
Ol ai'ngoyni
Ol saigorr
Orikiteng
Endapesi
engiteng
Osonash
Oswate
Household details
Female - natural
Total number of
TOTAL livestock
Young male calf
Weaned female
Asexual animal
Female - post-
calving sterile
Male - natural
Young female
Weaned male
holding (incl.
Early mature
Fully mature
Cow - many
households
Number of
Loans out)
household
Cow - few
people in
offspring
offspring
Bullock
sterile
sterile
Steer
Year
bull
bull
calf
calf
calf
379c 300 111 834 0 0 258 172 65 31 53 32 1 0 2236
2000
d
121.6 207.1 113.1 850 0 0 82.8 118.7 66.0 40.9 59.2 48.7 1.4 0 1709.5
621a 366c 304 92 829 0 0 283 184 59 54 66 31 2 0 2270

2001 150
502.0b 105.9 d
176 83.6 768.5 0 0 82.4 113.2 55.8 68.6 62.5 39.6 2.8 0 1558.9
373c 302 102 832 0 0 271 178 62 43 60 32 2 0 2253

Average
d
113.8 191.6 98.4 809.3 0 0 82.6 116.0 60.9 54.8 60.9 44.2 2.1 0 1634.2
16.6%c 13.3% 4.6% 36.9% 0.0% 0.0% 12.0% 7.9% 2.8% 1.9% 2.7% 1.4% 0.1% 0.0%
d
7.0% 11.7% 6.0% 49.5% 0.0% 0.0% 5.1% 7.1% 3.7% 3.4% 3.7% 2.7% 0.1% 0.0%
Totals
72.6%c 27.4%c
74.2%d 25.8%d
a
Number of people - adults and children combined; b Reference Adults (RAs)(ILCA 1981); c Number of individual animals in each category; d Number of Herd Livestock Equivalents (LEs - inter-annual herd
average =184.7kg: young ♀calf 79kg; weaned ♀ calf 170kg; cow [♀] 251kg; young calf 79kg; weaned ♂ calf 170kg; early mature [♂] bull 250kg; fully mature [♂] bull 325kg; young steer [♂] 275kg;
mature steer [♂] 375kg) in each category (adapted from King et al., 1984); e Ilparakuyo categories of cattle status and age. NB: Average LE for large stock herd over 2 years of data =0.74 cf 0.72 (King
et al., 1984) and 0.71 (ILCA, 1981).
171
Table 6.5: Ilparakuyo small stock (goats and sheep) herd composition and structure for 2000 and 2001
Female Male
Emkine omorumo
Orpejaroti oroti
Engerr morumo
Orrkerr lplaihe
Embale engerr
Embale engerr
Embale engerr
Orrbalo lekine
Orrbalo lekine
Okerr lembalo
Ositima lekerr
Ositima lekerr
Enkine lepong
Orrkine sapuk
Engerr lepong
Emablo ekine
Embalo ekine
Embalo kinyi
Ol meregesh
Ol meregesh
sapuk/kotok
Oloro sapuk
Oloro kinye
ekika onoti
ekika onoti
enkinee
botorr
botorr
botorr
botorr
Household details
lekine
kinyif
kitee
kinyi
kinyi
4th plus offspring
Total number of
TOTAL livestock
Mature castrate
Weaned female
Young castrate
Mature female
Mature female
Mature female
Young female
Weaned male
holding (incl.
Early mature
Fully mature
1st offspring
Young male
households
Number of
Loans out)
household
buck/ram
buck/ram
people in
offspring
kid/lamb
kid/lamb
kid/lamb
kid/lamb
2nd & 3rd
Year
502c 393 156 892 3 357 219 119 76 72 59 2851

2000
d
331.3 153.3 484.7
621a 374c 354 134 681 0 266 195 131 79 63 55 2332

2001 150
502.0b 262.1 d
134.1 396.4
438c 374 145 787 0 312 207 125 78 68 57 2592

Average
d
296.7 143.7 440.5
16.9%c 14.4% 5.6% 30.4% 0.0% 12.0% 8.0% 4.8% 3.0% 2.6% 2.2%
Totals 67.4d 32.6
67.4%c&d 32.6%
a
Number of people - adults and children combined; b Adult Equivalents (RAs)(ILCA 1981); c Number of individual animals in each category;
d
Number of Herd Livestock Equivalents (LEs ≈ 0.17 average conversion ratio for small stock (ILCA 1981)); e Ilparakuyo categories of goat status and age; Ilparakuyo categories of sheep status and age.
172
Herd structure and stock distribution
The Ilparakuyo herd structure is characteristic of many east African pastoralist herds (for example,
King et al. 1984; Rigby1983; Jacobs 1965; Swift 1981 cited in Lane 1996, 11) and is managed to
maximise (as far as possible) milk production (Bekure et al, 1991) and its reproductive potential with
a heavy bias towards female stock throughout the herd structure - see Table 6.4 and Table 6.5. The
overall ratio of female to male stock is in fact somewhat higher than that reported in other Maasai
herds with female constituting an average of 72.6 per cent of the herd (e.g. 66.4% in Kajiado Kenya
- King et al. 1984; see also Table 6.20). The overall Il Parakuyo cattle herd showed a slight increase
between 2000 and 2001, although the number of female livestock declined slightly, the number of
male cattle in the herd increased to the extent that the overall herd grew marginally - see Table 6.4.
The small stock herd declined markedly between 2000 and 2001 (refer to Table 6.5) for reasons
that are discussed further on in this section.
Male and female cattle are managed for very different purposes in the herd. Over 4.5 times the
number of female stock to male are retained in the herd for at least 6-8 years while they continue to
calve successfully and produce milk before they are sold (usually) for slaughter. In contrast to the
sustained productivity of the female portion of the herd, the male portion is managed for a wider
range of purposes - young male stock are far more predominantly used for cash and ritual exchange
requirements231 and only the strongest and most favoured animals are selected to remain in the herd
on a long term basis either as bullocks for ploughing or as bulls for siring the herd. In addition steers
may be kept for a limited period of time for later slaughter - especially for ritual ceremonial
purposes. A conscientious herd manager will be very much aware of the calving and mating record of
his cattle and, in consultation with the women (as the milkers of the herd) of the general milking status
of particular animals (especially those not lactating as expected). While the greater number of cows
is critical to the reproductive potential of the herd and central to household dietary requirements
(milk, and ghee especially), fully mature bulls with a good mating record are regarded with much
pride especially by the herd boys of the olmarei. Increasingly these days, bullocks have begun to
play a seasonally important economic role with the introduction of cattle ploughs, not least reflected
by the increasing dependence of the Ilparakuyo on agriculture as a mainstay to their livelihoods see
Boxes 5.3 and 5.4). While not only much reducing the highly labour intensive land-tillage
requirements faced by Ilparakuyo households (all of whom cultivate), bullock plough teams are
seasonally rented out to Hehe/Bena farmers for considerable profit on an acreage basis.
Small stock (sheep and goats) are managed on a similar basis to that of the large livestock but
largely as a source of meat - mostly, although not exclusively, for household consumption, and for
231 This is not to deny that a parallel role is played by female stock. Milk, which in itself plays an important
symbolic role in many rituals, aside from constituting a major dietary component to the household, is sold by the
women to neighbouring farming communities. However, in Idodi, the relative cash gains are marginal from this
activity (compared to those derived from stock sales) but nevertheless represent a small but significantly
important independent income for women.
173
sale. The small stock herd exhibits a similar (although not quite as great) bias towards females, which
ensures the herd’s ability to recover from bad years of drought and/or disease. While not culturally
as important as cattle, nevertheless small stock play an important part in the economy of the olmarei
as they reproduce faster than cattle and are easier to sell locally due to their considerably lower
value. As has been previously asserted (e.g. de Leeuw et al. 1991), small stock play a critical role in
post-drought herd recovery as they tend to better survive and recover faster from drought events. A
characteristic of small stock is that they appear to be locally far more susceptible to disease and are
generally prone to far greater inter-annual variations in morbidity and morality - but equally
conversely, productivity. For example, whereas the overall number of head in the large stock herd in
Idodi increased by 1.5 per cent between 2000 and 2001, the small stock herd declined by 22.3 per
cent over the same period. The small stock herd suffered extensively from tick borne diseases and a
widespread and crippling hoof-rot condition associated with a better than average year of rain in
2001.
The distribution of livestock holdings between ilmarei is depicted in Figures 5.5 and 5.6 for the years
2000 and 2001 respectively. Total livestock holdings are measured in Livestock Equivalents (LEs) per
Reference Adult (RAs)232. The graphs also show the relative proportion of large stock (cattle) to small
stock (goats and sheep) in each olmarei. Ilmarei have been classified into four wealth classes as
modified from King et al. (1984) - refer to Box 6.2.
232 See the explanatory notes below Tables 5.4 & 5.5 as to how the herd LEs and RAs were calculated. It should
be noted that LEs are not the same as Tropical Livestock Units (TLUs - set at 250Kg). LEs instead are a derivative
of the local herd’s overall average adult animal metabolic mass. Average herd LE for 2000 & 2001 was
estimated at 184.7kg – 0.74TLUs – which compares favourably to that arrived at by King et al (1984:26) of
0.72 TLUs for Kajiado herds.
174
Box 6.2: A description of the wealth categories adopted in relation to Ilparakuyo
livelihood strategies
Although King et al, (1984) do not provide an accompanying explanation, they adopt 3 wealth
categories – ‘rich’, ‘medium’ and ‘poor’ in relation to the number of livestock equivalents per
reference adult needed to support differing degrees of pastoralism. In this analysis, ilmarei which
fall into the ‘wealthier’ category are defined as having sufficient access to livestock to be able to
rely totally on their herd for their livelihood. While these households may practice farming, it is
complementary and not necessarily critical to household livelihood status. Households falling into
the ‘middle’ category may not have quite enough livestock to solely rely on their herd for their
livelihoods. Thus for households in this ‘middle’ category, farming is particularly important as a
complementary livelihood strategy, which also lessens the household’s reliance on its herd. In this
analysis the ‘poor’ category has been further subdivided into ‘poor’ and ‘very poor’ to simply
demonstrate the marginality of livestock holding amongst a large proportion of the Ilparakuyo in
Idodi. The two latter categories may be characterised by households with increasingly insufficient
access to livestock to be able to rely on the little stock they have for their livelihoods. Farming is
the main livelihood strategy with livestock providing a nutritional supplement as available.
These wealth designations are somewhat arbitrary and not definitive, but nevertheless help to
provide an indication of the relative importance and significance of livestock in individual olmarei
livelihood strategies. Although etic in their derivation, these wealth categories may be seen to be
reasonably representative of emic perceptions of enkarsisisho (Maa: wealth) which, for men at least,
centre on presiding over a large number of stock as well as a large kinship group (see also Rigby
1983, 145)233. All wealth categories are relative and not absolute. Some of the ‘very poor’
Ilparakuyo may still have livelihood resource endowments considerably greater than those of the
poorest Hehe/Bena farmers.
It is evident, that as mentioned above, small stock are much more variable in their productivity as
compared to large stock, and this can be clearly seen from the Figures 5.5 and 5.6. Thus while large
stock may be seen as being more productively reliable, small stock, while less reliable, have greater
productive potential, such that in good years small stock surpluses are generated which may be
converted into other investments - for example, large stock and agricultural holdings, or invested in
socio-cultural relations. However, despite the variability in the inter-annual small stock herd, small
stock holdings are reflected in large stock holdings such that those ilmarei with the largest cattle herds
tend generally to have the largest small-stock herds as well - as depicted in Table 6.6.
233 Despite other objects of perceived wealth, which may include radios, bicycles and other modern
conveniences, ultimately enkarsisisho still lies in cattle and kin, and perhaps more recently, farmland.
175
Figure 6.5: Livestock holdings for each olmarei during the year 2000
14.00
Small Stock
'Rich'
Large Stock
12.00
Increasing
Livestock Equivalents per reference adult
10.00 Wealth
'Middle'
8.00
6.00
4.00
'Poor'
2.00 'Very poor'
0.00
31
26
34
51
35
13
25
22
12
33
14
49
46
45
18
37
15
20
23
36
19
38
41
52
16
50
48
28
44
17
43
32
29
11
42
21
10
40
30
39
27
24
9
6
1
5
Super-household identifier
Figure 6.6: Livestock holdings for each olmarei during the year 2001
14.00
Small Stock
Large Stock 'Rich'
‘Wealthier’
12.00
Livestock Equivalents per reference adult
Increasing
10.00 Wealth
'M iddle'
8.00
6.00
4.00
'Poor'
2.00 'Very poor'
0.00
31
26
34
51
35
13
25
22
12
33
14
49
46
45
18
37
15
20
23
36
19
38
41
52
16
50
48
28
44
17
43
32
29
11
42
21
10
40
30
39
27
24
9
6
1
Super-household identifier
Allowing for some margin of error234, those ilmarei which fall into the ‘wealthier’ and ‘middle’ livestock
categories occur above the threshold of about 5 LEs per reference adult recognised as denoting the
level at which a (super) household can heavily depend on its stock for its livelihood. Thus only 27.5
per cent and 27.0 percent of ilmarei in 2000 and 2001 respectively had access to
234 While the data on livestock holdings would appear to be remarkably robust (see previous footnote), given
that livestock weights are estimated, and that the livestock categories employed may not entirely correspond to
those utilised and followed by King et al (1984) (from which the calculations are drawn), in calculating Livestock
Equivalents, the resulting analysis and data points can not be claimed as being totally exact but are instead
accurate relative to each other.
176
Box 6.3: A growing ilmarei exerting greater demands on its livestock herd
Ismaeli is one of the better-off Ilparakuyo herders in the Ikwavila valley. He lives together with his four
wives and six children (two girls and four boys) in Isanga, Makifu village. Ismaeli was born in Ilusi in the
Lunda section of the LMGCA in about 1958. His family moved from Ilusi to Malinzanga village while he
was still a herd boy. He was initiated into the Ildareto age set in 1973. In 1975 his family moved to
Tungamalenga village as they thought that the grazing for their livestock would be better there. The
family moved again in 1976 and 1977 to different sites in Tungamalenga and Mapogoro villages.
Unfortunately his father had an alcohol problem and the family lost most of their stock in one way or
another due to his drinking. Ismaeli and his brothers thus went to work as stockmen for another herder,
Mzee James. In 1978 the family moved to Isanga, where they have since lived. In 1982 his father died
and was taken to be buried in Ismani about 80 km to the east.
During the mid-1980s Ismaeli began to travel around Pawaga and Idodi and engaged in some stock theft
in order to start his own herd (since few cattle had been left by his father). He was caught at least once
stealing cattle but managed to escape from the village lock-up where he had been caught. After a
couple of years, he managed to build his herd up from three cows to twenty seven from stolen cattle
and livestock borrowed from relatives. In 1988 he was able to make a bride wealth down-payment for
his first wife of fifteen cows. At this point, he and his first wife Nailole moved to their own enkang’ and
from his brother’s place where he had been living. In 1990 he married a second time, and then again in
1996 and for a fourth time in 2001.
Today Ismaeli and his family depend on livestock keeping and farming for their livelihood. The family
keeps about 200 cattle, and until recently, a similar number of goats and sheep (unfortunately perhaps
up to one hundred were lost from disease). The family farms about 6 acres of dryland fields for maize
and has also recently bought about 2 acres of irrigated maize and banana farm plots in the bonde. They
had hoped to farm rice, but had decided not to try yet, since farming rice would be quite expensive,
and they were apprehensive that this would put too many demands on the family herd. This is because
the family herd supplies most of their cash income requirements, and there are many out-goings. For
example, two of Ismaeli’s brothers are often drunk and do not ensure that their families have access to
enough food. So Ismaeli’s brothers’ wives may quite often come asking to borrow maize from Ismaeli’s
wives, and the family may often end up feeding an extended family. In return, the family can
sometimes count on help from Ismaeli’s brothers’ families for herding, and in years when they take the
herds up into the higher ground in search of early rain and pasture, they will often share the herding
involved by combining their herds. Ismaeli has employed a general labourer (a Gogo) who helps herd and
farm, and another local part-time labourer from the village to help with the farming. But labour is often
in short supply in the family, and this is another reason why they decided not to farm rice. (In fact, the
following year, the family did farm rice.) Ismaeli has also trained two of his oxen to plough and he is
able to earn up to TShs 10,000 per acre hiring out his oxen team to other farmers.
Ismaeli sees his family as generally doing quite well. Two of his oldest children (both boys) are going to
school, and their mothers say that the girls will follow when they are old enough. Members of the family
are often ill though - malaria is a constant problem. Nailole nearly died from malaria in 2001. Ismaeli
took her to the Lutheran clinic (the best health facility in the valley) and made sure that the doctor
attended her until she had pulled through. The herd has also suffered in recent years from higher levels
of morbidity - for example, the family lost over 100 goats and sheep in one year due to disease - a
respiratory infection that lingered and hoof rot.
177
Box 6.4 An Ilmarei investing in agricultural production as a central livelihood strategy
Salum was born in Mahuninga village and spent much of his childhood in the Mahove bushlands as a herd
boy. Although his family moved away to Makifu for many years, Salum returned to live in the same area
that he had grown up in. Today he lives with his mother and two wives and their five children (four boys
and a girl). The family say they aren’t that well off compared to other Ilparakuyo - they have thirty
cattle and about sixty goats. The family had recently lost six goats to a hyena and Salum felt that his
livestock were not doing very well. He had gradually come to the realisation that he and his family had
to diversify into farming in order to secure their food needs and to generate more cash.
So Salum bought a shamba (Swahili: field plot) from a Hehe farmer for what he said was a ridiculously
small amount of money. Many people laughed at him - they thought he was a fool to have bought such a
shamba as the crops of the previous owner had done very poorly. But Salum had other plans. The field
lay next to a stream which flowed for most of the year. About 200 meters upstream, he built a weir out
of stones and logs. Together with the help of the family’s farm hand, John, he dug an irrigation furrow
up to one and a half metres deep in places to his field below. He said that he had never worked so hard
in his life as digging that trench. But after 3 months of work, the furrow was ready.
The weir and irrigation furrow worked very well. Salum managed to obtain a plate of thick iron sheet
from the village settlement with which he made a sluice gate so that he could control the amount of
water running into the furrow from the stream. The shamba was thriving. He had planted several
different bean crops together with tomato and onion which he planned to sell. He was still learning
about farming - his tomato seedlings had been transplanted too early and were in danger of withering in
the open sun. But together with John, they had created what they hoped would keep providing the
family with a sufficient income and a more varied diet. Salum had already been thinking about the other
side of the stream. There was an area that could be cleared for another field. Building another
irrigation furrow from the wear was not really possible, but he instead planned to use a hollowed-out
log to make a branch of his current irrigation furrow cross the stream.
H i l l s i d e
Trees
Planned irrigated shamba

Weir
Hut
200m
Irrigated shamba
H i l l s i d e
178
Figure 6.7: A dam built by an Ilparakuyo
pastoralist to irrigate his crops
Figure 6.8: The irrigated field with a crop of onions nearly

ready for market
179
livestock holdings above this threshold - as depicted in Table 6.6. The contribution of small stock in LEs
to attaining this threshold was substantial with only 15.7 per cent and 13.7 per cent of households in
2000 and 2001 still remaining above the threshold when the small stock herd was discounted.
However, the most important insight arising from the data is that nearly three quarters of ilmarei do
not have access to sufficient livestock to be purely pastoralists and therefore they have no immediate
choice but to be reliant in greater measure on agriculture for their livelihoods. However, this does not
necessarily mean that livestock poor ilmarei are absolutely poor, as if farm holdings are taken into
account, some of these households may still be relatively livelihood secure, particularly in comparison
to the poorer members of the Hehe/Bena farming community.
Table 6.6: The overall proportion of ilmarei in different livestock wealth categories
(N=52;
n=50235) All stock Large stock only
2000 2001 2000 2001
Wealthiera 5.9% 2.0% 0% 0%
Mediumb 21.6% 25.5% 15.7% 13.7%
Poorc 33.3% 35.3% 27.5% 35.3%
Very poord 39.2% 37.2% 56.8% 51.0%
a b
Livestock equivalents per reference adult (LE:RA) ratio >12.00; LE:RA= 5.00-11.99;
c d
LE:RA= 2.50-4.99; LE:RA= 0-2.49 (King et al., 1984; Bekure et al. 1991 - modified).
Table 6.7: Average small stock holding in ilmarei of different large stock wealth
categoriesa
Mediumb Poorc Very Poord
†
2000 1.9 ±0.45 1.3 ±0.29 0.6 ±0.12
1.4 ±0.27***
2001 1.0 ±0.15 0.5 ±0.11
(df=2)
†
A noticeable trend which approaching partial significance (p=0.131; df=2).
*Statistically significant difference - Friedman Test & Kendall’s W Test.
a
Data independent - wealth categories based on large stock holdings independent of small stock holdings in each
olmarei. All inclusive (large and small stock combined) wealth categorisations yield greater significance levels
(p<0.05 for all years) for differences in small stock holdings between wealth categories.
b
Ilmarei with LE:RA large stock ratio of 5.00-11.99; c 2.50-4.99; d0-2.49. No ilmarei fell in to the ‘wealthier’ category
of LE:RA>12.00 for large stock holdings alone - see Table 6.6.
A further aspect to the distribution of livestock holdings across ilmarei is the level of access to livestock
according to developmental cycle stage. As previously discussed, many Ilparakuyo ilmarei are a
235 All data analysis on livestock and farming are derived from census of 52 ilmarei of which data from 50
households have been used (one olmarei left the area and for another there is an incomplete dataset).
180
complex construct in terms of kinship, gender and age. Bearing this in mind, as an indicator of the
variability in access to livestock with developmental cycle stage, average livestock holdings per
reference adult for ilmarei at different developmental cycle stages are presented in Table 6.8.
Table 6.8: The livestock holdings of super-households at different stages of

developmental cycle
Senior elders1 Junior elders1 Warriors1

Year
(Il medoti2) (Ildareto2) (Il kimunya2)
4.5 ±0.70***
2000 (LE:RA3) 3.8 ±1.09 3.7 ±0.72
(df=2)
4.6 ±0.64***
2001 (LE:RA) 3.4 ±0.93 3.5 ±0.56
(df=2)
1
Age class of the male head of the olmarei (Senior elders the oldest; Warriors the youngest); 2Age set - see previous
3
box; LE:RA - Livestock Equivalents per Reference Adult; *Statistically significant difference - Friedman Test &
Kendall’s W Test.
Perhaps somewhat surprisingly, it is not the oldest ilmarei that have the largest number of stock per
person, as might be expected (but see Stenning 1971, 98) through a lifetime of accumulation. Rather
it is those ilmarei headed by junior elders which generally have the highest number of stock per
person - representing the stage in the development cycle at which ilmarei are at their greatest
productive potential at a point where they have become well established, a herd has been built up
and the ilmarei has several households (enkaji). Generally, demands on the herd in these ilmarei have
not grown as great as those of the oldest ilmarei which can consist of as many as eight to ten
households drawing heavily on the productivity of the super-household’s herd. Stock holdings of the
youngest ilmarei may be generally dependent on inheritance and the degree to which a murrani
(usually) has been able to build his stock up in the (relatively) short time since the inheritance
ceremony or since he moved away from his father’s olmarei.
Herd transactions
Herd transactions236 are a central component to herd management and livelihood strategy. Not only
are large stock sold for cash, exchanged for other animals, loaned out for either milking and or
grazing to kin and associates They fulfil important symbolic roles in cultural rituals- for example at
meat feasts (Maa: ilpuli), circumcision ceremonies (Maa: emurata), age set ceremonies (for example,
eunoto237, olng’esher238) and births. The relative incidence of different herd transactions that were
236 Transaction is used here in the sense of affecting the removal or addition of one or more animals out of or
into a livestock herd.
237 This is passage of junior ilmurran to senior murranhood.
238 The passage of senior ilmurran to junior eldership (Maa: olpayiani sing.; ilpayian pl.) and, of junior elders to
senior eldership and, senior elders to ‘retired elders’ (Maa oltasaati sing; iltasaat pl).
181
recorded in 2000 for the Ilparakuyo in Idodi is presented in Tables 5.8 and 5.9239. By far the
greatest proportion of animals leaving the herd are those sold for cash. During fieldwork the
proportion of the herd sold for cash was 14.6 per cent, which is, compared to other studies, average
to high (Homewood pers com). More male than female livestock tend to be sold - generally as
weaned calves or steers, many of which are destined in the short term for human consumption and the
urban meat market. Prices fetched at auction are seasonally variable and tend to be lower towards
the end of the dry season and early wet when a certain amount of distress selling may occur. Grazing
loans are more complex than initially described and may in fact be based on a number of different
reasons. Frequently a kinsman may approach a male relative or stock associate asking for one or
more cows as a milking loan. The animals will then be loaned to the kinsman or associate who does
not have a sufficient number of cows in his olmarei to support the requirements of his immediate
family. While agreement for such loans is frequently expressed as having been effected in the male
domain, women may often be the instigators and catalysts of such loans especially where close
kinship relations exist between them. Other reasons for a ‘grazing loan’ may include: the loaning of
steers for plough training; avoiding an occurrence of an outbreak of disease (frequently believed to
be associated with a curse) and; access to good pasture nearby the olmarei of a kinsman or stock
associate. A less common occurrence is the direct and permanent exchange of animals (usually of
different age class and/or sex) between ilmarei where a stockowner decides that the herd is lacking
in certain animals or is aware of an imminent social commitment for which he has to plan. Large stock
are only very rarely slaughtered for home consumption and usually as a result of an animal’s sickness.
However, depending on the developmental cycle stage of the olmarei, there may be years where a
significant number of large stock are slaughtered for rituals and/or paid out in bride wealth. Local
Ilparakuyo society still continues to follow traditional forms of social sanction and these too may from
time to time take their toll on a super-household’s herd.
239 These data are likely to have been under-reported for 2000 and certainly appear to have been heavily so
for 2001 – precluding the latter’s use. Nevertheless, the relative proportion of different transactions affected is
likely to have remained relatively robust, justifying the use of the data for the year 2000. An anomaly in the
data was the under-reporting of transactions in male stock which appear to be under-represented on two counts:
(i) The number of males leaving the collective herd is exceeded by females; (ii) The robust and consistent data
generated on stock holdings and herd structure, demonstrate an expectedly skewed sex ratio in favour of
females. A significant proportion of young male cattle that are later removed from the collective herd are not
accounted for in the dataset on collective herd transactions. This disparity only became evident towards the end
of post-field-research data analysis, and therefore, at present, cannot be easily followed up or explained.
182
Table 6.9: The proportion of different transactions reported for large stock leaving the
collective herd240.
Proportion of
Proportion of
Transaction category Number stock
Total Herd
transactions
Sold for cash 327 58.8% 14.6%
Grazing loan out 99 17.8% 4.4%
Fatality/sickness 77 13.8% 3.4%
Bride wealth paid 63 11.3% 2.8%
Exchanged out 4 0.7% 0%
Slaughtered for ceremony 2 0.4% 0%
Slaughtered for home use 1 0.2% 0%
Other - fine 1 0.2% 0%
Total 574 - 25.5%
Table 6.10: The proportion of different transactions reported for large stock entering the
collective herd241
Proportion of
Proportion of
Transaction category Number stock
Total Herd
transactions
Grazing loan in 120 70.0% 5.4%
Bought for cash 29 17.0% 1.3%
Bride wealth received 15 8.8% 0.7%
Exchanged in 4 2.3% 0%
Received as a present or as
3 1.8% 0%
compensation
Total 171 - 7.7%
A further aspect to stock loans already mentioned is that of wealth equalisation. A notable number of
the poorest ilmarei are dependent on their wealthier kin for access to large stock and many are able
240 The proportion of 25.2 per cent of cattle leaving the collective herd would appear very high and
unsustainable. However the figure does not denote the actual proportion of livestock permanently leaving the
collective herd since, largely with the exception of those stock sold for cash, the remaining stock re-enter the
collective herd at different intervals – either to the olmarei from which they left (for example those animals on
grazing loan) or to the herds of other ilmarei (for example, bride wealth and fine transactions).
241 It may be noted that the ‘grazing loans out’ and ‘grazing loans in’ to the collective herd do not balance each
other. This is because the system is not a closed one and many of the ilmarei have kin living not far outside the
field survey area to whom stock are seasonally sent as grazing loans.
183
to increase their herd significantly with stock loans - as depicted in Table 6.11. Nearly half of all
ilmarei falling into the poor stock wealth categories had received stock loans from their wealthier kin
and up to 54 per cent of the wealthiest ilmarei were loaning stock out to their poorer kin - as shown in
Table 6.12. This further demonstrates the fact that livestock loans, particularly from wealthier to poor
ilmarei, contribute an important component for poorer households’ livelihood strategies. Also the more
successful and wealthier an olmarei becomes, the more likely it is that it will loan stock out, in part due
to social obligations.
Table 6.11: The relative increase from stock loans in large stock available to ilmarei of
different livestock wealth categories
Wealthier Medium Poor Very Poor
2000 0% 3.2% ±2.32% 14.2% ±7.44% 12.8% ±4.67%
2001 0% 2.7% ±2.75% 20.6% ±7.08% 20.5% ±6.39%
Table 6.12: The proportion of ilmarei in different livestock wealth categories loaning
large stock in and out
Transaction Wealthier Medium Poor Very Poor
2000 100% 45% 0% 10%

Loans out
2001 100% 54% 6% 0%
2000 0% 18% 24% 40%

Loans in
2001 0% 8% 44% 47%
6.4.3 Farming
Today all ilmarei farm maize plots primarily for food subsistence requirements but more recently
many have also begun to farm cash crops - notably rice - that rely on furrow irrigation. Nearly all242
households (inkajijik) within each olmarei plant their own fields of maize although this does not
preclude sharing of the cultivation of maize fields between households and helping each other in
terms of labour requirements nor later in sharing the crop. Many male heads of ilmarei may also
invest in their own separate fields where the women often have much less involvement in their
cultivation or in the control of the crop. This may frequently be particularly the case for rice, where a
portion of the harvest is frequently retained for household use (often when receiving visitors) and
ceremonial occasions. However, much of the rice crop may be sold for cash. The proportion of
Ilparakuyo ilmarei farming rice paddy is increasing with time as Ilparakuyo rent and purchase fields
through a growing network of relations with rice paddy land owners and farmers. As already
242 In the limited number of incidences where individual households do not cultivate their own crop, they usually
instead contribute to the cultivation effort of other households or that of the olmarei as a whole.
184
discussed in Chapter Five, rice is significantly the most profitable and reliable crop in Idodi, and the
arduous and intensive labour it entails may be eased if the paddy is tilled initially by ox plough.
2000.
1.20
Rice - Acres:Reference Adult
Maize - Acres:Reference Adult
1.00
0.80
Acres/Reference Adult
0.60
0.40
0.20
0.00
9
14
33
30
51
28
37
6
16
10
3
44
36
43
42
45
50
20
12
41
7
13
23
18
35
40
25
48
2
1
5
24
21
46
22
38
8
17
29
49
39
34
15
19
11
27
32
4
26
31
Enkang identifier
Figure 6.10: Areas cultivated by each Olmarei for the two major crops - maize and rice -
in 2001.
1.20
Rice - Acres:Reference Adult
Maize - Acres:Reference Adult
1.00
0.80
Acres:Reference Adult
0.60
0.40
0.20
0.00
33
27
18
28
30
36
9
45
50
20
12
13
17
37
6
23
10
29
3
1
16
31
34
21
14
46
2
25
22
15
43
42
26
51
39
38
44
35
8
48
5
19
11
49
7
24
32
40
4
41
Enkang identifier
Ilparakuyo and Hehe dryland farming practices differ little since the Ilparakuyo have largely
acquired their farming knowledge from their Hehe neighbours. Nevertheless, individual Ilparakuyo
farmers may receive technical advice and help from Bena rice farmer associates in the husbandry of
185
their rice fields and also when the potential application of chemical fertilisers, herbicides and
pesticides requires additional knowledge and expertise. Figures 5.9 and 5.10 show the crop areas243
of maize and rice per reference adult cultivated in 2000 and 2001244 in each olmarei. Although the
variance in the amount that each olmarei cultivates as compared to the amount of stock directly
available is less pronounced, there are still distinct trends in crop area and crop types cultivated
between the ilmarei falling into the four different livestock wealth categories adopted on the basis of
livestock holding.
Generally, wealthier (in terms of livestock) ilmarei tend to farm slightly greater areas of maize per
reference adult than poorer ilmarei and relatively substantially more rice as they are able to afford
the much higher levels of investment required for rice farming - see Tables 5.13 and 5.14. Data on
the relative proportion of wealthier - as opposed to poorer households - cultivating rice is somewhat
inconclusive, not least due to the limited time series. An underlying reason for the fact that rice
farming is not limited to wealthier olmarei is that even some of the poorer olmarei may own or have
access to oxen and plough (via kin) and thus be able to more easily afford cultivating rice, both in
terms of labour and cost savings from the use of oxen. As in the case of poorer Hehe/Bena farmers,
assuming that a household has access to a rice field (which may not necessarily be the case), a key
consideration may be the decision of whether to farm rice, and when this decision is made relative to
the onset of the farming season and in relation to other (dryland) farming commitments. It is likely that
wealthier ilmarei are better able to afford delaying the decision of whether to farm rice or not, as
well as the investment required to expand rice cultivation in wetter years (2001) especially after a
drier one (2000) - see Table 6.14. This is especially the case where losses in the previous year’s
243 Unfortunately it proved impractical to collect data on yields as it was concluded (on the advice of the
Ilparakuyo themselves) that harvesting was frequently piecemeal and protracted over a period of several weeks
as well as carried out by a range of individuals such that establishing yields from recall would be too
inaccurate. Where necessary, average yields have therefore been derived from data collected from the
Hehe/Bena farmer component of the study and are used with reasonable confidence since they are unlikely to
significantly differ overall.
244 Data on individual enkaji (household) cultivation was collected, but for ease of cross-comparison with livestock
data, the data are presented at olmarei (super-household) level.
186
livestock wealth category
Maize Rice Total Average

Wealth Category
(acres per RA) (acres per RA) (acres per RA)
Wealthier & Mediuma 0.45 ±0.663 0.08 ±0.045 0.53 ±0.056
Poor 0.45 ±0.029 0.05 ±0.018 0.49 ±0.027
0.36 ±0.030* 0.03 ±0.012** 0.38 ±0.030***

Very Poor
(df=19) (df=19) (df=19)
a
‘Wealthier’ and ‘medium’ wealth categories are combined due to low sample size in the former category.
*Statistically significant difference as compared with the corresponding cultivation means in the ‘Wealthier &
Medium’ category - One sample T-Test.
livestock wealth category
Maize Rice Total Average

Wealth Category
(acres per RA) (acres per RA) (acres per RA)
Wealthier & Mediuma 0.39 ±0.031 0.11 ±0.031 0.50 ±0.045
0.04 ±0.018* 0.41 ±0.034

Poor 0.37 ±0.031
(df=16) (** df=16)
0.31 ±0.030* 0.03 ±0.015*** 0.35 ±0.033***

Very Poor
(df=18) (df=18) (df=18)
a
‘Wealthier’ and ‘medium’ wealth categories are combined due to low sample size in the former category. Wealth
categories are based on livestock holdings and are the same as those used in the preceding section.
*Statistically significant difference as compared with the corresponding cultivation means in the ‘Wealthier &
Medium’ category - one sample T-Test.
Table 6.15: The proportion of Ilparakuyo ilmarei cultivating rice

Super-households:
2000 2001
(n=140 inkajijik)(N=52 ilmarei)
Proportion of wealthier & medium cultivating rice 29%a 57%

Proportion of poor cultivating rice 35% 38%
Proportion of v. poor cultivating rice 25% 25%
Total Proportion cultivating rice 28% 40%
a
It is possible that this figure may be anomalous either as a result of the poor rain in 2000 or simply
as a result of under-reporting during data collection.
agricultural (dryland) activities may have been incurred and extra livestock may have to be sold to
cover household food requirements arising from shortfalls in grain harvests. Further, when total crop
area per reference adult is taken into consideration, poor and very poor households consistently, and
187
statistically significantly, farmed less per reference adult than medium and wealthier ilmarei in both
years.
The significance of hired labour in cultivation

A very high proportion of households - see Table 6.15 - will employ the labour of especially Hehe
farmers to till both rice and maize fields and to ease labour shortages especially during peak labour
demand times - typically in the early-mid wet season when the herd is calving and fields require
cultivation - planting and weeding245. This would strongly indicate that nearly all ilmarei have a
chronically short supply of labour, particularly during the farming season. Labour is typically hired for
cash and may frequently be for piece work - for example, the tilling of an acre of dryland field for
an agreed sum.
Table 6.16: Employment of labour in Ilparakuyo cultivation
(n=50 Ilmarei)(N=52) 2000 2001
Proportion of fields cultivated with the help of

94.5% 94%
hired labour
Labourers, as well as receiving payment, may enter into long-term client-patron relationships with
ilmarei and frequently will participate on the margins of Ilparakuyo ceremonial events. Long-term
labour relations with Hehe labourers may be critical for Ilparakuyo households in securing labour
during peak labour demand periods when the labourers themselves may be preoccupied with their
own agricultural labour commitments - especially in good years of rain246. As discussed earlier, while
some ilmarei have access to oxen and ox-ploughs - usually the households that are wealthier - many
ilmarei do not and are therefore dependent on reciprocal arrangements with those who do, or on
hiring manual labour. Moreover, the propensity of Ilparakuyo households to hire labour is also related
to a certain distaste humorously displayed towards manual cultivation and the hard labour involved,
not least by the most physically capable, but least inclined, ilmurran.
6.4.4 Livelihood strategies

Production strategies amongst the Idodi Ilparakuyo have become increasingly diverse as they develop
much closer trading, farming and labour relations with Hehe/Bena society. There are also a small
245 The early-mid wet season (and the wet season more generally) is characterised by much elevated levels of
malaria and noticeably increased levels of general malaise that may seriously impact on the ability of
households to carry out herding and cultivation activities. Indeed, such is the chronic incidence of malaria, that
many Ilparakuyo were frequently incapacitated by malaria and only when critically ill would they seek medical
assistance - instead relying on aspirin and induced vomiting (by imbibing ghee) to ease their condition.
246 A number of itinerant farm and herd labourers make their living from working with the Ilparakuyo, staying
and working for one super-household before moving on to another, frequently returning at a later date. Indeed,
some of them are colourful characters – one of whom is a flamboyant musician playing the Gogo ‘zeze’
(CiGogo: zither) instrument much popular at social events in the lives of both the Ilparakuyo and Hehe/Bena
communities.
188
number of Ilparakuyo, not necessarily stockless, who have been drawn to and reside in Iringa town
and who are largely traders in sundry supplies, stock marketing and traditional medicine. These
‘urban’ Ilparakuyo, often through circumstance, have had to diversify out of herding and farming but
very much remain part of the wider Ilparakuyo network - frequently being visited by, and maintaining
close ties with, kin living in the rangelands of Idodi and Pawaga.
Figure 6.11: The relative comparative index value (in cash equivalents) of combined
Ilparakuyo livelihood strategies for 2000
250000
Large stock recruitment in livestock cash equivalents
per reference adult
Small stock recruitment in livestock cash equivalents
200000 per reference adult
Rice produced in cash equivalents per reference
adult
Cash Equivalents (TShs)
Maize produced in cash equivalents per reference

150000 adult
100000
50000
0
31
26
33
14
9
2
6
51
13
12
22
35
34
25
46
23
37
45
49
28
18
1
19
48
43
7
30
38
16
42
36
44
29
11
8
39
41
15
10
3
32
17
24
40
21
50
20
27
4
5
Enkang Identifier
Figure 6.12: The relative comparative index value (in cash equivalents) of combined
Ilparakuyo livelihood strategies for 2001
250000
Large stock recruitment in livestock cash equivalents
per reference adult
Small stock recruitment in livestock cash equivalents
200000
per reference adult
Rice produced in cash equivalents per reference adult
Cash Equivalents (TShs)
Maize produced in cash equivalents per reference

150000 adult
100000
50000
0
31
26
13
9
34
33
46
25
1
22
51
14
35
8
23
15
2
28
12
45
6
19
36
29
18
43
37
49
42
16
32
17
38
48
41
20
27
44
5
50
3
30
7
39
40
11
21
4
10
24
Enkang identifier
189
A synthetic overview of the major components of Ilparakuyo production strategies is useful to compare
the relative socio-economic importance of herding to farming - especially given that all Idodi
Ilparakuyo are, in effect, now sedentarised and recognise themselves as increasingly relying on
farming as a central component of their production strategies.
Indices were developed for each olmarei by calculating the year’s cash value of livestock recruited
into the super-household’s herd (i.e. calves born into the herd) together with the cash value of the
acreage it will have harvested. It should be noted that the data presented here are not fully
comprehensive. In particular, they do not take into account other sources of productivity of overall less
economic significant value - for example, sales of milk and other livestock products, as well as other
miscellaneous sources of income in cash and in kind. The indices are a measure of productivity and the
potential for accumulation, and not necessarily a measure of overall socio-economic wealth. The
values derived are neither absolute nor exact, but serve to provide an indication of the relative
importance of herding to farming and the relative degree of variability in productivity amongst the
Ilparakuyo Maasai of Idodi. Despite the somewhat arbitrary nature of contingent valuation, the
process serves to throw some light on the relative socio-economic significance of herding versus
farming strategies and their implications for Ilparakuyo productivity. The relative variation in
productivity between ilmarei is substantial, with a number of ilmarei far exceeding others - a trend
maintained in both drier (2000) and wetter years (2001) - see Figures 5.11 and 5.12. Although
wealthier ilmarei tend to farm greater crop acreages in absolute terms, the data also reflect the
relative importance of farming as compared to herding for individual ilmarei production strategies.
There is a discernable and statistically significant trend in the average proportion of productivity
accounted for by farming between wealthier and poorer households. The trend strongly suggests that
poorer ilmarei proportionately rely to a greater degree on farming for their livelihoods as compared
to wealthier ilmarei - see Table 6.17. Furthermore, although the data are not independent of each
other, ilmarei falling into the wealthier and medium livestock wealth categories reflect even more
pronounced wealth differentials compared with those falling into the poor and very poor categories
as shown in Table 6.18.
Table 6.17: Relative importance of farming in the production of individual ilmarei
2000 2001
Wealth Category Production Production

accounted for by accounted for by
farming farming
Wealthier &
10% ±2.1% 18% ±2.4%
Mediumc
16% ±2.7%** 18% ±2.1%**

Poor
(df=15) (df=15)
31% ±5.2%*** 32% ±3.3%***

Very Poor
(df=19) (df=19)
* Statistically significant difference as compared with the corresponding

productivity means in the ‘Wealthier & Medium’ category - One sample
T-Test.
190
If the rank order of the households is compared in Figure 6.11 and Figure 6.12, households of middle
wealth category (as defined in Box 6.2 and Table 6.6) cumulatively increase their rank between
2000 and 2001 on average by 2.5 per household. Poor households also increase their rank on
average by 3.8 places, while the poorest show a decrease of 2.6 places on average. Although very
limited for a time series, this data suggest that the poorer an olmarei becomes, the greater the chance
it has of becoming poorer still. However, ilmarei of moderate means may still be able to improve their
livelihood status, particularly in good years, from a combination of herding and farming.
Table 6.18: Overall total productivity indices (in cash equivalentsa, b) of ilmarei grouped
by livestock wealth categories
2000 2001
Wealth Category (cash equivalents per (cash equivalents per
reference adult) reference adult)
Wealthier & Mediumc 125,190 ±10,493 119,128 ±9,786
67,844 ±4,050*** 70,966±5,253***

Poor
(df=15) (df=16)
27,384 ±2,871*** 33,418 ±3238***

Very Poor
(df=19) (df=18)
* Statistically significant difference as compared with the corresponding productivity means in

the ‘Wealthier & Medium’ category - One sample T-Test.
a, b
Cash equivalents have been derived from average local market prices.
a
For the year 2000: 1 Livestock equivalent (184KG) - equivalent cash value TShs 35,000/-247; One
small stock unit on yields on average an equivalent cash value of TShs 8,000248; 1 Acre of rice
yields on average yields 8.7 sacks - equivalent to TShs 104,400; 1 Acre of maize on average
yields 1.6 sacks - equivalent to TShs 12,800.
a
For the year 2001: 1 Livestock equivalent (184KG) - equivalent cash value TShs 35,000/-; One
small stock unit on yields on average an equivalent cash value of TShs 8,000; 1 Acre of rice
yields on average yields 10.8 sacks - equivalent to TShs 129,600; 1 Acre of maize on average
yields 2.8 sacks - equivalent to TShs 22,400.
c
‘Wealthier’ and ‘medium’ wealth categories are combined due to low sample size in the
former category.
247 Large stock, as recruited into the herd, are valued as an estimate of the average contingent sale price they
can be expected to fetch as weaned calves. The value used does not reflect any inter-annual or inter-seasonal
fluctuation in prices. This is because livestock may be sold – or exchanged - in a range of different social
contexts and market locations depending on price differentials. Price differences between years, seasons and
locations are extremely difficult to account for without reliable and often unobtainable records.
248 Small stock recruited into the herd are also valued as an estimate of the average sale price they are
expected to fetch, bearing in mind that different stock categories are often sold at different stages of maturity.
However, there appears to be an anecdotal tendency to sell small stock while still not fully grown and thus
prices fetched may often be lower than those possible – this is reflected in the relatively low value estimate of
small stock utilised here.
191
A number of inferences may be drawn from the data which are supported by qualitative information
drawn from conversations with Ilparakuyo together with field observations. There is likely to be an
increasing disparity among the Ilparakuyo in terms of wealth and socio-economic wellbeing - despite
compensatory socio-cultural wealth equalisation practices249. While some ilmarei retain relatively high
and sustainable levels of per capita livestock holding, and have additionally entered into farming -
successfully combining cultivation with herding over the last 15 to 20 years, a substantial number of
ilmarei are increasingly stock-poor and unable to farm as productively. The issue of rice farming is a
case in point.
Households with access to the most stock are able to farm larger areas of rice, a crop which as
discussed in Chapter Five, can be very profitable. Moreover, rice may frequently - assuming that
irrigation furrows continue to flow especially during drier years250 - be far more reliable and
productive as crop than rain-dependent dryland maize cultivation. However, poorer households, if
they can afford it, may invest in rice cultivation if at all possible, as such an option constitutes a more
reliably profitable production strategy than maize cultivation, not only in terms of deriving a further
form of cash income, but particularly in securing household grain requirements251. However, the
poorest (in terms of livestock) households are simply unable to afford the investment required for rice
cultivation and are compelled to farm maize only, a much less reliable and lower value crop.
The availability of livestock is a strong determinant of inter-annual productivity - especially after a

period of drier years. Availability of livestock may also be strongly related to the position of an
olmarei in the development cycle - as shown in Table 6.19. Although household size may not
necessarily always be an accurate indicator of a household’s position within the developmental cycle,
it is taken here to be a reasonably robust indication.
249 As Rigby (1992, 147) states, radical variations between domestic groups in terms of livestock holdings,
particularly with regard to milk cows, are at least partially levelled out by mutual access rights among kin and
affines, and livestock trustee arrangements (as discussed in the main text). Thus, the concept of ‘inclusive control’
of most livestock (Rigby 1985, 142) may be seen to play at least a partial role in maintaining or improving
olmarei or even enkaji access to livestock.
250 As previously discussed in Chapter Five, the location of a paddy field may strongly influence whether or not
the irrigation furrows supplying water will continue to flow sufficiently and for long enough for the rice crop to
grow properly. Indeed relatively substantial areas of paddy have lain unutilised during recent drier years and
their supply furrows have remained dry as up-stream farmers monopolise the water available. Access to paddy
fields receiving reliable and less contested irrigation water is at a premium and a key factor in securing reliable
and bountiful rice yields.
251 Rice has a consistently significantly higher value than maize (around 1.5 times that of maize) and given that
rice yields (in 80kg bags) per acre may currently be quadruple or more those of maize, the marginal value of
rice production becomes salient.
192
Table 6.19: The impact of olmarei (household) size on production and wealth status
2000 2001
Average herd and farm
Average herd and farm

Reference Adults per
Average Livestock
Average Livestock
Reference Adultd
Reference Adultd
Reference Adult
Reference Adult
productivity per
productivity per
Average Wealth
Average Wealth
Equivalents per
Equivalents per
Sample sizee
Categoryb
Categoryb
olmareia
LE:RAc
LE:RAc
0-5 3.6 2.7 59098 3.5 2.8 62121 13
6 - 10 3.4 2.8 65648 3.4 2.8 65034 22
11 - 15 3.1 4.5 81807 3.0 4.6 93851 11
16 - 20 4.0 1.6 45612 4.0 1.8 55389 3
Note: Household size is assumed to be indicative of the position of ilmarei in the development cycle.
a Households were classified by their total size (in Reference Adults) into one of four inter-quartile groups
b The average wealth category of the households falling into each inter-quartile size category
The average LE:RA ratio of the households falling into each inter-quartile size category
c
d The average combined herd and farm production for 2000 or 2001 of the households falling into each inter-
quartile size category
e The number of households that were classified into each inter-quartile category (n=50).
The data in Table 6.19 show that per capita household livestock holdings and herd and farm
production increase as the household increases. However the data suggest that the largest ilmarei
which are likely to be nearing the end of the development cycle suffer from a marked decline in per
capita252 access to livestock and substantial reductions in their annual herd and farm production. Thus
those ilmarei with sufficient capital accumulated in stock (i.e. those ilmarei tending to fall in the
‘wealthier’ and ‘medium’ wealth designations - or alternatively at the peak of their development
cycle) are able to transfer and invest a proportion of it in seasonally productive irrigated agriculture
especially during wetter years, as an alternative but reliable and effective accumulation strategy.
However, those ilmarei who do not have access to sufficient livestock capital (for example ilmarei at
the beginning or end of the development cycle as inferred in Table 6.19) are far more vulnerable to
inter-annual perturbations in rainfall - both in terms of household food security during drier years and
in being able to mobilise capital during wetter years to take advantage of higher (value) agricultural
production (rice). The greater inter-annual reliability in terms of livestock production as contrasted
with farming is presented in Figures 5.13 and 5.14253. These graphs demonstrate the continued
importance and significance of livestock in Ilparakuyo livelihood strategies as a reliable mode of long
252 In terms of Reference Adults.

253 After analysis, the data points in Figure 6.7 are not clumped in any particular way in terms of household
wealth. The two most extreme outlier points (the first above the line and the second below) belong respectively
to enkang’ 51 and 28. The former falls into the ‘middle’ and the latter into the ‘very poor’ wealth category for
the year 2000.
193
Figure 6.13: The inter-annual variation in livestock (both large and small stock) available
per reference adult
14.00
Livestock equivalents available per reference adult -

12.00
10.00
8.00
2001
6.00
4.00
2.00
0.00
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00
Livestoc k equivalents available per referenc e adult - 2000
Figure 6.14: The inter-annual variation in area cultivated per reference adult
1.20
1.00
cultivated per reference adult - 2001
0.80
0.60
0.40
Acres
0.20
0.00
0.00 0.20 0.40 0.60 0.80 1.00 1.20
Acres cultivated per reference adult - 2000
194
term accumulation and source of capital for short-term, higher risk but frequently profitable,
agricultural investments.
Herd decline and comparative livestock holdings

There is good reason and circumstantial evidence to suspect that the Ilparakuyo herd has declined
substantially over the last 20 years, as supported by both contemporary and historical comparative
data from other locations in Tanzania. If crude livestock indicators are compared for the Idodi
Ilparakuyo against other pastoralists (see Table 6.20), it can be immediately seen that the Idodi
Ilparakuyo have by far the lowest number of cattle per capita at 3.6. There is little immediate reason
to suggest that in the recent past the Idodi Ilparakuyo may not have had cattle holdings similar to the
Ilparakuyo in Bagamoyo District of about nine livestock per capita. If this is the case, then the Idodi
Ilparakuyo have certainly experienced a substantial decline in their livestock holdings. In contrast, it is
likely that the Barabaig in Idodi still retain substantially higher stock levels, and that they may not
have necessarily suffered from the extensive decline in stockholding that the Ilparakuyo in Idodi very
likely have (given the data in Table 6.20).
Table 6.20: Comparative livestock holdings between the contemporary Ilparakuyo of

Idodi, the Il Parakuyo of Bagamoyo District and the Barabaig of Hanang District
Herd Composition
Av. Number of cattle

persons per enkang’
Average number of
livestock in sample
Calves
Total number of
Adult females
per person
Bullocks
Female
(oxen)
Bulls
Male
Ilparakuyo - 2000-2 in Idodia 55.5% 7.1% 6.4% 18.7% 12.2% 11.9 3.62 2,253
Il Parakuyo - Mid-1970s in Bagamoyob 55.0% 11.0% 4.0% 16.3% 13.3% 11.3 8.33 2,160
Barabaig - Early 1990s in Lagaujadc 40.0%d 2.0% 10.0% 32.0%d 16.0% 11.0 5.62 1,231
Barabaig - Late 1980s in Dirmae 40.0% 8% 14.0% 25% 13% n.d. n.d. n.d.
Barabaig - Mid 1970s in Hanangf 55.0% 8.0% 14.0% 14.0% 9.0% 9.5 5.34 2,020
a
Data collected during fieldwork in Iringa District
b
Data collected by Rigby (1983, 134) in Bagamoyo District
c
Lagaujad is in south-eastern Hanang District - data was collected by Lane (1996, 48)
d
The proportion of cows was relatively low, and female calves high, as the herd was recovering from the effects of a
recent drought
e
Dirma lies just to the northeast of Lagaujad in south-eastern Hanang District - data collected by Borgerhoff Mulder
and Kjaerby, cited in Lane (1996, 48)
195
The continuing long term decline in the Ilparakuyo herd in Idodi was a re-occurring theme during
fieldwork. The underlying causes for the herd’s decline are complex but are likely to strongly
comprise an interplay of reasons.
The increasing sedentarisation (particularly post Ujamaa and villagisation) of the Ilparakuyo has
reduced their ability to provide sufficient low disease risk and good pasture for their herds,
particularly as farming competes increasingly for limited household labour. Thus ilmarei are now often
less able to graze their herds greater distances away from the enkang’, a capability which in some
years may be more important that in others, depending on rainfall, range condition and livestock
disease levels. A further contributory factor is that as the extent of farm fields expands, and as the
landscape becomes increasingly divided and zoned, pastoralists are becoming increasingly restricted
in where they may graze their livestock.
The increasing sedentarisation and proximity of Ilparakuyo inkang’itie to farming settlements has led
to growing demands on olmarei herds, as households are drawn further into the village cash economy.
This has resulted in the herd’s productivity being converted more frequently and extensively into cash
for an increasing number of uses such as health, school and village contributions. The herd’s production
is also perhaps spent a little too often by men254 (of all age grades excluding ilaiyok) in beer clubs,
a practice which has contributed, in more extreme cases, to the marked attrition of an olmarei herd.
Finally, as already discussed, the herds appear to suffer from a relatively high chronic disease
burden, most associated with the arrival of CBPP. Despite skilled management by the Ilparakuyo using
traditional and basic modern drugs, tick-borne diseases as well as trypanosomiasis continue to have a
substantial impact on herd morbidity and, therefore, productivity. This observation is supported to an
extent by Rigby’s (1983,160) assertion for Bagamoyo District during the late 1970s that Ilparakuyo
herds were beginning to succumb to an increasing disease load, particularly from East Coast Fever,
as well as trypanosomaisis (see also Rigby 1992, 142).
It is likely that these factors may act in such a way that the adverse impact of each reinforces the
other, in what might be described as a spiral of dwindling pastoralism. Despite this fact, some ilmarei
have remained relatively better off and substantially pastoral. These wealthier households are often
led by Ilparakuyo men and women who, through their own careful and concerted labours, have
persevered and prevailed against sometimes substantial odds - for example, Ismaeli and Nailole
referred to in Box 6.3. That is not to say that other Ilparakuyo households have not also similarly tried,
but, through no fault of their own, encountered less success than others in remaining pastoral - for
example, Salum referred to in Box 6.4.
254 Several ilpayian and senior ilmurran reflected that alcohol had played its part in the decline of their herds
and their livelihoods. It is not possible, however, to draw any conclusion as to whether the varying levels of
alcohol use that exist in Ilparakuyo society in Idodi have grown, or whether they are as much as they ever were.
196
Overall, it is very likely that the sedentarisation of the Ilparakuyo in Idodi, as an indirect result of
villagisation and as a more direct result of an increasingly bounded landscape created as a result of
government policies, has led to long-term declines in their herds. A similar process of sedentarisation
and stock loss was directly associated with villagisation as independently observed by Rigby,
Parkipuny and Ndagala in Bagamoyo District (Rigby 1985, Parkipuny 1975 & 1979, Ndagala 1974
& 1986 cited in Rigby 1992, 146-147).
6.5 Conclusion
The rangelands of Idodi to which the Ilparakuyo first came some fifty years ago have undergone
substantial change. They have been demarcated into different land-use zones, and those parts of the
landscape still open to settlement have been occupied and developed by a diverse and growing
assemblage of farmers and herders. The Idodi Ilparakuyo, in contrast to their initial semi-transhumant
existence in a lightly populated landscape when they first entered Pawaga and Idodi, have become
largely sedentary and now live around the periphery of expanding farming settlements. While
maintaining seasonally spatially variable grazing patterns, the Ilparakuyo are increasingly relying on
farming as a key component of their production strategies - both for cultivation and dry season
livestock grazing. The agro-pastoralist Ilparakuyo have suffered from marked declines in their herds
as they have become confined to the margins of agricultural settlements and as the herds have
succumbed to the influences of a cash economy and to higher levels of disease. A growing proportion
of Ilparakuyo ilmarei no longer have access to sufficient livestock upon which to base their production
and instead have, for all intents and purposes, become agriculturalist. While those Ilparakuyo still
endowed with larger herds have been able to successfully adapt and invest in higher value and more
reliable forms of agricultural production, a growing underclass of poorer ilmarei - despite the
continued practice of traditional forms of wealth equalisation among kin - are increasingly dependent
on much more unpredictable and lower value dryland farming as a key production strategy. As
overall per capita livestock holdings continue to decline, there is likely to be increasing wealth
differentiation between the relatively richer Ilparakuyo domestic units, and those who are the poorest.
This trend may be seen to be a result of past and present state development policies and the
emergence of an increasingly strong formal market economy in which the social inclusivity of
pastoralist production has diminished (see Rigby 1992, 164).
Although this analysis has not adopted a gender-focused approach to household production, the
results of this study in the light of previous findings on Ilparakuyo production strategies, beg an
important question. Rigby (1983, 163) predicted that growing levels of cultivation in Ilparakuyo
society would result in a reduction in the status of women. The reason he made such a prediction was
that, as the importance of cultivation in relation to livestock keeping for household production
increased, women could varyingly lose their once stronger position as regulators - or mediators - of
household production. Rigby (1983, 163) found that olmarei production had become increasingly
focussed on agricultural land, labour and commodity relations, placing the major portion of
production - particularly in stock-poorer households - in the hands of men This process reflects a
similar loss of control by women over production and a reduction in their status in Kisongo Maasailand
197
described by Hodgson (2001, 68 & 253). It would be appropriate to investigate this issue as part of
further research on the Idodi Ilparakuyo.
Finally, the Ilparakuyo, while remaining an independent and distinct cultural identity, have developed
increasingly close social ties and client-patron relationships through labour and exchange with the
wider hegemonic Hehe/Bena farming community. As they become totally reliant on their access to the
finite grazing and farming resources of the Idodi rangelands, the future of the Ilparakuyo is entirely
bounded by the measure of success with which they will continue to negotiate and secure access to
these growingly contested resources through their social relations with the Hehe/Bena farming
community. This may be contrasted with the situation for the Barabaig. Although some Barabaig
families now follow similar livelihood strategies to the Ilparakuyo, other families remain far more
mobile in the landscape. It is these latter Barabaig who will likely find it increasingly difficult to
continue their current way of using the landscape as other herders and farmers seek to limit their
movements which are seen as a source of conflict over land use.
198
7 Negotiating the political ecology of landed resources
Megel nkishu ‘pere - do not separate cattle with a spear (you may destroy many)
Ilparakuyo proverb
7.1 Introduction
In this chapter I investigate the nature of land-use relations between farmers and herders, in the
context of Peters’ (2004) assertion that there is growing inequality, competition and conflict over land
at local level. As the Idodi landscape has become increasingly peopled, so the level of contest for
access to key resources - for example, arable land and dry season on-farm grazing - has increased.
These resources continue to be perceived differently by herders and farmers in terms of rights of
access and cultural norms of use, leading to dispute and, in recent years, sometimes violent conflict.
While most Ilparakuyo and some Barabaig herders have successfully begun to avoid conflict as they
engage with farmer-based perceptions of resource access and use, others - especially transhumant
Barabaig herders have continued repeatedly to enter into conflict situations with farmers.
In attempting to understand land-use practices and land-use conflict in Idodi, it is important that a
simple but robust analytical framework be adopted. In this regard, Goheen and Shipton (1992, 309-
311) identify three key, yet straightforward, sets of questions which provide an entrée to
understanding the socio-ecological complexity of land holding - an approach particularly
appropriate in the context of a polyethnic landscape in which multiple and contested forms of land-
use and production co-exist:
What does land mean and to whom?’ - What kinds of resources do people use and
how are land and its resources defined and categorised in local cultures?
What kinds of social affiliations affect land-use and control?’
Who controls the terminology?’ - Who gets to interpret and define the meaning not just
of land per se but of ‘the group’ itself?
As Goheen and Shipton (1992, 309) note at the outset, the answers to these questions are rarely
simple. Any analysis may only necessarily be able to draw out a particular set of contextual nuances
and leave others untouched. This chapter therefore employs Goheen and Shipton’s three questions as
a basis for providing some ethnographically grounded and focused analysis of how the Idodi
rangelands are used today by herders and farmers, who controls their use, how this use is decided,
and why and how land-use conflict continues to occur.
In this chapter I deconstruct some key components of the complex socio-cultural milieu underlying and
moulding current trajectories of landscape occupancy that, nested within a wider policy context,
favour farmer over herder. Thus, drawing on the parallel themes developed in Chapters Four and
Five, I examine in this chapter how the landscape has come to be defined in terms of herder-farmer
relations, within the context of a demarcated landscape, much of which has been reserved exclusively
199
for wildlife conservation. While seasonal herder-farmer conflict remains a prominent feature of
landscape relations, I investigate through the use of extended case study, how social relations and
negotiation between herder and farmer have become central for modulating land-use agreements
and in influencing landscape and livelihood outcomes. I will argue that while conflict over land has
increased in recent years, negotiated land-use agreements between farmer and herder have helped
to resolve - or lessen - these conflicts. I argue that the local legitimacy of these agreements has only
been possible as a result of the long-term and increasingly interdependent nature of socio-economic
relations between herder (particularly the Ilparakuyo) and farmer. In villages in which land-use
conflict has recently broken out, imposed and hastily contrived land-use plans - largely at the behest
of the State, have not improved land-use relations, as they have failed to address underlying tensions
and the socio-political dimension of land-use conflict. I will show how land-use plans as envisioned by
the authorities cut across existing production relations between herder farmer, while locally based
agreements may form more of a flexible agreement with fuzzy boundaries in which socio-economic
relations and networks are allowed to continue. Despite the relative success of locally negotiated
land-use agreements, I argue that herders’ access to key landed resources for their pastoralism
remain insecure. While negotiated agreements have enabled their continued access to range
resources, they are liable to reversal as farmers seek to secure more land for an expanding
population, and as land-use intensifies with an expanding irrigation network. I argue that negotiated
land-use outcomes are alone insufficient for herders’ access to key range resources and that other
forms of action are required if their access as a marginalised and minority group is not to deteriorate
further. The insights I draw from the extended case studies in this chapter are then combined with
those from the preceding chapters to discursively answer the central questions I asked in Chapter Two
- leading into a final conclusion centred around Peters’ (2004) discussion about inequality and social
conflict over land.
7.2 The socio-ecological context to resource dispute: what does land mean and
to whom?
In Idodi, people’s past origins and experiences may often bear heavily on their current perceptions
and attitudes towards the landscape, land holding and production. Many people living in the Idodi
villages, particularly the older generations of newcomers, have memories of eviction and land
insecurity. For example, some newcomer Hehe families were forced to move up to three times in the
space of twenty years due to state interventions (see Chapter Five). While insecurity is remembered
in the past by Hehe newcomers, insecurity has become an increasingly key issue for herder newcomers
in the present, who continue to remain little more than squatters on village commonage. Previously this
de facto status was not viewed as a constraint by Ilparakuyo herders, as mobility was more important
for their livelihood strategies. However today, as Ilparakuyo and Barabaig herders have become
increasingly agro-pastoralist and thus more sedentary, access to secure land and grazing rights in
their local village commonages has become a critical issue. The growing importance for herders of
obtaining secure tenure of grazing land may be seen as part of the continuing conflicts of interest and
discord between herders and farming peoples in the Idodi villages over access to key resources.
200
As farmers and herders have come to live together in the Idodi villages, each group has brought their
own socio-ecological, economic and cultural understandings of landscape occupancy. These different
normative perceptions, although flexible and negotiable, have variably led to conflicts of interest and
tensions between herder and farmer over what are acceptable and unacceptable uses of the
landscape, and more fundamentally, which group has socio-political power over the other.
In the following two sections, I first review some of the key findings of Chapters Five and Six that are
particularly relevant for understanding the socio-ecological context to dispute, before I continue on to
examine the nature of complementing and conflicting herder and farmer land-use practices.
7.2.1 Farmers: expanding frontiers, fertility decline, and new technology

For farmers, the Idodi rangelands have been a source of fertility and an expanding frontier for
agriculture. The Idodi rangelands are today characterised by two very different forms of agriculture
- dryland and wetland farming - as discussed in Chapter Five. Dryland farming can be perceived as
representing a long-established means of production as practised by Hehe farmers throughout the
Idodi rangelands. However, whereas dryland farming was formerly practised in an extensive way,
with the populating of the Idodi rangelands and the villagisation campaign of the mid 1970s, farmers
increasingly came to cultivate the same fields year on year as the remaining uncultivated land was
allocated by village governments to immigrant farmers. Continual cultivation of dryland fields has led
to a slow but continual decline in the overall fertility of the dryland farmland. Declining levels of soil
fertility may be compounded by the unpredictable rainfall regime leading to heightened levels of
risk in dryland farming. Yet despite the unpredictable nature of dryland farming in a semi-arid
rangeland, the area under dryland cultivation continues to form an expanding frontier as the farming
population has grown, augmented by a continuing inflow of immigrant farmers seeking land (see
Chapter Five).
Wetland rice agriculture was brought to the Ikwavila valley in Idodi from the Usangu in the 1970s255.
Adopted particularly by newly arrived Bena and Wanji farmers, the riverine wetland area under
rice cultivation expanded through the 1980s and 1990s as the growing irrigation furrow system
seasonally allowed. Although wetland agriculture requires substantially higher investment in labour
and agricultural inputs than dryland farming, it can be very profitable (see Chapter Five), and it has
led to the emergence of a wealthy entrepreneurial class of farmer. These farmers now represent a
new face of agricultural production and prosperity in the Idodi rangelands, one which is hardly
constrained by fertility and is not as limited by rainfall perturbations as dryland agriculture.
Many dryland farmers are unable to successfully take-up wetland agriculture due to its labour and
input requirements as well as due to a shortage of easily available irrigable land. Therefore dryland
agriculture remains the most important form of livelihood for many farming households. The dryland
255 It is probable that wetland rice technology was also brought to other parts of Idodi from Pawaga, itself a
significant rice growing area.
201
soils are extremely difficult to maintain - both in technical terms and also in terms of the availability
of labour and inputs. With growing populations, the increasing commodification of land and its rising
value, there is a new and growing class of landless farmers who are compelled to rent land from
others in order to cultivate (see Chapter Five). It is likely that increasing numbers of destitute farmers
are entrapped in a cycle of poverty and poor agro-ecological productivity, unable to sufficiently
support themselves from the land, even in years of good rainfall. In the face of these challenges,
farmers are increasingly clamouring for the political right to cultivate areas informally ‘reserved’ for
pastoralists.
7.2.2 Herders: seasonal grazing systems, demarcated landscapes and expanding fields
For pastoralists, the Idodi rangelands provide a variety of both wet and dry season grazing range
as well as corridors of movement between different grazing areas. As for pastoralists in any semi-
arid rangeland, the spatial and temporal variability in rangeland productivity is often a key driver in
herder landscape occupancy. Thus seasonal flexibility in rangeland-use remains essential for the
viability of Ilparakuyo and Barabaig herds in Idodi (see Chapter Six). Sufficient access to dry season
grazing areas is an important determinant of the productivity of their livestock. As was concluded in
Chapter Six, the status of pastoralist herds is a key determinant of the ability of pastoralist
households to invest successfully in agricultural - particularly wetland - production. In this regard,
many pastoralists, particularly the Ilparakuyo, are becoming increasingly agro-pastoral, investing in
agricultural production, particularly in wetter years, and then re-investing any surplus back from
farming into their herds. Thus agricultural production now constitutes an important part of many
pastoralist livelihoods and forms a key determinant of overall pastoralist livelihood success.
As the landscape has become increasingly demarcated and populated, so the ability of herders to
seasonally move in the landscape to maximise the reliability of their grazing has become increasingly
difficult and constrained. The rangelands available for grazing have shrunk as the area under
cultivation has expanded as a result of population growth, and as the areas reserved for wildlife and
restricted in one way or another to people, and in particular pastoralists, have been extended. In
several of the Idodi villages, pastoralists have found themselves compelled to withdraw into the
remaining range lying between farmers’ fields and the proscribed parts of the Lunda section of the
LMGCA (see Chapter Four). This has also meant that whereas agricultural fields did not previously
form an important part of the annual cycle of grazing patterns in the Idodi rangelands, pastoralists
have now in incorporated farmer’s fields into their grazing cycles, to the extent that field crop
residues have become a key, indispensable, grazing resource during the dry season.
The agro-pastoral Ilparakuyo have responded to these developments through achieving higher levels
of assimilation into farming communities, particularly by engaging in client-patron relations. A number
of the Barabaig have much less so, leading to easier expression of differences of interest and to
open conflict and violence over access to dry-season grazing. Driven by shortages in grazing, an
increased livestock disease burden (see Chapter Six) and heightened demands of a growing
population on a diminishing herd, many Ilparakuyo Maasai are today investing more in agricultural
production to augment their livelihoods. In contrast, many Barabaig have remained highly mobile,
202
exploiting the range in seasonal patterns contingent on inter-annual climatic variation and local socio-
political relations. Other Barabaig have settled in localised areas and are engaged in more
sedentary agro-pastoralism, and follow similar livelihoods to the Ilparakuyo.
7.2.3 Farmers and herders: complementary and conflicting land-use practices

Parts of the Idodi landscape are used by farmers and herders at the same time of year; other tracts
are independently used by both groups in different seasons through different periods of the year.
The simultaneous use by herder and farmer of similar resource sets may occur with or without
disagreement dependent on the relations between them, and contingent on their respective
understandings of how these resource sets should or can be used.
Defining terms and describing the nature of disagreement

Before proceeding, it is important to clarify different terms - or labels - that may be used to describe
different levels of disagreement between people - for example, between herders and farmers. The
objective here is to use definitions that are appropriate for a simple yet robust analytical framework
of disagreement, drawing from the work of others. Thus the definitions and concepts defined and
described below are adapted from Gulliver (1979, 79) & Nader and Todd (1978, 14-15) cited in
Hagberg (1998, 68); and Hagberg (1998, 20 & 68).
A ‘claim’ is defined as an alleged infringement of a law or customary rule or entitlement perpetrated

by one party against another. For example, if a farmer finds that his/her field has been grazed
without his/her consent, then the farmer may make a ‘claim’ against a herder who is alleged to have
perpetrated the incursion. The substance of the ‘claim’ may often be negotiated and privately
resolved within a dyad. A ‘claim’ may sometimes become public knowledge, particularly if tensions
are high between herder and farmer, making a private and relatively quick resolution less easy.
A ‘dispute’ occurs when the person to whom the ‘claim’ has been directed denies culpability and the
disagreement becomes concrete and public (Gulliver 1979, 79 cited in Hagberg 1998, 68). The
‘dispute’ is most likely to be resolved by adjudication through jural proceedings - usually by village or
ward level courts, or in other circumstances, brought before a customary jural body (see Chapter
Four). However, depending on the circumstances, disputes may not always be brought before a jural
body, and may remain unresolved for long periods of time, to be re-expressed at a later date.
A conceptual distinction can thus be made between a ‘claim’ and a ‘dispute’. A ‘claim’ may be seen to
be mediated often privately through social relations, which avoid adjudication. The disagreeing
parties may perceive greater benefit in privately resolving their disagreement through negotiation
thereby maintaining good social relations with each other in light of wider socio-economic relations.
Contrastingly, a ‘dispute’ becomes publicly expressed as a perceived breach of legal right or the
law, and is most likely to be adjudicated by a jural institution. Adjudication may not necessarily lead
to the resolution of the underlying disagreement, and may only serve to maintain or worsen the
damage incurred by the dispute to social relations (see Gulliver 1979, 6-7 cited in Hagberg 1998,
70).
203
Finally, ‘violent conflict’ occurs when either the claimant/disputant or the defendant or both feel
compelled to defend their perceived rights or express their grievance through violence. Although
‘violent conflict’ may occur spontaneously over a relatively small conflict of interest, it is likely to be as
much an articulation of a larger set of circumstances or conflicts of interest.
A disagreement can develop its own dynamics - bouncing back and forth among and between levels
and parties, and may have implications far beyond the actual dispute (Nader and Todd 1978, 15
cited in Hagberg 1998, 68), impacting on how other (similar) disagreements unfold and are
approached by the claimants/ disputants.
Complementary land-use practices

In some instances, herder and farmer understandings of normative resource use can concur. For
example, farmers’ tenurial rights over their fields are respected and enforced through popular
sanction during wet season agricultural growing period. However, once a field is harvested, a herder
may secure grazing rights for crop residue from the farmer during the late wet season and or in the
early dry season. The grazing rights for a field are normally exclusive and are negotiated between
farmer and herder for a payment usually in cash but sometimes in kind. Despite the repeated
iteration by District authorities that such practices are now illegal, farmers voluntarily enter into
grazing agreements with herders as they may often form part of a wider client-patron relationship,
which may also encompass labour and drinking relations. These agreements also continue despite
farmers maintaining that cattle may damage soil fertility256 (see Chapters Four and Five). Farmers
also complain about the incidence of incursions in neighbouring dryland fields, which may or may not
be resolved quickly between herder and farmer (see below).
Although dry season field grazing can constitute a common negotiated and accepted land-use
practice for pastoralists, it has also become a source of conflict in some villages in recent years.
Sometimes a herder may assume rights to graze a field without the consent of its owner, and if
detected in time, this may lead to a dispute, depending on the type of field and whether the farmer
perceives it worthwhile or possible to pursue the dispute (see Section 7.4.3). Farmers associate this
practice most with pastoralists - particularly the Barabaig - some of whom who seasonally arrive in
the Idodi rangelands with their herds during the dry season to graze their stock on farmers’ fields.
Idodi is only one of several dry season grazing areas that these transhumant herders may use. Thus
these herders may pursue a more opportunistic strategy in which fields may be grazed without the
prior consent of the right holder. Since these herders are not ordinarily members of the village, they
may stand to loose relatively little by breaking normative expectations, which lead to poor social
relations with farming communities. Resident herders tend to conduct their gazing strategies in an
256 As discussed in Chapter Five, farmers did not perceive the manure left by cattle grazing fields as important
or significant for improving soil fertility.
204
opposite fashion, paying greater deference to normative field grazing practices, as the maintenance
of good social relations with their farming neighbours and associates is key for their livelihoods.
Conflicting land-use practices

There are other instances in which herder and farmer understandings of normative resource use do
not concur. Perhaps the most significant example of discordant landscape use in the Idodi rangelands,
aside from field grazing, is that of fire management257. Hehe farmers typically set fire to the grassed
areas of the rangelands in the early dry season for a number of reasons. Fire use has strong roots in
declining and perhaps previously more important uses of the landscape, such as wildlife hunting and
honey gathering. Setting fire to the undergrowth of the Acacia-Commiphora bushlands is perceived as
enabling hunters to more easily track and hunt down their quarry. Also, for honey gatherers, the
smoke from the ‘green’ fires may drive bees away from their wild hives, often located in baobab
trees. Early dry season fires tend to be cooler fires (than those later in the year) as much of the
vegetation still has residual moisture, and therefore woody plants, which are often fire-hardy, survive
without damage. However, for pastoralists, these early dry season fires are often a nuisance as they
result in substantial areas of early dry season grazing being burnt and rendered useless for their
livestock. Pastoralists are then forced to move their herds elsewhere. The early fires lit by farmers
represent an element of range insecurity for herders. In addition, these early cooler fires may result in
the maintenance of a more wooded landscape, as young trees especially are less susceptible to fire-
based mortality from these events. For pastoralists later burning of the already grazed range,
although depending on climatic conditions, may be more preferable. Later burning may not only
conserve the early dry season range, but it also can open out the landscape, limiting tree recruitment
and therefore improving and expanding the rangeland’s grasslands.
The divergent fire management interests of herder and farmer create underlying tensions, with
pastoralists perceiving farmers as wilfully mismanaging fire so as to ruin grazing areas as a pretext
for discouraging a herder presence. Because fire management does not constitute a readily
recognised basis for dispute among the Hehe, there are few avenues of recourse available, or
precedents set, for pastoralists to contest the use of fire in grazing areas, in which they also have no
recognised formal permanent tenurial rights. While village governments will sometimes arrange
collective fire-fighting initiatives for large and out-of-control fires, particularly those close to
settlement, these initiatives tend to be predicated on the normative perceptions of district extension
officers and other educated district council employees who tend to perceive most fire as being
undesirable and destructive. Thus these fire fighting initiatives, when they occur, are not specifically
carried out for pastoralist benefit, although herders may sometimes benefit from them.
257 See Laris (2002) for an insightful paper on mosaic burning in Mali, which stresses the need for understanding
not only when and where fires occur, but why they are lit, what they are lit for and who lights them. Far from
being a solely destructive agent, fire may be crucial for rangeland management. It is often used at different
stages of the dry season by local farmers and herders to create a mosaic and diversity of different vegetation
types maximising the utility of a rangeland (although this is not to say there may be conflicting interests over fire
management – as exist in Idodi).
205
Changing land-use priorities
In addition to specific overlapping land-uses, such as grazing and fire which may or may not lead to
conflict, there is a further dimension to the coexistence of alternative forms of production and
understandings of landscape that farming and herding represent. In particular, the expanding
farmlands of Idodi are beginning to impact on the way herders use the landscape and on how the
remaining landscape can be used. There are long-term farmer hopes that the area under irrigation in
Idodi will be further expanded. These expectations are reflected in national policy which sets out
objectives to expand agricultural productivity and improve national food security particularly through
rice production (Agriculture and Livestock Policy 1997, 48 - 51; National Irrigation Development
Policy 1997) 258. Thus the lack of access to irrigated wetland agriculture, particularly for many poorer
farmers in the lower reaches of the Ikwavila Valley, has been the recent focus of a World Bank
funded irrigation development project (World Bank 1996, MAFS 2002). The project’s aim was to
more efficiently tap the flow of the Mazombe River, and through infrastructural development, further
expand the area under irrigation.
There are a number of implications arising from the expansion of irrigation in the landscape that,
although not directly expressed by herders or farmers, can be identified through the analysis of field
data. The description that follows is not necessarily based on herder and farmer descriptions, but it
fits in and supports other farmer and herder narratives discussed later on in this chapter.
While herders and farmers have benefited from negotiating seasonal dryland field grazing
agreements, the expansion of wetland irrigated agriculture presents a different situation. For
example, during the early dry season of 2001, herder field grazing was tracked by randomly
sampling post-harvest field grazing. Post-harvest dryland field grazing was found to constitute 76
per cent of all field grazing of fields sampled (241 fields) in that year259. Although only 24 percent
of fields sampled were wetland fields, nearly all these fields were grazed by Barabaig herders, who
that year were also involved in violent conflict with farmers specifically over wetland field grazing
(see Table 7.1). Court cases heard by the Baraza la Mahkama ya Kata in Idodi between the years
2000 and 2001 show that the Barabaig were relatively heavily prosecuted - they were defendants
in 36 per cent of all court cases260. In contrast, no Ilparakuyo came before the court as defendants.
258 The Agricultural and Livestock Policy (1997, 48) states, ‘…irrigation seems to hold the key in stabilizing
agricultural and animal production’. Expanding the area under irrigation has been realised to a certain extent in
the Idodi villages through a World Bank funded smallholder irrigation project in Mapogoro and Nyamahana
villages which was started after field work had been completed (URT 2002) – see Chapter Three.
259 A list of households in Idodi village was compiled and every fourth household was visited and asked about
the details of any fields they had rented out for grazing.
260 In one case alone, a Barabaig herder was penalised TShs 255,000 in fines and TShs 49,000 in court charges
(a total of TShs 304,000 or about USD$380 at the time).
206
Table 7.1: An inventory of court cases heard by the Baraza la Mahkama ya
Kata for 2000 and 2001
Type of Case Number Percentage
Farmer-herder (Barabaig) dispute 16 18%
Robbery/theft 10 11%
Environmental degradation and/or water

8 9%
pollution (Barabaig)
Illegal incursion into village area (Barabaig) 8 9%
Gross bodily harm 8 9%
Slander 6 7%
Farmer-farmer land dispute 5 6%
Domestic dispute (Farmer) 5 6%
Divorce (Farmer) 4 4%
Debt 3 3%
Farmer-herder (Ilparakuyo) dispute 0 0%
Other 16 18%
Total 89
In contrast, no conflict occurred in the dryland areas, which for the most part were grazed by
Ilparakuyo herders, although this is not to say that claims and disputes do not occur over dryland field
grazing. The data suggest an underlying trend as follows. Dryland field grazing constitutes a
temporally and spatially delimited use of the landscape over which herder and farmer can have
common interest; wetland agriculture instead constitutes an increasingly exclusive zone for farmers in
which herders’ access is likely to be much diminished.
The phased expansion (realised and anticipated) of wetland fields and the area in and around which
herders’ potential for access is being challenged, has precipitated pressure by village councils,
popularly supported by much of the farming village assembly, for pastoralists to move out of the
upper river valley reaches of some of the Idodi villages. In other Idodi villages, pastoralists occupy
rangeland that is not suitable for irrigation and therefore they are not as subject to this pressure.
Thus, through overt pressure in village council meetings, from which pastoralists are often absent (see
Chapter Four) but at which their removal is often called for, pastoralists are being slowly compelled
to move down into the drier rangelands that form part of the LMGCA, despite substantial resistance
on their part. The withdrawal of pastoralists to the periphery is being formalised through land-use
planning facilitated and initiated by the district council and partner institutions, in response to land-
use conflict and incidences of violence. These more extreme conflict events that have occurred in Idodi,
and which form a central part of the rationale behind district sponsored land-use planning, are those
which are nearly all centred around wetland field grazing dispute. In addition, the separation of
207
herder and farmer domains is part of a wider policy initiative by the government in its reaction to
preventing and solving protracted and sometimes violent land-use conflict between herders and
farmers that has occurred elsewhere in the country (e.g. Rural Development Strategy 2001, 32;
Agricultural and Livestock Policy 1997, 62-68).
Summary
As production intensifies and land-use in the more productive areas of the landscape becomes more
exclusively reserved for irrigated agriculture, so the potential for divergent understandings of the
landscape between herder for herding and farmer for farming continues to grow. Farmer-based
understandings of the landscape and control over its use stem from overlapping local power centres
both of locally powerful lineages and village councils. This power base which underpins prevailing
landscape meanings and understandings has substantial support from the government in its drive for
improved agricultural production. As farmers are well aware of the considerable increase in
productivity and reliability that wetland farming can constitute over dryland, their understandable
hopes for the continued expansion of intensive wetland production are supported by national
government and global institutions - such as the World Bank - through irrigation expansion
programmes. It is therefore of little surprise, that in the face of locally prevailing farmer-based
meanings of landscape supported by national agendas, that pastoralists find themselves retreating to
the spatial periphery of the landscape and their understandings of landscape management
marginalised261.
7.3 Social affiliations, land-use and conflict
7.3.1 Law and social process

The new land law of Tanzania potentially provides a legal framework for engendering equitable
rights over land for all at village level, but it is clear that the law in itself is insufficient to guarantee
these rights. There are two reasons for this. Firstly, as discussed in Chapter Three, the State continues
to heavily control and define the lie of the landscape and retains the power to appropriate land
from villages. Secondly, the underlying socio-political processes at village level are as important, if
not more so, in defining people’s access to resources on village land. Different socio-political
261 This is not to say that that the government does not view herders as potentially important producers. The
government has identified livestock as a nationally important yet under performing sector. In short, the State
currently plans that the livestock sector will grow from contributing 2.7 per cent to 5 percent of Gross Domestic
Product by 2005 (RDS 2001, 32). Several strategies are identified including the creation of Disease Free Zones
and the demarcation of permanent grazing lands to put an end (in effect) to pastoralist transhumance and
mobility. These permanent grazing lands are supposed to have set carrying capacities (Agriculture and Livestock
Policy 1997, 55). An underlying rationale is to promote the (commercial) production of export beef, in line with
export-orientated macro-economic policies. Insufficient attention has been given to supporting pastoral
livelihoods based on a milk-producing economy, where livestock may be accumulated for a range of reasons
including long-term household livelihood security (given frequent droughts) and as an alternative form of
savings.
208
affiliations and interests use and interpret the law (when the law is known and applicable) for their
own ends, and may often fill the legal vacuum that the lack of application of formal state law often
leaves with their own customary and socio-cultural interpretations of how the land should be used,
and who should benefit.
The social affiliations that modulate land-use and its control in the Idodi rangelands are, as might be
expected, locally diverse, complex and fluid. Attempting to provide a comprehensive ethnography of
these social affiliations would be hardly possible in a single field study, especially given the
polyethnic nature of society in the Idodi rangelands. However, it is possible to provide a selection of
cross-cutting case examples about the on-going social affiliations and processes that have led to
different farmer-herder outcomes in the landscape. I have chosen here to focus on farmer-herder
land-use relations, as herders are a significant minority group who remain marginalised, but who
have varyingly negotiated access to key landed resources.
There is a noticeable degree of variation in land-use relations and land-use control between herder
and farmer in the Idodi villages. In some villages, herders and farmers have negotiated an
understanding of how the landscape can be used. These understandings have evolved over time,
remain fluid and are still prone to underlying tensions and conflicts of interest over land-use in these
villages. Yet, dispute is often purposefully avoided and when it occurs, it is usually resolved relatively
quickly. Violent conflict very rarely occurs, if at all, in these villages.
In other villages, herders and farmers are engaged in a more acute and prolonged struggle over
land-use and access rights, and disputes are often more frequent. In one village in particular, Idodi,
there have been repeated outbreaks of violent conflict, which are likely to be reflective of peoples’
low expectations of sufficiently resolving their underlying conflicts of interest through local jural
institutions or, in the longer term, through social relations. While the different physical geographies of
the villages are in part a contributory factor, it is the underlying matrix of social affiliations and
production relations between herder and farmer that arguably accounts most for these different
outcomes.
I present and use two sets of extended case studies to illustrate and discuss the very different herder
and farmer land-use outcomes in different parts of the Idodi rangelands.
7.3.2 Social affiliations and negotiated outcomes between herder and farmer
The following two case studies show how herding and farming groups have negotiated a rights-based
land-use agreement regulating access to key resources and the management of the landscape. I use
the term ‘rights-based’ to mean that farmers or herders are allocated specific use-rights (including
settlement) over an area of land or zones - whether for grazing or farming. In public discourse, these
land-use zones often infer exclusivity for one or other land-user - particularly from the standpoint of
the state. In this regard, herders may re-negotiate access to key resources from which they have been
formally precluded in the rights-based land-use agreement through using their own individual
networks.
209
Group: herder - farmer and land-use agreements in Mahuninga ward
Mahuninga ward consists of two villages - Makifu and Mahuninga (see Chapter Four). Makifu village
has among some of the richest and most fertile cropland in the Ikwavila valley. There are three major
agro-ecological zones: the irrigated wetland fields, the dryland fields and the non-cultivated
bushland used for grazing. Several seasonal tributaries of the Tungamalenga River that are
seasonally used to irrigate rice paddy run through the village area. One of these tributaries forms
part of the boundary between the relatively intensively cropped fields to the west and north and the
higher ground to the south east where, in addition to some dryland farming, the Mahove bushland
begins. The village has four sub-villages: Isanga in the northeast; Makambalala to the northwest;
Makifu to the west and; Mkanisoka to the south-west - beneath the foot of Mkanisoka hill. A
longitudinal ridge of hills in the west forms a physical boundary between farmland to the east and
the village’s wilderness area to the west that also forms part of the Mkupule section of the LMGCA
(see Figure 4.2).
Mahuninga village lies south of Makifu, stretching to the end of the Ikwavila valley. It is bounded on
three sides by hills, with the Mahove bushlands forming the north-eastern periphery of the village. The
village has six sub-villages, of which Uyamba and Majengo in the north border with the Mahove
area. Mahove has an inexact boundary to the west and to the south follows the course of a tributary
of the Tungamalenga River.
In comparison to other villages in the Ikwavila Valley, a relatively large number of Bena, Wanji and
Kinga farmers live in the Makifu sub-villages, especially Makambalala. While farming is the major
livelihood occupation in the village, a number of these families also keep livestock, particularly in
Makambalala, where there are twelve agro-pastoralist households.
Ilparakuyo herders have had a presence in Mahuninga Ward since the late 1960s and currently live
outside the farmer settlements in scattered enclosures on the western and southern margins of the
Mahove bushland area. There are up to eight Ilparakuyo enclosure clusters in the Mahove area262,
which are organised on the basis of clan affiliation as shown in Table 7.2. Some of the families and
clans - such as the Ilmarumai and Inkelingishu - were the first to arrive in the area and have retained
a presence ever since. Today, a majority of the enclosure heads are of the Ildareto age set, many of
whom grew up together in the area as ilaiyok and ilmurran. The other enclosure heads are comprised
of the younger Ilkimnya age set and are mostly from newer families in the area. As many of the older
heads of enclosures have known each other for many years, often living quite close to each other,
there is a greater degree of trust and cooperation among these combined families than might be the
case elsewhere in the Idodi rangelands.
262 During field work, one enclosure head died and one other moved away to Ruaha Mbuyuni about 150kms to
the east.
210
Table 7.2: Ilparakuyo enclosures in Mahuninga Ward in 2000-2002
Enclosure
Division Enclosure Head Clan Age Seta
Cluster
Ismaeli Katei Inkelingishu Ildareto - Errokorr
Suberi Katei Inkelingishu Ildareto - Errokorr

Odo Mong’i
1
(Red bullock)
Rashidi Katei Inkelingishu Ildareto - Errokorr
Sendeu Katei Inkelingishu Ilkimnya - Errokorr
Salum Lebere Ilmarumai Ildareto - Ikelimboti

Ola Kiteng
2
(Black bullock)
Mbungjai Mshoka Ilmarumai Ilmedoti - Errokorr
3 Ola Kiteng Ndagoala Sumaeli Ilorborkishu Ildareto - Errokorr
4 Odo Mong’i Ismelwa Palinoo Inkelingishu Ilkimnya - Errokorr
Kombeti Palinoo Imosiyekuu Ildareto - Errokorr
5 Odo Mong’i Rashidi Palinoo Imosiyekuu Ilkimnya - Errokorr
Kifutu Palinoo Imosiyekuu Ilkimnya - Errokorr
Kanaiya Mataali Ilmarumai Ildareto - Ikelimboti
6 Ola Kiteng Salimu Lebere Ilmarumai Ildareto - Ilpalingotwa
Kangai Lebere Ilmarumai Ildareto - Errokorr
7 No data Rasta Sejengo No data Ilkimnya - Errokorr
8 Ola Kiteng Sangwea Makaluma Ilatayo Ilkimnya - Errokorr
a
See the appendices for further information on the opening and closing of age sets in Idodi.
In recent years, as the farming population has grown and the area under farmland has been
expanded, the Ilparakuyo263 say that they have found themselves increasingly surrounded by fields,
which have become a source of disagreement. Both herders and farmers identify a particular number
of years during the 1990s when tensions between farmers and herders were running particularly
high. Disagreement centred on dry-season grazing with the Ilparakuyo tending not to always seek the
permission of farmers to graze their field crop residue, despite their claims to the contrary. There
were also incidents of cattle being grazed on un-harvested crops that may or may not have been
accidental. While Ilparakuyo pastoralists were often successfully prosecuted by the village
government and farmer plaintiffs, paying fees and compensatory fines, these incursions led to
increasingly high tensions both in regard to farmers (particularly in Isanga) angry at the repeated
incursions by cattle into their fields without their prior consent, and also Ilparakuyo, who felt that their
grazing area was constantly being encroached upon by expanding fields. The Ilparakuyo claimed
263 The Ilparakuyo are referred to here as a single unified entity with a single set of narratives. This is of course
not the case, for either Ilparakuyo society or for their many and diverse narratives. However, in the context of
the case study, and in regard to the collective action taken on the part of the Ilparakuyo families in Mahove, the
treatment of the Ilparakuyo as a collective whole can be seen as being justifiable here.
211
that as long term inhabitants of the area, they had a right to have their own grazing land. The
tensions were underpinned by farmer narratives of destructive pastoralists wantonly invading fields
and destroying soil fertility (see Chapters Three and Five). Pastoralists claimed farmers wanted them
to leave the area, that they were de facto second-class citizens with no land rights, and that they
were given little opportunity to participate in village government to represent their interests (see
Chapter Four). The Ilparakuyo also claimed that the recently arrived Barabaig were making farmer -
herder relations worse. The Ilparakuyo alleged that the Barabaig were committing many more field
incursion offences than the Ilparakuyo, but as herders, the Ilparakuyo were being held equally
responsible by farmers for the incursions committed by the Barabaig. Irrespective of the veracity of
these allegations made by the Ilparakuyo about the Barabaig, again the underlying reason for these
narratives on the part of the Ilparakuyo can be explained by a sense of range insecurity and a
shortage of dry season grazing compounded by newly arriving Barabaig pastoralists. Farmers’
narratives of herders as being a destructive and an unwanted presence in the landscape were also
often expressed simultaneously with narratives of fertility decline, a shortage of irrigable and arable
land, adverse changes in the climate and a general decline in their overall fortunes (see Chapter
Five).
The disagreement about land-use practices had taken on its own inertia as accusations and counter-
accusations bounced back and forth (see Nader and Todd 1978, 15 cited in Hagberg 1998, 68)
among and between herders and farmers, often with acrimony and little attempt to engage in a
process of negotiation about how the landscape could and should be used. Although these accusations
and counter-accusations continue to occur, of late they have become less acrimonious and less
meaningful than previously. An important reason for this is that the Ilparakuyo, through collective
action, sought to reduce tensions by coming to closer agreement with the village governments of
Makifu and Mahuninga over how to modulate farmer - herder relations.
The Ilparakuyo first began to negotiate with Mahuninga Ward over the status of Mahove bushlands in
1994264 during a period in which farmer-herder relations were becoming increasingly tense. A
meeting was called by the village governments of Makifu and Mahuninga to which both farmers and
herders were summoned. A farmer-herder committee was subsequently set up to investigate and
agree the boundaries of the new pastoralist grazing area which they agreed should be created. It
was agreed that Mahove would be reserved for pastoralists in an effort to reduce conflict and that
the minutes of the meeting at which this had been agreed would serve as a record of this. The
boundaries of Mahove were described definitively in some areas, especially those close to farmers’
fields, but in other areas where conflict had not occurred, less so. In order to create an indisputable
boundary in a zone where the grazing area and farmer’s fields bordered each other, it was agreed
that a tributary of the Tungamalenga River would form the southern part of the border. In return for
the allocation of the pastoralist area, no further grazing of livestock was to occur in the wetland fields
264 This information is pieced together from minutes of meetings retained in Idodi’s divisional office, as well as
from interviews with Ilparakuyo pastoralists, some village government officials and farmers.
212
and those apprehended would be liable for a penalty of TShs 50,000 (about US$100 in 1994 - a
significant sum). In addition all dry-season grazing on dryland fields was to stop so as to prevent
further herder-farmer conflict. All pastoralists were to move into this area forthwith.
Leaving a rights-based dispute unresolved - Mahuninga village

Unfortunately, the southern border of the Mahove grazing area as had been agreed included some
farmer’s fields on the northern river bank. Although it was agreed that these farmers would be
allocated fields elsewhere, this never happened, perhaps since it would have been politically difficult
for the village government to be seen evicting farmers for the benefit of pastoralists. Evicting farmers
from land that they could lay rightful claim to (given that they had cleared the land they were
farming and were part of the autochthonous community) would have been controversial among the
wider Hehe farming community and damaged the political credibility of village and ward leaders. In
short, many Hehe farmers would have viewed the evictions as illegitimate, particularly in the context
of previous state-perpetrated evictions in the wider landscape. The continued presence of farmers
farming their fields, and even re-opening some older fields was a continued source of irritation for
the Ilparakuyo, who saw the continued presence of farming in their reserved grazing area as
undermining their land-use rights as had been agreed. They viewed the continued presence of some
farmers in the village commonage assigned to them as being reflective of their inferior collective
status in relation to the farming majority. An effort was made by Makifu and Mahuninga village
governments in 1998 to prevent further encroachment of farmers into the southern section of Mahove,
and despite some forceful language employed in the letters that were written by the Idodi Katibu
Tarafa (Swahili: Divisional Secretary), and which were copied to the District Commissioner, the
Mahuninga village government has not stopped farmers from continuing to farm in parts of what are
designated the Mahove grazing area reserved for herders.
It is clear that there was an overt dispute between Ilparakuyo pastoralists and farmers over the use of
the Mahove rangelands. Although the Ilparakuyo had lobbied the local government at different levels
to make the farmers move elsewhere, the farmers were viewed by the wider farming community -
and thus the village government - as having sufficient legitimacy to continue their use of the Mahove
rangelands. The Ilparakuyo could have attempted to take their dispute to court, but it is likely that the
adjudication would have remained unimplemented, and their social relations with the wider farming
community would have likely suffered. The Ilparakuyo had considered obtaining title to the land, but
their efforts at village level had been frustrated by the village government, which was reluctant to
give formal land tenure (as opposed to occupancy) rights to the Ilparakuyo in the village commonage.
An underlying reason for this was that the wider Hehe farming community did not view the Ilparakuyo
as legitimate owners of the land, merely its occupants. While pursuing a rights-based approach to its
conclusion might have been the most obvious choice for an observer or external actor, such an
approach would have led to heightened tensions between all parties and to deteriorating social
relations.
213
Resolving disagreement through negotiation - Makifu village
Despite the framework land-use agreement which designated the Mahove area in 1994 as a
pastoralist land-use zone, disputes over dry-season field grazing had continued to occur - particularly
in parts of Makifu village (see above). In 2000, the Ilparakuyo and village agreed that Mahove
would be more strictly reserved for pastoralist grazing (i.e. no more dryland fields cleared), although
other village uses such as firewood collection would still be allowed to continue. Significantly, no
further transhumant pastoralists would be allowed into the area. The underlying significance was that
Barabaig families currently living in the west of the village, who had occasionally encamped in the
Mahove area, would be prevented from moving again into the Mahove rangeland. In return, the
Ilparakuyo agreed that they would not graze their livestock in some parts of their village -
particularly a relatively extensive area of fields farmed by Bena agro-pastoralists who wanted to
reserve the fields for their own livestock grazing during the dry season. The most important outcome
for the Ilparakuyo as they saw the unfolding situation was that, with Barabaig herders less present in
the immediate landscape, they would be able to better control and modulate relations with farmers.
A further advantage was that there would be less competition for dry season grazing - both within
the Mahove rangeland as well as in farmer’s fields.
Significantly, during the same period the Ilparakuyo also acquired de facto recognition from the
village government that they could act as a sub-village and would have their own chairman. This
development was a pragmatic move on the part of Mahuninga and Makifu villages and the
Ilparakuyo. In the past the creation of a separate administrative unit on the basis of ethnicity would
have been arguably controversial since the State has nationally heavily suppressed political
expression of ethnic identity. However, although the kitongoji in Mahuninga Ward was informally
created on the basis of ethnic identity, its justification was seen by those involved to lie in improving
land-use and administrative relations between different categories of land-user. As discussed
previously, national policies support land-use planning and the delineation of land zones according to
herding and farming categories - a far less politically charged approach to dealing with an issue that
may often include but avoid expression of an underlying dimension of ethnicity. In addition to being
recognised by the local village governments as running their own internal affairs, it was agreed in
return that the Ilparakuyo would cooperate in collecting their own taxes (the district development levy
and livestock head tax) thereby helping the village governments more easily accomplish an often
tense and unpopular tax collection process.
The process of collective action on the part of the Ilparakuyo pastoralists in Mahove can be seen as
having tangibly improved farmer-herder relations and having reduced conflict through the re-
negotiation and reaffirmation of boundaries (which nevertheless remain porous). While the land-use
outcome in Mahuninga ward has been agreed and framed in terms of a rights-based land-use
planning approach, critically the agreement has been negotiated as a result of evolving relations
between herders and farmers. It is also clear that the land-use agreement in itself does not presently
guarantee herders formal tenurial rights over the Mahove rangelands. Instead, herders continue to
build legitimacy through their growing and evolving relations with farmers and local government. It
may be the case in the future that herders will be able to secure formal tenurial rights over the
214
Mahove rangelands as a result of their continued investment in social relations with farmers - a
development which on its own and in the absence of this investment in social relations, would not be
legitimate in the view of many farmers and thus the village government.
Networks: individual herder-farmer relations and land-use outcomes

The land-use (rights-based) agreement in Mahuninga Ward may be viewed as a public expression of
how people perceive the landscape should be normatively used. Yet the agreements negotiated at a
collective level between herder and farmer can be viewed as accounting for only part of the land-
use outcomes that have developed in Mahuninga Ward. In reality, the rights-based land-use
agreement may be less definite and strictly adhered to than people imply. An underlying reason is
that the socio-economic and production relations of herding and farming households cut across
boundaries. A useful way of demonstrating this is through examining the impact and importance of
networks in re-defining land-use outcomes.
Thus the following example demonstrates how networks can cut across rights-based rules and
boundaries that have been agreed at a collective level and which form the socio-political backdrop
within which people use the landscape.
Ismaeli Katei lives on the northern edge of the Mahove bushlands, having initially lived with his late
father and brother in this area from the late 1970s onwards. In the late 1980s after having married
his first wife Nailole, they established their own homestead, and have since lived in close proximity to
Ismaeli’s three brothers. Ismaeli together with his four wives are now perhaps the most successful
pastoralist family of his age-set (he is a junior elder) in the Ikwavila valley (see Box 6.3).
In order to obtain sufficient grazing and nutritional security for their herd through the year, Ismaeli
seasonally secures access to different range areas, mostly through the networks that he has built up in
different localities. During the late wet season, much of the family herd is taken to Mlowa where the
herd is lodged with relatives (his fourth wife’s father) to avoid the seasonally elevated level of tick-
borne disease that the livestock are exposed to in Mahove. When the herd returns to Mahove in the
early dry season, it increasingly relies on field grazing as the range dries out and is grazed down.
Ismaeli has developed two sets of networks in Mahove that enable him to secure sufficient field
grazing: The first network focuses on Hehe farmers in Mkanisoka sub-village with whom he maintains
a client-patron relationship, strategically donating small stock for village and other social events. He
often frequents the kilabu and buys beer for his associates, particularly prior to the onset of the dry-
season field grazing period. During this process, he also agrees prices for the grazing rights for the
crop-residue of farmer’s fields in Mkanisoka. As his patronage has grown among farmers in recent
years, so he has found it necessary to build a satellite enclosure closer to the fields which are
seasonally used by the family to reduce the amount of trekking his stock have to do between his
enclosure and the field grazing area. The investment Ismaeli has made in relations with Hehe farmers
has meant that he has gained increasingly secure access to an important dry-season grazing resource
for his herd, despite a continuing policy by the district that dry season field grazing is not permissible.
215
Although access to farmers’ fields in Mkanisoka is an important part of Ismaeli’s annual grazing
strategy, the family has also managed to acquire access to fields in another sub-village that they are
supposed not to be able to graze as they are owned by agro-pastoralist (Bena) farmers who reserve
them for their livestock. Ismaeli gains access to some of these fields close by their enclosure through
loaning stock to a poor yet skilled Bena stock associate who then grazes his stock on these fields
through agreement. The family gains further access to this area of fields through occasionally renting
one or more of these fields from Makambalala agro-pastoralists farmers to farm at the beginning of
the wet season, a tenancy which also engenders the right for his livestock to graze the crop residue of
these fields after harvest. In addition, by lending his plough and training the bullocks of the Bena
agro-pastoralist farmers at other times of the year, he maintains a sufficiently strong network of
relations with these farmers for them to ignore his flexible interpretations and circumnavigations of
the boundaries and rules that have been agreed at a collective level.
Groups and networks: ‘on-stage’ and ‘off-stage’ discourses

The process of developing land-use agreements between collective herder and farmer groups as
described in the previous section can be seen as part of an ‘onstage’ discourse (Feierman, 1990). In
contrast to this ‘onstage’ discourse, there has also been a parallel ‘offstage’ process, sometimes
contradicting the ‘on-stage’ discourse. Thus Ismaeli Katei was able to use his personal - offstage -
networks to gain access to disallowed grazing resources. In this regard, the offstage reinterpretation
of a rules-based framework was beneficial for some (Ismaeli and his associates), and likely to have
been benign for others (other herders and farmers).
However, offstage practices can cut both ways and the offstage renegotiation of a rules-based
framework may sometimes be harmful. For example, in Mahuninga village, pastoralists were
summoned to a village meeting convened to address some recent incursions of livestock onto farmers’
fields. The Ilparakuyo were accused of carelessly herding their stock through the fields, an accusation
that they vociferously denied. Instead, they argued that it was some immigrant Barabaig pastoralists
(in contradistinction to other Barabaig who are more permanently resident and registered as village
members) who had recently arrived and had been allowed to temporarily stay in the village. The
Ilparakuyo asked why these Barabaig had been allowed temporary access to the village when
transhumant pastoralism (in contradistinction to seasonal rangeland mobility by locally resident
pastoralists which retains a degree of local legitimacy) had been banned in the district. The village
government was unable to answer this question. The Ilparakuyo alleged, and farmers have likewise
accepted at other candid moments, that key village government officials often allowed transhumant
pastoralists access to village rangeland in order to derive illicit cash payments from the migrant
pastoralists. The Ilparakuyo saw the illicit and ‘off-stage’ allocation of their rangelands by members
of the village government, together with the perceived elevated level of farmer-herder conflict, as a
significant reason why they would continue to be unable to achieve sufficient tenure security of their
rangelands, and why herder-farmer tensions would continue to occur. Village government officials
were re-interpreting - or to the Ilparakuyo, breaking - the agreed land-use framework that had been
216
negotiated, but with little regard as to the implications of their actions for the maintenance of good
social relations between herders and farmers.
While this was an understandable viewpoint for the Ilparakuyo, this perspective can be contrasted
with that of transhumant Barabaig herders’. The Barabaig suffer most from rangeland insecurity since
they often do not have as developed a set of social networks with farmers or with resident herders,
yet they employ opportunistic herding strategies with considerable skill and effectiveness, and
personal networks with particular village government officials can be key for securing rangeland-use.
Finally, in relation to contemporary developments in Tanzanian local government reform (see

Chapters Two and Three), this example also demonstrates how continuing issues of accountability and
poor governance at village level may sometimes adversely impact on herder and farmer land-use
agreements and relations. This theme is taken up further in the next section.
Summary
Growing social and economic relations have created a sufficient basis and commonality of interest
between herder and farmer within different parts of Mahuninga ward (not least helped by a
modulating [Bena] agro-pastoralist presence), to address long-standing disagreements on how the
landscape should be used. Group negotiations have created a simple set of rights-based rules and
boundaries as a means for regulating landscape use by different categories of land-user and as a
public expression of how farmers and herders should use the land. These simple ‘onstage’ rights-
based rules and boundaries have remained flexible and porous as they are re-interpreted and re-
negotiated ‘offstage’ within a complex network of social affiliations between herder and farmer. Thus
it is the dynamic interplay between more formal group agreements and their informal play-out in
individual social networks that has led to the particular land-use outcomes that we see today in
Mahuninga Ward. Growing social relations have been the catalyst for the creation of formal rights-
based land-use agreements, and they have remained as important in modulating how these
agreements are played out in reality. Finally, growing social relations between farmer and herder
may be viewed as serving to increase the legitimacy of access to land and the rights claims of
otherwise marginalised herding groups, who are viewed as ‘strangers’ by an autochthonous farming
group.
7.3.3 Conflict and the undermining of social affiliations

In direct contrast to the relative success of negotiated agreements in Mahuninga and Makifu villages
between herder and farmer, there has been a protracted and on-going conflict between farmers and
herders in Idodi village over the last six or more years. The drawn-out nature of the conflict has
resulted in the polarisation of different parts of the community along lines of ethnic affiliation. Much
of the conflict has centred on access to farmers’ wetland fields by Barabaig pastoralists for dry
season grazing, and the use of fire in the landscape by farmers to control pastoralist land-use. The
case of Idodi village presents a useful opportunity for examining why, in contrast to Makifu and
Mahuninga villages, farmer - herder conflict has continued despite the intervention of the state.
Collective groups, although ostensibly agreeing to resolve conflicts of understanding in public (on
217
stage) through a rights-based land-use agreement, continue off-stage to resist and evade imposed
solutions which they perceive as being, at least in part, contrary to their underlying interests. In Idodi,
the state has pressured farmers and herders into creating a land-use agreement, almost in an
adjudicatory fashion. Thus far, the land-use agreement has been still-borne and the process has
unravelled as it has been characterised by a lack of extended negotiation between the conflicting
parties, in large part due to the poor social relations that exist between some (not all) herders and
farmers.
The context to conflict: resource use pressure

Idodi village has an extensive area of wetland agriculture fed by reliable streams flowing off the
well wooded highland scarps immediately to the south of the main village settlement. Idodi’s wetland
farmlands are not only a centre of agricultural productivity which has been exploited over the years
by an increasing number of farmers, but they also are an invaluable dry season grazing area for
herders. In particular, Barabaig herders have been drawn to graze their livestock in these wetland
fields at the onset of the dry season. Although there are a number of Barabaig families who have
been resident in Idodi village since 1984 (see Chapter Six), there are also a relatively substantial,
although unknown, number of Barabaig pastoralists who opportunistically use the wetland farmlands
during the dry season for grazing, before moving away again, often north to Pawaga. As the
numbers of both farmer and herder using the wetland farmlands have increased, so the likelihood
and frequency of conflict has grown. Conflict between Barabaig herders and farmers in the wetland
farmlands has been made all the more likely due to different and conflicting landscape use practices
and separate jural systems (see Chapters Three and Five) compounded by poor social relations.
Explaining Barabaig - farmer dispute

An important question arises as to why many of the Barabaig - in contradistinction to the Ilparakuyo -
have often not avoided dispute or violent conflict with farmers and, indeed, why they have not sought
to more proactively attempt conflict resolution, particularly as a collective group. Any explanation
attempted here is necessarily likely to be incomplete, especially since only a limited amount of
interaction with the Barabaig was possible during field work. However, an attempt at an explanation
is necessary in order to better develop the context and case analysis of land-use conflict in Idodi.
Before continuing, it is also important to underline the fact, in contradistinction to Barabaig, that some
Barabaig maintain stronger social and economic relations with farmers.
The Barabaig have a saying, ‘cattle graze themselves’ Lane (1996: 45) 265 and it is likely that young
Barabaig men have not been too particular as to where they allow their cattle to graze within a
locality - a tendency borne out by numerous complaints from farmers. Moreover, when Barabaig
cattle have damaged farmer’s fields and crops, the farmers have had little or no recourse to
Barabaig jural institutions which would be the most effective means of sanctioning the young men
responsible. Barabaig elders - who are often heads of households and overall responsible for making
265See also Chapter Six, Section 6.2.3.
218
stock grazing decisions - have perhaps seen little cause to control or sanction their herdsmen as to
where and how they graze their livestock in relation to farmers’ fields and standing crops. This may
particularly be the case since no Barabaig in Idodi were known to farm wetland fields (with only a
minority farming dryland fields). Even if Barabaig elders were concerned with controlling the grazing
practices of their herders in and around farmers’ fields, it is uncertain whether Barabaig jural
institutions function to the extent they might be expected to. Customarily, Barabaig land relations are
mediated at the clan level (Klima 1970, 84; Lane 1996, 142; see Chapter Six). However, in Idodi,
there are no clan lands (since the Barabaig are not autochthonous), and this institution cannot be used
for resolving land-based disputes. Another institution that might be more appropriate for modulating
land-use practices by Barabaig in Idodi is the ‘council of neighbours’ (Lane 1996, 143; see Chapter
Six). This institution plays an important role in mediating matters of public concern, and may come to
the fore in enabling collective decision-making in Idodi by the Barabaig. However, decisions made by
the ‘council of neighbours’ may be little known and remain unobserved by more itinerant Barabaig
herders who seasonally visit the Idodi rangelands in search of grazing. Moreover, the ability of
farmers who are not Barabaig to effectively participate and bring their grievances before a ‘council
of neighbours’ would be minimal.
Farmers have therefore necessarily only had recourse to the village government and the government
jural system (see Chapter Four) to attempt resolution of their disputes with Barabaig herders. In order
for a farmer to successfully pursue a dispute, and other factors not withstanding, it has been the norm
that the farmer has had to indisputably identify the herd owner and provide evidence of field
damage. Where the livestock owner is not immediately known, then the livestock have had to be
rounded up by the farmer and taken to the village office as evidence. There have been incidents
where groups of Barabaig men have intercepted their cattle being driven to the village office, and
then made off with them, leaving the farmer without the evidence with which to pursue his/her
dispute. The Barabaig in turn have claimed that farmers have sometimes purposefully driven cattle
into their fields in order to try and be awarded large compensatory sums by the village and ward
courts.
As with the villages of Makifu and Mahuninga, there have been allegations of village government
officials receiving illicit payments from Barabaig families wanting to use the Idodi rangelands and
farmlands for their dry season grazing. It is further alleged by farmers and Ilparakuyo herders that
village government officials have also regularly taken payments from some Barabaig herders
apprehended by farmers and under prosecution by the village and ward courts. The Barabaig
herders are thus able to avoid the larger payments that they would have to make in terms of
compensation to farmers, fines and court fees (the latter two items often equal the compensation due
to the farmer[s]). Indeed, it was well known during 2000/2001 in Idodi village that one Barabaig
herder boasted that he could graze his stock in any field he wished as he would be able to out-bribe
any farmer and that he had a particular village official under his patronage. Poor governance,
particularly within Idodi village government, was implicitly recognised by the District Commissioner,
when in October 2001 she ordered that all herder - farmer disputes were no longer to be heard by
219
the village and ward courts and should instead be heard by the primary (but see Chapter Four) and
higher courts.
Thus overtime, repeated field incursions by Barabaig herders and the difficulty experienced by
farmers in successfully pursuing their cases through the village and ward courts led to a build-up of
frustration by farmers. There were several reasons for this frustration. Firstly, the adjudication of
dispute in village and ward tribunals was not always impartial (due to back-deals struck between
herder defendants and tribunal officials). Secondly, farmers did not always feel that they had been
adequately compensated. However, thirdly and most importantly, the adjudication of individual
disputes had not addressed the continuing - and unresolved - problem of field incursions and wider
land-use conflict between Barabaig herders and farmers. This had led to a situation in which even
relatively minor disputes began to rapidly escalate towards violent conflict - an articulation of the
wider unresolved and underlying land-use conflict and of farmers’ frustration and anger at being
unable to stop the field incursions. On a number of occasions the Field Force Unit (paramilitary police)
had to be dispatched from Iringa (the district headquarters) to prevent any further violence and to
make arrests. Violent conflict broke out during field work in 2001, and also according to divisional
records, in 1998.
Crisis and beginning the negotiation of landscape use

In 2001, the early dry season months had seen multiple herder - farmer disputes over field grazing,
several of which had led to flair-ups in violent conflict. Herder - farmer tensions were brought to a
head in late September when a dispute over field grazing between Barabaig and Hehe farmers
escalated into violent confrontation. Cattle were lacerated and half a field of crops were up-rooted
in retaliatory reactions after the first confrontations. The village then dispatched a messenger to
Iringa (the district headquarters) to call the police. The messenger allegedly claimed that 200
Barabaig herders were about to ‘invade’ Idodi and graze their livestock in farmers’ fields. A Field
Force Unit (FFU) was promptly dispatched to investigate and arrest both the herders and farmers
involved in the dispute. The presence of the FFU suppressed any further outbreaks of violence, but
tensions between herder and farmer remained high.
A previous meeting in June, that had been initiated by the Divisional Secretary at the behest of the
District Commissioner266, brought farmers and herders (both Barabaig and Ilparakuyo) together to
discuss land-use conflict and to begin a land-use planning process. During the meeting, the
participating farmers and herders agreed that the village should be divided into farming and
herding areas so as to help reduce land-use conflict through providing pastoralists with their own
266 The District Commissioner for Iringa, Mrs Tatu Ruta, was newly appointed in 2001 and made a point of
trying to ensure that farmer - herder conflicts were resolved as equitably as possible in the district. In an
introductory meeting to Idodi Division in July 2001, she firmly warned the meeting that, ‘Certain ethnic groups
could not and would not be discriminated against – this was Tanzania! Why should a whole community suffer at
the behest of a few errant individuals? Instead firm steps would be taken against those individuals who erred
and not communities’.
220
rangeland areas, in which no farmers would be allocated fields. Each particular pastoralist area
within the village would have to elect its own kitongoji chairman who would be responsible for his
hamlet and its relations with their farming neighbours. It was agreed that the grazing of livestock in
farmers’ fields could not occur without the consent of their farmers. Furthermore, it was agreed that
newly arriving pastoralists should inform the village of their presence and that pastoralists wishing to
seasonally move between rangeland areas (particularly between villages) should first attain the
permission of the relevant village authorities. The following day this overall agreement was put to the
wider village assembly in a public meeting at which only five pastoralists were present (pers obs.) out
of about 150 people attending267. The meeting was remarkable for the degree of active
participation by the village assembly who are often reticent in such meetings to air their views
publicly. The contributions to the meeting as they were made by the members of the village assembly
are presented in Box 7.1. These comments are generally representative of the range of opinions and
attitudes held by farmers in regard to pastoralists. The opinions in particular show the open degree
of hostility felt by farmers towards the Barabaig, and some of the underlying conflicts of interest over
future range use.
After the confrontations of late September, in mid-October the District Security Committee268
travelled to Idodi to meet and discuss the recent herder-farmer conflict and to assert their authority
over providing a solution and end to the conflict. During the meeting, different parts of the village
were identified and exclusively allocated for herding areas and farming areas with an order that
herders move immediately to herder areas and farmers to farming areas. It was further declared
that field grazing was to be banned forthwith despite some farmers expressing their dissent - saying
that field grazing was a useful source of income. All further registration of immigrant pastoralists
(mostly Barabaig) was to be stopped not least because of herder networks leading to pastoralists
inviting in their kin and also lodging their kin’s stock in their own herds. Livestock tax records would be
henceforth monitored to ensure no extraordinary increases in numbers of livestock occurred. The issue
of maintaining herders’ long established seasonal range-use patterns within the Idodi rangelands was
not mentioned or discussed. Since Idodi’s grazing range was identified as being ‘full’, de-stocking was
put forward as a means to coping with perceived (on the part of the security committee) range
depletion problems. The Idodi village government supervised by the Divisional Secretary was to carry
out and implement the provisions as soon as possible. The measures put forward by the District
Security Committee - zoning, sedentarisation and destocking - directly reflected the more widely held
vision and approach of government administrators for managing herder-farmer relations and
herders’ livelihood practices.
267 The village government interestingly mentioned ‘453’ people as having attended the meeting in their report
to the District Commissioner.
268 Including the District Commissioner, the Police Officer Commanding the District (OCD), the District Chairman
of Prisons, the District Chama Cha Mapinduzi (CCM) party chairman as well as the District Agriculture and
Livestock Officer (DALO).
221
Box 7.1: Farmer opinions about pastoralists and herder-farmer land-use conflict
· I say turf the Barabaig out of the village completely!
· Rice paddy crop residue grazing in particular should be banned as it destroys the irrigation system!
· A team of villagers should be created which should establish the exact numbers and whereabouts of
the Barabaig!
· All herders should move from higher up in the drainage basin to lower down to protect water sources
and river flows.
· Even crop residue grazing on dryland fields should be banned as it is not compatible with perennial
cashew nut crops! (the individual was a cashew nut farmer)
· Ihehero (a rangeland area designated as a grazing zone and which has had a long association with
pastoralists - especially Ilparakuyo) should be evacuated by herders as it is too close to cultivation!
· The Barabaig should leave because of their tendency to aggression - they are a nuisance and instead
the Maasai should remain - at least they are no problem!
· The Barabaig should leave! The Maasai can stay as they are generally beneficial to the community
providing meat and milk.
· There’s no point discussing all this if the law doesn’t help or isn’t implemented - especially when those
responsible for doing so - the village leaders - disregard the law!
· The pastoralists should leave - even though we have got used to them, because they are always
inviting in their relatives to join them and then they take what is ours!
· I say all pastoralists out!
· Why are the problem pastoralists allowed to stay when only the good ones should be allowed to?
· If the law is broken, what steps will be taken?
· When Barabaig are caught grazing illicitly in the fields, they are often very aggressive - how should
these situations be handled?
· Crop residue grazing should be stopped as it conflicts with crop residue incorporation! (when preparing
land for planting - a fertility measure)
· There’s enough grazing for 10,000 head of cattle in Kibikimuno (a rangeland area to the north east of
the village settlement extending deep in the LMGCA) - so let them graze there! We border the park
anyway and there’s a lot of grazing available! (this is an overstatement)
· What about the future and our children and their children? Surely we farmers need expansion areas?
· In the event of a farmer being beaten up (by herders) and the perpetrator running into the bush
(away), what should we do? It seems to me as though the village militia and the MBOMIPA scouts are
often reluctant to follow the matter up!
· I agree that Ihehero was undoubtedly a livestock area, but we farmers have increased in numbers and
need to expand our cultivation areas. I recommend that the herders de-stock in order that they have
sufficient grazing (and farmers sufficient farmland).
· I’m against Barabaig sheltering each other! Often when one of them has done something wrong, his
friends or relatives hide them and then represent the accused on their behalf. I think this should be
discouraged!
· I think we should pay the village militia to ensure security in the fields!
· Who’s going to help anyone when the accused is fined 40,000/- (approximately USD $40) and those
who helped to round up the cattle and bring in the accused get nothing? People just won’t help if this is
the case!
· Why aren’t there any Barabaig here? (Unfortunately, a young Barabaig child had gone missing the day
before and many of the Barabaig were out searching – their absence was seen as having ulterior
significance)
222
After the meeting no further action was taken by the village government - as had happened before
after similar meetings held in June. Farmer-herder conflict broke out again two weeks later when
fifteen farmers sent complaints to the village government about herder incursions into their fields. An
emergency meeting was convened to discuss the new round of herder-farmer dispute and to reflect
on the continued lack of action to carry out the orders of the District Commissioner and District
Security Committee. This further outburst of conflict (although not violent, enough to make the village
and ward governments apprehensive about the reaction of the District Commissioner should she learn
of their inaction) was sufficient to provoke a hurried demarcation in the following days of Idodi
village area as had been ordered. Members of the village and ward committees together with the
farmer-herder committee and the Divisional Secretary chose trees and other physical landmarks as
land-use zone boundaries which were recorded in a set of meeting minutes. Particular farmers living
in the newly exclusive pastoralist areas were identified for eviction and vice versa for herders.
Two months later, when I left the field, farmers and herders had still not moved into their land-use
zones, and with the onset of the farming season could no longer do so effectively for another 5-6
months. Indeed, despite the delineation of land-use zones, it is understood that farmer-herder conflict
broke out again during 2002 resulting in the eventual transfer of the Divisional Secretary to another
part of the District on the orders of the District Commissioner, due to his failure to ensure farmer and
herder compliance with the now irate District Commissioner’s orders. Most difficult of these, was
preventing the continued dry season field grazing by pastoralists, as many farmers were reluctant to
miss out on the opportunity of generating further income from herders grazing their fields.
The case of the Ilparakuyo: undermined social affiliations and landscape occupancy
The protracted conflict between Barabaig and farmers had already had an adverse impact on the
Ilparakuyo who had built up strong labour and exchange relations with farmers over the 50 years
that they had been in Idodi. In this time, the Ilparakuyo have developed conflict aversion practices -
for example, purposefully avoiding grazing their herds in wetland fields, and have generally sought
to resolve disagreement quickly269. Yet their social relations and status within the village, as
perceived by farmers, as a benign and sometimes beneficial presence (see Box 7.1), has been
increasingly overshadowed by the acrimony associated with the Barabaig. Previously, the Ilparakuyo
had been able to continue landscape use practices - such as placing their enclosures upstream of the
village settlement and grazing their stock in dryland fields - that had been technically proscribed one
way or the other but which the village government had continued to allow and farmers to tolerate or
even co-operate with. As relations between categories of landscape users - farmers and herders -
have continued to deteriorate, these practices have now increasingly come into question.
269 Although there are one or two cases in which particular Ilparakuyo herders have evaded paying
compensation to farmers through delay tactics, eventually leading to the farmer giving up after repeated court
hearings.
223
The elevated and prolonged levels of conflict in Idodi have resulted in several Ilparakuyo families
being forced to give up their positions in the landscape and to move to areas allocated to them as
part of the land-use planning process but not of their choice. Salum, a senior Ilparakuyo elder,
explained with reserved anger and frustration how he saw these imposed changes to their landscape
occupancy as being particularly symbolic of the Ilparakuyo’s marginalisation and their feeling of
second class citizenship as a collective group within the village. He pondered how it was that he and
his family, although having arrived in Idodi 50 years previously and before the great majority of the
farmers now living Idodi, were regarded as outsiders. Whereas previously the Ilparakuyo had
managed to retain access to rangeland areas they had long used by continuing to live in these
localities despite farmland expansion, he blamed the Barabaig for having polarised and damaged
herder - farmer relations to the extent that the previously porous boundaries and flexible rules
governing the use of the landscape were becoming increasingly - and unnecessarily in his view -
enforced. Although the senior elder’s personal network and relations remained strong with individual
farmers, he worried about the nutritional security of his and his associates’ herds and their personal
networks with farmers related to field grazing agreements. If during the dry season he was not to be
allowed to graze his stock on dryland fields, where could they be grazed?
Conclusions
In contrast to Makifu and Mahuninga where categories of land-users - herders and farmers - have
reached a growing understanding of how the landscape is to be occupied and used through formal
‘on-stage’ agreements played out in an informal ‘off-stage’ matrix of networks, in Idodi, the situation
has been markedly different. Here, relations between groups - Hehe farmers and Barabaig herders
have become increasingly polarised due to the absence of appropriate or sufficiently functioning
dispute resolution forums, in part due to nascent collective group relations between a number of
Barabaig herders and farmers, undermined by poor village governance. The Ilparakuyo, out
numbered by the Barabaig, have been unable to negotiate or otherwise sufficiently influence the
deteriorating relationship between the Barabaig and farmers - which has overshadowed their well
established presence in the landscape and their strong relations with farmers. Despite heavily
sponsored attempts by the state to coerce agreements between categories of resource users over
land-use and the demarcation of the landscape, compliance has been poor and conflict has continued.
This situation can be analysed from three different yet complementary perspectives.
Firstly, despite the polarisation of relations and rhetoric, herders and farmers have varyingly resisted
the onstage demarcation of the landscape, and particularly the enforcement of rules that cut across
off-stage and growing herder - farmer socio-economic relationships including, ironically, consensual
field-grazing. In short, rights-based landscape demarcation and the attempted strict implementation
of exclusive land-use areas has thus far failed since it is not consistent with underlying and growing
herder - farmer relations and networks.
Secondly, a higher level of concord between herder and farmer over acceptable land-use practices
is realisable in Idodi. Indeed, despite the worsening conflict situation over land-use practices between
Barabaig herders and farmers, the Ilparakuyo had largely achieved a fluid agreement over land-use
224
mediated through their social relations with farmers and the village government. The level of dispute
in Idodi can be viewed as having escalated to violent conflict because the underlying conflicts of
interest over land-use between Barabaig herders and farmers had not been addressed and resolved.
The village and ward courts had only served to adjudicate disputes, not resolve the wider and
underlying land-use conflict issue. In short, there was an absence of sufficiently effective dispute
resolution mechanisms in Idodi, particularly between Barabaig herders and farmers. Although a
combined and potentially helpful farmer and herder conflict resolution committee had been set-up by
the local government as a result of the growing levels of tension, it is not clear that the creation of the
committee by itself was necessarily a sufficient approach to addressing the conflict. Firstly, artificially
created institutions take time to evolve (if they do at all) towards socially-embedded and effectively
functioning structures that are seen to be sufficiently legitimate by the communities that they serve and
represent. Conflict resolution mechanisms cannot materialise instantaneously, but may be seen as
evolving over time, through social process. This is particularly the case when highly indignant disputing
parties are brought together to begin a dialogue from which an almost immediate solution is
expected by others (e.g. the State). Secondly, additional approaches may be necessary for bringing
disputing parties to agreement, since it is unlikely that all individuals will respect or heed the
agreements struck by new institutions which have yet to gain legitimacy. For example, it may often be
appropriate that an impartial third party intervenes to facilitate a negotiation process between
particular disputing parties - or individuals - who are perhaps at the centre of the wider dispute. In
this way an interim agreement and solution may be developed with the objective of building trust and
better relations between the disputing parties.
Finally, a group of farmers - particularly in one part of Idodi - considered the continued arrival of
more herders - particularly Barabaig herders - as undesirable. While their complaints over field
incursions may have been entirely justified, they had begun to use these field incursion disputes and
out-breaks of violent conflict to politically mobilise (see below) against the continued and increasing
herder presence in the village. It was in their interests that the more overt conflict of interests over
access to land be maintained until it was resolved in their favour.
7.3.4 Controlling the terminology

Thus far, certain terms and labels have been used to classify particular groups or categories of
people and interests - for example, we have ‘farmers’, ‘herders’, ‘the state’ ‘village government’.
While I have attempted to use these labels to dispassionately describe and analyse people’s
livelihoods and land-use relations in the Idodi rangelands, the (emic and etic) terms I have employed
are based on my field data of the different socio-economic production strategies and socio-political
processes that I observed in Idodi. In describing and analysing the nature of people’s livelihoods and
land-use relations, I have controlled the terminology and attempted to use it to construct and convey
a certain set of perspectives and meanings. Thus my choice of terminology has framed the reader’s
understanding and perception of people’s livelihoods and land-use relations in Idodi.
In a similar vein, land-use outcomes in Idodi can be seen as having been heavily modulated by
people’s use of terminology - particularly in three ways: Firstly, in the way a range of different terms
225
and meanings were employed for different groups of people and types of land (see this and
previous chapters); secondly, in the way certain people’s definitions counted the most and; thirdly, in
the way different terms and meanings were deployed (or not) in particular situations in preference to
others. The first point has been woven into much of this discussion and that of preceding chapters.
However I have yet to bring in the latter two considerations, which are helpful for better
understanding the underlying dynamics of land-use outcomes in Idodi. I discuss each in turn as follows:
During fieldwork, it became evident to me that although the Idodi rangelands formed what might be
termed a ‘frontier’, in the sense that the rangelands continued to be a destination for in-migrant
farmers searching for fertile farmland and herders seeking pasture (including field grazing), the Hehe
considered themselves as being the only autochthonous group. Certainly, as the largest collective
group, they were hegemonic and controlled village government. Their understandings of how the land
should be used were dominant and they controlled (other factors not withstanding) who was allocated
land and where. Herders, as a minority group, were therefore dependent on their socio-political
relations with the Hehe for legitimately securing access to the land270 and resources they needed for
grazing their livestock and building their homesteads. Over an extended period, herders had had to
negotiate with village governments and Hehe farmers to secure and maintain their position in the
landscape. In other words, herders now occupied and used village commonage not by right, but as a
result of their negotiation with Hehe understandings and perceptions of how the landscape should be
used, and in relation to what level of herder access to the land the Hehe considered acceptable.
While negotiated land-use outcomes had many advantages for both farmer and herder, herders
remained dependant on the good will and disposition of the autochthonous Hehe farming community
for their continued legitimate access to land and resources. However, herders have recently had to
increasingly resist changing Hehe land-use priorities (particularly in Idodi, Mapogoro and
Tungamalenga villages) which overtly and covertly seek to eject and move them elsewhere in the
landscape - ‘out of the way’. For example, Salum, the senior Ilparakuyo elder, described how his
access to a rangeland area - Ihehero - that he and his associates had long used was being
increasingly challenged by the Hehe. At the time he wryly noted the symbolic significance of another
rangeland area in Idodi having its name changed - once known widely by villagers as ‘Kibikimuno’ -
after an Ilparakuyo pastoralist who had lived there many years ago, it was now to be known as
‘Lianziwa’ - a Hehe name.
Many of the terms used during public meetings convened to discuss land-use conflict emphasised the
distinction between: different land-use zones - ‘maeneo ya wakulima’ (Swahili: farmers’ areas) and
‘maeneo ya wafugaji´ (Swahili: herders’ areas); different land users - ‘wakulima’ (Swahili: farmers) and
‘wafugaji’ (Swahili: herders); different ethnic groups - ‘hawa Wamaasai’ (Swahili: those Maasai
[Ilparakuyo]), ‘hawa Wabarabaig’ (Swahili: those Barabaig), and even sometimes, ‘hawa Wahehe’
(Swahili: those Hehe). These distinctions did not always reflect reality and were often employed for
270 However, herders have as much right to gain access to farmland as autochthonous farmers – in that they can
legitimately purchase, rent and borrow farmland.
226
public posturing and political manoeuvring in public meetings (see below). It was evident that these
distinctions varyingly broke down the minute people left the meeting and as everyday social and
economic relations recommenced. The terminology in public meetings largely emphasised the
differences between people living in the Idodi rangelands, and not their commonalities. In reality
many herders were farmers, and some farmers were herders. Very many Hehe, Barabaig and
Ilparakuyo were friends, clients, patrons, stock associates, employers, employees, landlords, tenants,
herbalists, butchers, mid-wives and so on. Yet these terms of commonality were rarely, if ever, part of
the public debate over land and land-use relations. The use of terms articulating difference in public
meetings was being used by some farmers (in particular) as a political strategy to set out their claims
to land and landed resources for the future. In this context, the everyday terminology of commonality
was conveniently discarded and momentarily forgotten.
7.4 Conclusion: Landscape and livelihood outcomes
Conflict over access to land and its use has become a frequent occurrence in the Idodi landscape in
recent years, as the human population grows and as competition for access to resources and control
over their use increases. Historically, and as much today, land-use planning has been viewed both
locally and more widely by the government as a key tool for regulating landscape occupancy,
resource use practices and as a means for reducing land-use conflict. Unfortunately, the record of
land-use planning in Idodi (see Chapters Five & Six) in its various guises - from the evictions of people
and the extension and creation of protected areas in the 1950s & 1960s, the villagisation of the
Ujamaa period of the 1970s, to the creation of LMGCA in the 1980s and the latter evictions of
pastoralists that occurred in the 1990s - can be seen as having had a largely deleterious impact on
people’s access to resources. If anything, previous land-use planning has precipitated the levels of
tension and conflict over land-use that exist today in Idodi. People now live hemmed in by large
expanses of wildlife estate, once theirs and from which they currently derive a derisory benefit. It is in
this remaining and relatively restricted area of landscape which herders and farmers must interact
with each other to secure their livelihoods, and varyingly contest and negotiate their different
understandings of land and its use.
As the wetland farmlands of Idodi expand, supported by national policy priorities, pastoralism is
becoming increasingly marginalised as a livelihood strategy. In Idodi pastoralists have adopted two
different strategies to cope with the increasing threats to the pastoralist way of life and their access
to rangeland resources. Many of the Ilparakuyo and some Barabaig have increasingly diversified
and invested in agricultural production and in labour and exchange relations with their farming
associates, to enable them to access resources and to exert influence in local centres of power. An
unknown number of Barabaig herders remain heavily pastoralist following highly mobile landscape
use patterns, which enable them to have temporary access to spatial and temporal variations in
rangeland productivity at the expense of investing in stronger spatially tied socio-economic relations
with local farming communities. The case studies in this chapter demonstrate that the former strategy
of long-term investment in social networks may lead to more equitable and socially sanctioned
(legitimate) access to resources for all. However, the latter strategy of high mobility has been
227
increasingly resisted by locally resident resource users and is now much discouraged (if not made
varyingly illegal) through the state’s policies of zoning, sedentarisation and destocking.
While more formally agreed land-use and landscape demarcation can be seen as having had an
important framework role to play within herder - farmer relations, socio-political relations and socio-
economic networks can be seen as being key to the process of negotiating land-use outcomes. In
Makifu and Mahuninga the Ilparakuyo, through their numerical predominance and long-term
relationships with the farming community, have successfully negotiated a flexible set of
understandings over land-use and access. In comparison, in Idodi, the nascence of social process
between many herders (Barabaig) and the farming majority has meant that ‘on-stage’ (albeit
somewhat state-coerced) agreements have failed ‘off-stage’.
Largely state-imposed - and ‘onstage’ - formal rights-based land-use planning solutions aimed at
reducing land-use conflict may currently, at best, only constitute a framework within which different
resource users, through developing social networks and affiliations, flexibly renegotiate and gain
informal ‘offstage’ access to different resources in the landscape. The development of socio-political
networks and the negotiation of how the landscape can be used, particularly in a polyethnic setting,
may necessarily occur over an extended period of time, be recursive and involve re-occurring
dispute. To date, emphasis has been given by district and central government to ‘quick-fix’ solutions
that focus on land-use planning, often unrealistically discounting on-going and socially-grounded
land-use practices and relations in an all-out rush to stop conflict. Instead, conflict might be better
recognised as being part of a wider unfolding social process, one in which the development of social
networks and locally appropriate jural institutions and other fora are key to enabling resource users
to flexibly negotiate and achieve more equitable land-use and livelihood outcomes.
Yet at the same time, it is clear that herders continue to have inferior land rights in the Idodi villages -
particularly in regard to their occupancy of what is currently village commonage. According to state
law and as village members, herders’ legal entitlement to secure land tenure for their herding
livelihoods has not been sufficiently recognised by village governments. Although herders have
negotiated their position in the landscape and their access to village rangelands, it is clear that this
access is heavily dependent on the disposition of the hegemonic Hehe farming community. It is also
clear that while locally negotiated understandings of land-use practices have served herders’ (and
farmers’) interests relatively well, these understandings remain at risk from being rather transitory,
particularly as farmers clamour for more land for their livelihoods. Herders may need to gain secure
tenure to the Idodi rangelands using state law, since prevailing interpretations of customary law and
practice are unlikely to sufficiently facilitate this. Yet formal rights-based herder tenure of the
rangelands must have - or quickly acquire - local legitimacy if land-use conflict is not to increase. De
jure rights to resources (according to state law) may not automatically engender local social
legitimacy - i.e. rights and social legitimacy may not necessarily equate - particularly when prevailing
customary understandings of land-use and land tenure differ from those set out by state law.
228
It may be the case that the facility for ‘group’271 customary rights to village land as afforded by the
new Village Land Act (1999) is, in theory, a potentially innovative instrument for enabling pastoralists
to secure equitable rights to land and landed resources. However, the comparative examples from
Mahuninga Ward and Idodi village indicate that careful navigation of the underlying politics of land
will be necessary if herder ‘group rights’ are to be established and become locally legitimate. Recent
experiences of the Ilparakuyo in Tungamalenga are a case in point. They have thus far unsuccessfully
attempted to secure group rights to land and their claims have met with overt resistance from village
government. As with land-use planning, it will be the nature of political affiliations and networks
within the social landscape which will substantially determine how the new law is applied. The
playing-out of these socio-political processes will determine whether marginalised resource users -
such as pastoralists, will have sufficient social capital and political support to legitimately mobilise
and employ the innovative parts of the Village Land Act to secure their rights and ‘entitlements’272 to
village land resources.
As herders have gradually come to realise (see Chapter Four), local fora for collective action - in the
guise of pastoralist associations in Idodi - may be an increasingly important means for building their
political power base for negotiating, and importantly retaining, their access to key rangeland
resources. In this regard, the State - through the Ministry of Livestock - can be credited with having
catalysed the creation of these associations. However, the underlying motive for their creation has
been more for developing incentives for better livestock marketing than for supporting pastoralist
livelihoods and resource rights. Although livestock marketing is an important part of pastoralist
management - after all, male stock may be sold to maintain a predominantly milking herd - the focus
on marketing was too narrow an approach which discounted other major constraints to pastoralist
livelihoods and production. Furthermore, and perhaps unsurprisingly, there has been very little follow-
up in providing support (for example, in terms of capacity and skills training) to these associations for
improving land-user relations and developing simple and sustainable demand-driven range and
livestock management improvements in support of livelihoods. Scoones (1995, 31) states that
pastoralist organisations should start small and focus on forging collective action around common
interests - particularly socio-economic issues such as livestock healthcare or marketing, before
developing their capacity to tackle more complex issues such as range management. True, but this
approach may not always be sufficient, particularly when land-use conflict between herders and
271 The village Land Act (1999) allows for group registration of land holding such that, for example, a family or
association can lawfully secure collective rights of customary occupancy over an area of village land. The law
ascribes grazing land equal status as that of farmland and makes it possible for pastoralists and agriculturalists
to hold different rights in the same land through ‘land sharing agreements’. Importantly, the law provides for
collective pastoralist land rights across different villages.
272 The use of ‘entitlements’ does not refer here to people’s rights in a normative sense – what people should
have – but rather the range of possibilities that people can have (Leach et al. 1999, 232). Thus entitlements may
be viewed as representing, ‘the set of alternative commodity bundles that a person can command in a society
using the totality of rights and opportunities that he or she faces’ (Sen 1984, 497 cited in Leach et al. 1999,
232).
229
farmers leads to growing levels of tension and distrust in a polyethnic pastoralist association as one
group of herders is seen by the other as being the culprits for a deterioration in key land-user
relations with farmers - as has occurred in Idodi.
Experience from elsewhere in Africa, as reviewed by Sylla (1995, 135-149), points towards the
need for pastoralist organisations to develop flexibly - both in terms of structure, membership and
size; that starting small is wise; that across an ecosystem or large range area, smaller organisations
working together are better than one large organisation; that wider herder inclusiveness - both in
terms of (poorer) socio-economic and (weaker) socio-political status is important; that the strengths
and weaknesses of traditional systems and their structures need to be taken fully into account; that
planning must be iterative, adaptive and flexible; that neither privatisation nor collectivisation may
provide appropriate range tenurial frameworks; that subsidiarity should be an underlying principle,
and; although not easy to achieve, long-term dependency on external support (particularly financial
and technical) should be avoided (Sylla 1995, 149-152).
In Idodi, the development of pastoralist associations is likely to be all the more difficult given the
polyethnic nature of the herding community, their relatively poor social cohesion and their
marginalised status. If these associations continue to remain weak, in the absence of social and
political support by village assemblies and within local government, the new land laws will remain
outwith the grasp and benefit of these marginalised resource users.
230
8 Conclusion
Enkiterunoto oo ‘lMaasai o ‘lMeek: The beginning of the Maasai and the

Cultivators273
When Leeyo (the first Maasai man) became a great elder, he called his
children and said, ‘My children, I am now an elder of many days and I want
to instruct you”. He then asked his eldest son, “What is it that you want from
my treasures?” And the eldest son replies, “I want everything in this country”.
And the old man said, “Since you want everything, take a few head of cattle,
a few goats and sheep, and some food of the earth, since there are a large
number of things”. The eldest son replied, “Very well.” Then Leeyo called his
youngest son and said, “And what is it that you want?” And he said, Father, I
wish that I should be given that fly-whisk in your hand”. And his father said,
“My child, because you have chosen only this fly whisk, may God give you
prosperity, so that you will have control among your brothers.” And so the one
who wanted everything became a cultivator, and he who took the flywhisk
became the father of all the Maasai.
In this chapter I relate the major findings of my field study to the central theoretical questions that I
raised at the beginning of the thesis. I then discuss the relevance of the insights drawn from this thesis
for the land-tenure reform process and land management outcomes in the dryland-wetlands of
Tanzania.
8.1 Disaggregating trends in access to land in Idodi
In Chapter Two, I asked three questions. The first question was,
What are the major wider factors that have influenced increasing socio-economic
differentiation and unequal access to land and landed resources at local level?
Using selected examples, my underlying intent was to investigate the extent to which local land-use
practices and systems have been impacted by the state and what impacts these developments have
had on people’s livelihoods in the Idodi rangelands.
It is clear that land-use relations in the Idodi rangelands have been heavily shaped and impacted by
the state. Local people’s occupation and use of the landscape has been transformed over a 50 year
period through state-mediated landscape demarcation and land resettlement, principally as a result
273 Translation taken from Rigby (1992, 195).
231
of protected area creation, Ujamaa villagisation and appropriations of land elsewhere in the country.
In-migrant farmers and herders continue to move into the rangelands of Idodi and Pawaga, often
attracted by the possibility of rice farming and field grazing. Thus the once relatively sparsely
populated rangelands of Idodi are now relatively densely settled, and increasingly characterised by
competition (particularly between herder and farmer) over access to key resources. The concentration
of a growing population of farmers and herders in a relatively small proportion of the former extent
of the Idodi rangelands has exerted increasing strains on people’s land-use relations and access to
land.
As Peters (2004, 304) states, relations around land are not just socially embedded, but are
embedded in unequal social relationships. One way of examining the unequal nature of these
relationships is to investigate some of the socio-economic factors underpinning people’s local access to
various different types of land at village level. Thus in Chapter Five I showed for farmers that, for
less valuable and more widely available dryland, access was relatively equitable, although there
was an emerging class of landless farmers. Contrastingly, for higher value and much more productive
wetland, access was far less equitable, and was strongly skewed towards a relative minority of
wealthier farmers. This trend strongly points towards an ongoing process of land accumulation and
social-differentiation, supporting Peters’ concerns about growing inequality in access to land at local
level. In Idodi, wealthier farmers have generally legitimately increased their holdings of wetland
fields through a strategy of gradually purchasing fields, although perhaps more insidiously, processes
of debt entrapment (as described in Chapter Five) accelerate wealth differentiation and may have
indirectly expedited land accumulation. Yet at the same time there are some exceptions to the trend.
Some marginalised and poor farmers have been purposefully extended access to relatively high
value riverine gardens by village governments for ensuring their food-security. By no means have all
poor farmers been allocated riverine gardens, but the data gathered indicate that wealthier
households do not own disproportionately larger amounts of this category of land. In summary, socio-
economic differentiation is definitely occurring in the Idodi rangelands, driven by access to a
restricted amount of highly productive land. However, overall access to the remaining dryland
continues to be relatively equitable between farmers, although there is an emerging class of landless
people.
In Chapter Six, I concluded that there was almost certainly a trend towards increasing socio-economic
differentiation among the Ilparakuyo, despite the on-going traditional practice of wealthier
households providing (milch) stock loans to support poorer households. This reflects a wider trend
documented by, for example Ellis (2000) and Bryceson (2002)274, income diversification is a major
factor in enabling households to become better-off. Although national and local political economy has
led to all Ilparakuyo pastoralists in Idodi becoming more sedentary, owning less livestock and being
increasingly reliant on agriculture, some pastoralist households have adapted well to this
274 Although both Ellis and Bryceson focus on off-farm income diversification, given the limited opportunities for
off-farm income earnings for people living in Idodi, on-farm income diversification can be seen to be an
important factor underlying increasing socio-economic differentiation among the Ilparakuyo.
232
transformation and others less so. The reasons for this difference are not understood, although I would
suggest that they may lie more within households and extended families rather than with external
factors. Certainly there is a multiplier effect at work as households that have sufficient livestock to sell,
may temporarily invest the proceeds in rice farming before re-investing the profits back in livestock
or other forms of accumulation - such as wetland fields. Those households without sufficient livestock
are unable to farm rice, and therefore miss out on the accumulation opportunity. Dryland farming
rarely, if ever, is a means for wealth accumulation and only serves to provide household grain
provisioning needs. At a collective level, Ilparakuyo and Barabaig herders have been able to acquire
sufficiently secure tenure to farm fields, and to similarly benefit from agriculture as the farming
community. However, they have been far less able to sufficiently obtain secure tenure to village
commonage for grazing their livestock. Extensive grazing systems, in effect, are not perceived by
farmers as comprising a sufficiently legitimate use of the land, particularly since only a very small
minority of the autochthonous farming community keeps livestock. Moreover the farming community is
apprehensive of permanently making-over areas of village commonage to herders for grazing as this
would be seen as endangering farmers’ longer term interests and their need for expanding the area
under cultivation. In this regard then, herders’ access to grazing land is embedded in an unequal
social relationship with the majority farming community.
In drawing conclusions about the increasing inequality, competitiveness and conflict over access to
land, it is important that I not make over-simplifying assumptions about the complexity of land-use
relations. Thus at a local level, I have not been able to examine in detail, for example, intra-
household inequalities in access to land that may impact on, for example, youth and women in
particular. Nor have I been able to examine local patterns of inheritance or land distribution within
kin groups, which in the patrilineal societies of Idodi, may discriminate against vulnerable groups -
such as widows or divorced women. With regard to the state, I have largely treated the government
as a black-box and I have not examined - except in passing - the impact of other key state policies,
for example, the legacy of twenty years of structural adjustment on herders and farmers. Instead, at
the beginning of the study I chose lines of inquiry which I considered would most effectively provide
some incisive insights into key changes in people’s access to land, their livelihoods and land-use
relations in the Idodi rangelands. Thus despite the above caveats it is clear that the state has had a
very significant impact on people’s land use and livelihoods in the Idodi rangelands. The State’s
legacy has been compounded, perhaps increasingly in recent years, by growing inequality in access
to the most productive land - a result of local processes of accumulation and social differentiation.
Thus, while highly productive land is slowly being accumulated in the hands of wealthier farmers,
herders are struggling to maintain sufficient access to the rangeland resources they need for their
pastoralist livelihoods. Undoubtedly, people’s access to land has been heavily structured - or
constrained - by state land-use policies, and their livelihoods have been adversely impacted by
national macro-economic policies (for example, the withdrawal of crop marketing support,
agricultural inputs subsidies and livestock health services). Yet within this context, I argue that people’s
land use-relations in Idodi are today increasingly modulated by local level politics and socio-
economic relations between different land-users - in this case, herders and farmers. This is the focus of
my second question.
233
8.2 Negotiated land-use agreements: too much flexibility?
The increasing importance of local-level power relations provides the context to the second question,
When and how does local social flexibility and negotiation over land lead to
inclusion or exclusion at a local level between different resource-users? How does
this take place?
In asking this question, I set out to examine some of the underlying socio-political processes behind
land-use relations between farmers and herders that have been characterised by increasing
competition and conflict in recent years. Herder’s rights to the Idodi rangelands - or village
commonage, as compared to their access to farmland, may be described from a legalistic standpoint
as comprising rights little better than those of squatters. In this context in Chapter Seven, using
extended case studies, I found that in some villages in Idodi, socially negotiated land use outcomes
achieved by herders with local village governments and the autochthonous farming community remain
key for maintaining herders’ locally legitimate access to land. Where socially negotiated land-use
practices are still developing, or have been eschewed by one or other parties, pronounced levels of
competition and conflict of interest have occurred. I found that hastily constructed and imposed land-
use plans cut across production relations and had, at least for the time-being, quickly unravelled.
These plans were developed without an extended process of social negotiation necessary for their
wider legitimacy. Neither herder nor farmer fully accepted them, and village governments and local
courts did not have sufficient legitimacy, incentive or resources to enforce them. Furthermore, imposed
land-use solutions, which are carried out with the intent of effecting swift control of land-use conflict
and also sometimes safeguarding minority user rights, may only serve to mask the underlying conflicts
of interest over land-use, not resolve them. Yet at the same time, the flexible nature of negotiated
outcomes is as much a disadvantage as an advantage for marginalised groups such as herders, as
their access to resources remains at risk from the changing priorities of more powerful resource users.
When herders have attempted to negotiate more formal rights to land and to secure certificates of
occupancy, they have been frustrated as village governments have manoeuvred to evade or delay
their requests. This outcome in Idodi holds strong parallels with the observations of other researchers
who have drawn attention to the considerable social inequality in many customary systems (Platteau
1996, 2000; Ribot 2000; Carney and Farrington 1998; Woodhouse et al. 2000 and Amanor 2001
cited in Peters 2004, 277). In summary, while social negotiation between marginalised resource users
and more powerful autochthonous farming communities has been crucial for the formers’ continued
access to key landed resources in Idodi, negotiated land-use outcomes remain insufficient for herders’
long term land tenure security in the context of an expanding farming population.
8.3 Looking to the future: the case of herders in Idodi
I now turn to considering the third question which asks,
Scholarship has shown that past and continuing efforts to reinvent, subsume and/or
extinguish customary law within national law throughout sub-Saharan have
234
contributed to growing social differentiation and landlessness. In this context, to
what extent does an alternative approach recognising the legitimacy of customary
land tenure practices reinforce the growing trend of social differentiation and the
increasing numbers of landless and resource poor observed across Africa? Under
what circumstances might such an approach limit or reverse the trend?
In posing this final question, I set out to investigate whether the new - but varying - emphasis on
recognising customary land tenure will merely reinforce the growing trend of social differentiation
and the increasing numbers of landless and resourceless poor observed across Africa. The new
Village Land Act of Tanzania devolves substantial control over land management to the village level
– and in this regard may, at least to a certain extent, be seen to be evolutionary in intent. Although
the law attempts to improve the equitability and accountability of village land management, and
pays deference to the rights of vulnerable and marginal categories of people, it does so through a
necessarily legalistic and heavily bureaucratic approach. The approach focuses on formalising local
or customary land rights and procedures, sometimes to the extreme, in attempt to sufficiently regulate
local land management. An underlying assumption is that the formalisation of land tenure and land-
use relations will lead to more efficient and equitable land management. This is hardly likely to be
the case. The law as it stands cannot stipulate, far less control, the underlying social relationships
between land-users which comprise and embody people’s rights in reality. Two immediate
implications follow:
Firstly, it is clear then that if the law is to facilitate more equitable and accountable village land
management, less focus needs to be given to the bureaucratisation and formalisation of local land
tenure practices, and far more consideration given to facilitating the interplay between formal and
informal systems of land tenure management. Variably pluralistic jural institutions do exist at local
level (village, ward and primary courts), but they remain weak, are vulnerable to manipulation by
elites and litigants and are often not the most appropriate institutions for modulating equitable land-
use relations and safeguarding the land entitlements of vulnerable and marginalised people.
Although the new land laws set out the creation of dispute resolution institutions (village land councils,
ward tribunals and district land and housing tribunals), these institutions have not started to operate in
Idodi. The new land laws also provide for village land adjudication committees overseen by village
adjudication advisors. The emphasis of the law heavily lies with legalistic processes of adjudication
(albeit with some form of consultation between parties being stipulated) and the assumption that
these institutions will operate impartially and competently. Given that these institutions currently do
not exist, it is not clear as to how they will work (or not) to equitably mediate between complex state
land laws and increasingly varied interpretations of local or customary law (given the polyethnic
nature of rangelands such as Idodi). It is also not clear what priority the entitlements of minority,
vulnerable and marginalised groups will receive, particularly in polyethnic rangelands. It is one thing
for land courts and adjudication committees to adjudicate disputes fairly between, for example, two
relatively wealthy farmers, and entirely another to ensure the wider development of equitable yet
flexible land management systems that safeguard the interests of vulnerable and marginalised
groups.
235
Secondly, the reality of unequal socio-political and socio-economic relations between different land-
users - particularly in relation to vulnerable and marginalised groups - is hardly liable to be changed
by the law. In Idodi, herders as a marginalised socio-political category had recently started to
organise collectively, partly in response to land-use conflict and partly as a result of the potential
prospect of attracting livestock development funds from the state. I argue here that these associations
comprise an opportunity for herders to build their socio-political status as a group and to collectively
develop the legitimacy of pastoral range tenure and range-use practices. As I discussed in Chapter
Seven, the challenges faced by these associations are considerable. They are characterised by low-
levels of credibility and legitimacy among the pastoralists who they are supposed to represent and
support, and they are currently prone to varying levels of mistrust between members, particularly
since the associations are comprised of Ilparakuyo and Barabaig factions. In trying to improve the
socio-political status of pastoralists at different levels of local government, the associations will also
have to start to contend with conflicts of interest among their own members - for example, in
persuading individual herding households to accept some restrictions on seasonally accommodating
the livestock of stock associates and kin so as to reduce the incidence of farmer-herder land-use
conflict and local pressure on key resources in the landscape. Balancing the interests of individual
herder households against the wider collective herding community through the flexible development
of new norms will necessarily take time and be subject to recurring dispute and dissent.
The local development of range-use norms and practices will necessarily depend on integrating
different customary range-use practices of different herding groups and simultaneously negotiating
with formal government range-use policies and interests. This strongly suggests that the pastoralist
associations in Idodi, as is likely to be the case for other polyethnic herding groups in other
rangelands in Tanzania, will need to develop into ‘hybrid’ formations of a formal institution and
informal organisation. The distinction between formal institution and less formal organisation may
seem merely conceptual here, but it is important and Swift (1995, 154) draws attention to the
difference. Institutions can be seen as a structure or set of rules that set out how people should
interact. Contrastingly, organisations may be viewed as groups of people bound by some common
purpose in order to achieve a set of objectives (North 1990, 5 cited in Swift 1995, 154).
Organisations can thus be viewed as operating within the framework - the structure and rules -
provided by institutions (Swift 1995, 154).
As hybrid formations, herder associations may benefit from being recognised as formal institutions by
local and central government, yet as informal organisations they may derive their local legitimacy
through the support of different groups of herders who are able to find commonality in their diverse
interests. Thus these ‘hybrid’ formations may be an important component for achieving the interplay
needed (for example, see Maganga 2003 and Swift 1995) between formal state laws and
structures, the informal matrix of social relations that embody people’s rights and frame people’s
livelihoods.
236
8.4 Business as usual?
In returning to Peter’s central concern over increasingly competitive and conflictual relations over land,
it is clear that the new evolutionary approach to land is double-edged. On the one hand the principle
of subsidiarity and deferring management to a local level presents a very important opportunity for
re-investing in long-marginalised customary or local land management institutions and practices.
Arguably such an approach is the most appropriate for the development of locally legitimate,
responsive and equitable land management systems. On the other hand, if at least some elements of
the new Village Land Act embody the new evolutionary approach to land tenure law, then there is a
substantial risk that the devolution of land tenure will reinforce growing social differentiation and
landlessness, and increasing competition and conflict over land that we have seen, for example, in
Idodi. This is because the new Tanzanian land laws make inter alia highly unrealistic assumptions
about local administrative capacities (Sundet 2005, 16) and the likely equitability of local
administrative processes. Enabling rural Tanzanians to benefit from the Village Land Act while
minimising its substantial risks, will demand a high level of new skills, well allocated and managed
resources and a different level of commitment on the part of the state and other third parties
focussed at village and district level.
Ironically, where rural villages have taken advantage of the Village Land Act (1999) and have
independently strengthened their land management systems with the support of nongovernmental
organisations, the State has sometimes sort to undermine their achievements. Thus many of the
pastoralist villages in Loliondo Division in Ngorongoro District in northern Tanzania have developed
and are implementing their own communal land use management plans. These land-use plans are
based on well-established range-use patterns, and accommodate the interests of private tourist
operations that pay the villages for their access to particular areas of village land. The State –
through the auspices of the Wildlife Division – has attempted and failed to impose restrictions on
village land-use which would have resulted in the State re-asserting its control over village lands and
the business agreements struck between the villages and the tourism industry.
Finally, a sober reflection of reality suggests that currently neither the new skills, nor real priority on
the part of the State, nor well allocated and managed resources currently exist for addressing the
growing inequality and competition over access to land in Tanzania at village level. A key (draft)
strategy document for implementing the new land laws (MLHSD 2005) has adopted a highly
technocratic, formulaic and interventionist approach focussed on formalising and constructing new
land management processes and systems in Tanzania - at great projected cost. The Strategy is
correct to identify village lands as a priority area for implementing and applying the new land laws,
but the strategy as it is laid out is much less than satisfactory. For example, the strategy goes so far
as to set-out the need for a National Village Resettlement Scheme that amounts to a blue-print land
reform reminiscent of Ujamaa (the creation of new separate ‘farmer’ and ‘pastoralist’ villages) for
resettling ‘landless’ people (URT 2005, 28). The strategy again returns to the old chestnuts of
targeting ‘nomadic cultures’ and ‘excessive stock holdings’ as the underlying causes of land-use
conflict, yet also identifies the violation of land rights (quite correctly) as a fundamental issue. Overall
the strategy is hopelessly optimistic about the pace of projected implementation of parts of the new
237
Village Land Act (most of the actions identified in the strategy, it would seem, are supposed to be
realisable in three years for the whole country). Far more attention and projected expenditure is
given in the strategy towards building state institutions and technical processes and relatively very
little to supporting village-based institutions and village-based land management capacity and
systems. Given the findings of this thesis, and the insights of others (e.g. Shivji 1998; Sundet 1997 &
2005), such an approach to developing land management in Tanzania given the long process it has
entailed thus far, leads one to ‘not a cheerful conclusion’ Sundet (2005, 16). It appears that rural
Tanzanians have again come last in line as the State pre-occupies itself with an expensive, hardly
affordable and centrally controlling system of land administration. It is likely that the highly
technocratic land administration system will largely benefit the (urban) wealthy and the elite at the
expense of the majority of Tanzanians, and particularly the poor. Instead, and with regard to the
central questions of this thesis, a well focussed and resourced initiative at village level targeted at
enabling marginalised and vulnerable groups in society to safeguard their entitlements would be
have been a good start to opening up and more constructively addressing critical land and resource
issues for many rural people across Tanzania.
238
Appendices
A1: A short overview of Local government in Tanzania

Village government and administration, was formally begun in 1975275. Before this time, villages
were not legally recognised entities. For many years, from the late 1970s until the mid 1990s, village
government was an instrument of central government, in practice allowed little autonomy and
expected to implement and/or fully cooperate with central government policies and programmes.
This has gradually changed over a period of 20 years, as village councils have been allowed more
autonomy, and as attempts have been made to improve their democratic functioning (see Table A1).
However, despite steps taken towards strengthening village councils, village government remains
weak and plagued by revenue shortages, major issues of accountability, and low administrative
competence - particularly in the remoter and poorer areas of Tanzania such as Idodi.
Table A1: Key developments in village-level government and administration - modified

from Shivji and Peter (2000, 46-53)
1975 Villages and Ujamaa Villages Act of 1975 - and subsequent supporting legislation
Villages were legally recognised after the villagisation campaign from 1971-1975,
which created many new villages.
A process of village registration for those villages with defined boundaries was begun
- now over 10,000 registered villages.
Institutions of village administration were established but under the auspices of the
central single ruling party (CCM): (i) A village council with five committees - finance
and planning; production and marketing; education, culture and social welfare; works
and transport; security and defence; (ii) A village assembly consisting of all the
members of the village above 18 years of age.
The main role of the newly recognised village assembly is to elect the Village Council.
In turn village councils had the power to call a village assembly consultative meeting
as needed.
The main role of the new village councils was to manage and coordinate the
development affairs of the village for social and economic development.
Village councils were conferred corporate status (with the right to sue and be sued).
Village councils did not have any government powers except for those necessary to
implement decisions made at higher party and governmental levels.
The village was thus considered a site of autocratically carried out development
administration rather than a locus of devolved government.
1982 Local Government (District Authorities) Act of 1982 - and supporting legislation
The village assembly was recognised as the supreme policy-making body in relation to
village affairs and was now able to remove any or all the members of the council -
although its powers remained theoretical and were never really realised.
275 With the passage of the Villages and Ujamaa Villages Act of 1975.
239
Village councils were enabled to make bylaws in consultation with the village
assembly, but subject to the overriding powers of the Minister. The respective
District Council had to approve the village bylaws for them to take effect.
Executive and legislative powers are therefore formally merged into one body - that
of the Village Council. In addition, village councils often played a de facto ‘judicial’
role, particularly when primary courts were not locally available.
Divisional and ward secretaries were given powers to enforce all bylaws within their
areas of jurisdiction - both district and village bylaws.
Village councils were permitted to raise revenue through licenses, permits, fees etc
subject to the directions of central government and district council authorities.
Previous more liberal revenue-raising powers introduced in 1979 were rescinded.
Overall, the power of the village council was subordinate to that of the Minister and
corresponding district council, and thus village councils remain devoid of any
significant autonomy.
Other minor changes included: (i) Extending the tenure of office for village
councillors from one to three years; (ii) The election of one village chairman (within
the District) to sit on the District Council, and; (iii) A statutory requirement that the
village assembly hold at least three ordinary meetings per year.
1992/5 Eighth Constitutional Amendment of 1992 - leading to Local Government Laws

(Amendment) Acts of 1992, 1993 (twice), 1994 and 1999
Eighth Constitutional Amendment introduced multi-party politics to Tanzania.
Each village is now sub-divided into to a maximum of five vitongoji as decided by
each village council and approved by its respective district council.
Vitongoji chairpersons are now to be elected by each kitongoji’s members.
Literacy, tax clearance and party affiliation were made qualifications for election
eligibility in addition to existing adult age and Tanzanian citizenship
requirements.
Vitongoji chairpersons are to be automatically appointed ex officio members of
the Village Council.
Village councils now consist of no less than 15 and not more than 25 members
(although the latter is now usually taken as the norm).
Periods of office for village council members have been extended from three to
five years.
A village chairperson must now be elected by the village assembly and, in theory,
need not necessarily be a member of the ruling CCM party.
Other council members are to be elected by the village, and not less than one
quarter of the total complement of council members are to be women.
Village council chairpersons are now removable by a two-thirds majority vote of
the village council although the village assembly was not given similar power.
Vitongoji chairpersons are now removable through a simple majority vote of a
kitongoji’s residents.
The 1999 amendment places an emphasis on decentralising and devolving power
not only form central to local government, but ‘within the local government
system, from district council levels to lower levels of local government’ - i.e. the
village level (Shivji and Peter 2000, 53).
240
Local government is today structured in three tiers - district, ward and village level, the most
important of which are arguably the district and village levels (see Figure A1). In addition to the
three levels of local governance, there is an administrative divisional level. In contrast to elected local
government, the ‘division’ is part of a different top-down administrative system that is not accountable
to the electorate. Although a legacy of the colonial past, this system - consisting of regional and
district commissioners and their subordinates - has been retained as it facilitates direct high-level
government control and intervention at a local level. On administrative maps, divisions fall under
districts, wards fall into divisions, and villages fall under wards.
Figure A1: The normative structure and functioning of local government in the Tanzanian
mainland
[Adapted from Shivji and Maina Peter 2000]
Level of Local Government Executive structure Legislative Structure
District Executive District Council

District
Director Chairperson
District Council
Ward Executive Ward Councillor

Ward Officer
Ward Development Committee
Village Executive Village Council

Village Officer Chairperson
Village Kitongoji
Village Vitongoji Members Chairperson
Village Council
Village Council
Members
Village Assembly
Legend:
Lines of accountability and responsibility Electoral process Provision of service
The village assembly is now ostensibly considered the foundation of local government (Shivji and
Maina Peter 2000, 35-36). The village assembly directly elects a village council, a village
241
chairperson and, with other village assemblies in its ward, a ward councillor, who is the chairperson of
the ward development committee (WDC). The ward councillor is a member of the district council, and
together with ward councillors from other wards, elects the district council chairperson. The district
executive consists of a district executive director (DED) and district department heads for agriculture
and livestock, natural resources, lands, water, health, education, and other services. In theory, the
district council now employs its entire executive staff on a competitive basis, and it is only the regional
and district commissioners, and their administrative and divisional secretaries who are presidential
appointees. This underpins the difference between the two administrative and governance systems:
local authorities are, at least in theory, increasingly democratic whereas commissioners and their
respective staffs are agents of the central state and continue to fulfil an authoritarian role.
242
A2: A FAO land cover scheme for the Idodi and Pawaga rangelands
Figure A2: A FAO land cover scheme for Idodi and Pawaga villages (FAO land cover data 2002)
Agricultural land cover Forest / woodland land cover
Ruaha National Park Pawaga Ruaha National Park Pawaga
LMGCAS LMGCAS
6
6
7 7
4
4
3 5 3 5
2
Idodi Idodi
2
Usangu Usangu
Game Game
1 1
Reserve Reserve
Open grassland / range land cover Legend:
Ruaha National Park Pawaga
LMGCAS
6
7
4
3
5
2 Idodi
Usangu
Game 1
Reserve
1 – Mahuninga 5 – Idodi
2 - Makifu 6 – Malinzanga
3 - Tungamalenga 7- Nyamahana
4 - Mapogoro
0 20 30 40 km
Scale:
FAO Land Cover Metadata base, 2002 http://www.africover.org/ (free public access)
Overlay grid projection: UTM Zone 37s, Arc 1960
243
The FAO data, which only became available after fieldwork was completed, is remarkably
representative although not entirely accurate. Some areas of agriculture have been well captured by
remote sensing (such as in Idodi), other areas - such as the more extensive rice paddy in Pawaga,
appear to be less well portrayed. In this regard, rangeland category RL6 (closed herbaceous
permanently flooded grassland) appears to be incorrect in that the corresponding area denoted on
the map largely comprises rice paddy fields. Also, the agricultural categories AG1A-C need to be
qualified by the proviso that most of these areas are only sporadically farmed, and instead in many
places consist of a variety of tree-grassland mosaics.
244
A3: A Provisional Gazetteer of the Idodi and Pawaga Villages
During field work, as part of tracing some of the past events of the Idodi and Pawaga rangelands, a
gazetteer of former settlements in the Ruaha River valley in what is now the Lunda-Mkwambi Game
Control Area and the southern extent of the Ruaha National Park was developed with the aid of key
informants. The gazetteer is not intended to be a definitive historical record, but rather mostly a
description of changes that occurred in the 1940s - 1970s as a result of protected area creation and
villagisation during Ujamaa. The gazetteer is a work-in-progress.
Table A2: A provisional gazetteer of former settlements in the Ruaha River valley in what
is now the Lunda-Mkwambi Game Control Area and the southern extent of the Ruaha
National Park
[See Figure 5.1 for the corresponding map]
Settlement
Coordinates Name Description
Number
Abbreviations: NFD = No further details; MDA = More details available.

Notes:
1. The names of particular patrilineages given are not necessarily exclusive to each settlement as there was some movement
of people between different settlements over the years.
2. Grid references are estimates - derived through discussion of landscape features and the number of hours taken to walk
between settlements where no defining landscape references were available. There is a reasonable amount of confidence
associated with them unless otherwise indicated.
3. Informants: The descriptions are largely based on those given by: Saleh Petwa (Tungamalenga 1-48); George Mtati and
Mikelo Ndindile (Mahuninga; 50-56; 58-61; 86-90); Andreas Nganylika, Wazee Nyemile, Kihunye, Kaundama (Tungamalenga and
Kitisi 49;57; 86-90).
1 7 21 91 93 Kwa Mlele An old 19th century settlement which by the early

20th had been abandoned due to drought. The
inhabitants moved to Igawa and Ilolo in Pawaga. NFD
2 7 24 91 85 Chemchem ya Magombwa A series of natural seepages/springs that were

used/frequented by elephant
3 7 31 91 84 Magombwe An old settlement in a low lying area between two

watercourses that was swampy. The inhabitants left
sometime in the mid 1940s (1946) during the Njaa ya
Ihambwa due to the drought
4 7 36 91 81 Makutupa A settlement by a hill overlooking the Mbuga ya

Chamguite which was abandoned during the 1940s
(1946?).
5 7 35 91 73 Chamguite Inhabited until 1955. The drought of 1949 resulted in

most people seeking refuge south of the river at
Igawa. Post 1949 there were about 3 households left.
Names of families remembered are Mkonongo and
Mwagiyumbile.
6 7 33 91 70 Mdeya Inhabited up until the 1974 Uhamisho, it had

previously taken in evictees from the north bank of
the Ruaha. Livestock were kept - especially goats.
Cattle described as dying off post 1955 from tsetse
fly invasion. About 20 households in the late
1950s/early 1960s. The major lineage was that of
Nyambarazi
245
Settlement
Number
7 7 37 91 71 Igawa Inhabited until the 1974 uhamisho. Received a

substantial number of people from the north
bank/area of the Ruaha post 1955. In 1949 a
substantial number temporarily (some permanently?)
moved to Idodi area to farm the Vynungu, returning
the following late dry season to farm again. There
was at one time a substantial number of stock -
perhaps a '1000' (sic) pre 1949 which were grazed in
the Chamguite grasslands on the northern bank
among other places (MDA). This village belonged to
the Kayela and Petwa patrilineages who were the
local elite - both administratively and culturally. The
term nchi ya Kayela was used. At its zenith, the
settlement had between 30-50 households - after the
1955 Ruaha evictions and before the 1974 uhamisho.
Another patrilineage name mentioned was
Nyambarazi. Post 1955 cattle numbers are described
as having declined due to the invasion of tsetse fly
and sleeping sickness.
8 7 39 91 74 Mwaliganzi Post 1949, abandoned due to drought and famine.

Inhabitants moved to Pawaga and Idodi and
subsequently did not move back once rain had
returned. A notable patrilineage which moved to
Isanga (Makifu) was that of Mayanga.
9 7 37 91 72 Nyirenge Abandoned post 1949, inhabitants moved to Igawa,

Ilusi and Idodi. Apparently a large settlement of
Kosisamba. Consisted of about 20 households and
notable patrilineage names included: Ngabala*,
Mterere and Wanzalilo. * Descendants live and work
in Msembe for TANAPA
10 7 45 91 74 Ikelekembe Abandoned post 1949 due to famine, the inhabitants

moved to Ikorongo, Pawaga and Idodi. Pre-
abandonment there were about 5 households with
substantial cattle holdings. The major patrilineage
was Mwisangu.
11 7 46 91 76 Uzimoto Abandoned and not resettled after the 1949 famine.

Identified as exclusively a 'Hehe' settlement of about
20 households. 2 major patrilineages were Mwisangu
and Mnemba.
12 7 42 91 83 Chauumu Settled by one old man and his family and

descendants - Mzee Kuwalunda Kinula. A 'substantial'
number of cattle were kept which were seasonally
grazed in the dry season on the banks of the Ruaha
River and moved to pasture in the vicinity of Urungu
hill during the wet season. The foloboto famine of
1949 forced the whole extended family to move to
Ikolongo whence they never returned.
13 7 21 91 62 Mwagusi These settlements were inhabited until the late

14 7 23 91 62 Ifugulu 1940s, people then moved to Pawaga and Idodi and
across the river to Makuluga. There were about 30
15a 7 23 91 67 Ididimo(a)* and Kikoga (b)*
households on the north bank of the Mwagusi at
15b 7 24 91 72 *Exact location uncertain - Ifugulu, not far from the confluence some of which
but both on the north kept cattle. Notable patrilineages included Kayela,
escarpment, Ididimo lying 6- Mhegele, Mwivhiga, Mwasimba as well as Chaka in
7kms south of Kikoga. the hill settlement of Ididimo. Cattle were grazed up
on the escarpment during the wet, being moved
down to the river during the dry.
246
Settlement
Number
16 7 30 91 63 Mkwata A hamlet that was established post the 1955 evictions

from the north bank, it was only occupied until 1958
before people gave up and moved to other villages
such as Igawa. Originally the inhabitants came from
Makuluga (principally). The reason for abandoning
the hamlet was wildlife depredation of crops.
Notable patrilineage was Mwisamba.
17 7 23 91 61 Makuluga An important farming area with riverine shambas.

Formerly stocked with cattle, with increasing tsetse
infestation, the cattle were superseded by small
stock (goats). The main crop remembered as being
maize but also finger millet. Two events - the 1949
famine together with the 1952/3 flood resulted in
most people moving to Kiganga. Mdeya, Igawa and
Ilusi. Patrilineages remembered are: (Hehe) Mwilwa,
Kidunya, Kisinga Mwasimba; (Kimbu) Matoweskola,
Matidole.
18 7 20 91 60 Ilala Simba A peripheral settlement to Makuluga - abandoned

sometime in 1949 as a result of drought. Maize and
peanuts were grown and some finger millet. May
have had approaching 20 households. Patrilineage
names remembered are Matoshola and Kidunye.
19 7 42 91 65 Ilusi A major settlement extant until the 1974 uhamisho

and then moved out by lorry. Principally an
agricultural area with maize and a lesser amount of
finger millet grown in Mipogoro (Acacia albida)
farmland. At its most populous, perhaps towards the
late 1950s, about 50 households. Goats came to
replace cattle as the latter declined due to tsetse fly
and sleeping sickness. Flood irrigation was practiced
in between the confluence of the Ilusi and
Tungamalenga rivers as well along the outer banks.
People were identified as being Hehe descended
from Nyambarazi and Kosisamba. Patrilineages noted
included (there were 'many') Kaundama, Mulimbwa
and Kigelelo as well as Ngabala (who latterly) moved
from Igawa.
20 7 42 91 63 Nyawagulu (grid ref Wet season cultivation area along a narrow riverine
uncertain) flood-plain. Maize principally grown. Farmers
returned to Ilusi during each dry season.
21 7 50 91 70 Wheluka Not far from a swampy area, an area where finger

millet was cultivated, with drinking water having to
be fetched as far as from the Ruaha. During the 1949
famine, the majority of the settlement (about 10
households large) moved to Ikolongo and Ilunda -
especially for water. While some people appear to
have returned or remained behind in Wheluka, all
eventually moved to Ikolongo and thereafter in 1974
to Mlowa during the uhamisho. Inhabitants identified
as being Hehe and Kosisamba, patrilineage names
remembered as: Malinga, Mulumbwa and Mbwawa.
22 7 58 91 75 Uwalinzota A small hamlet which farmed maize and some finger

millet. The inhabitants were moved out in the 1974
uhamisho and during the 1946 and 1949 famines,
people temporarily sought refuge in Ilusi and
Ikorongo before returning. Patrilineages remembered
as Mlimbwa and Muyovela.
247
Settlement
Number
23 7 63 91 77 Kihanga A wet season farming area. Farmers used to come

from the Pawaga villages to cultivate (especially
from Kisoloka and Kisanga).
24 7 69 91 81 Kisoloka A mixed farming settlement with goats and cattle.

The settlement remained post the 1974 uhamisho.
Predominantly a Hehe settlement. Patrilineages:
Lulawa, Mbimbi and Lulinga.
NB: The current villages of Pawaga - Kisanga, Itunundu, Mboliboli, Isele and the post Ujamaa settlements are not
included here.
25 7 64 91 85 Cheleganza A settlement that was started by and received a

substantial number of refugees from the hamlets in
Igawa and environs affected by the 1949 famine.
Subsequently used as a drought refuge in subsequent
years. The patrilineage graves of Petwa and Kayela
lie here.
26 7 65 91 87 Kikulwe A small agricultural clearing in the then extensive

groundwater forest - NFD
27 7 60 91 83 Mgonadele Settlement established in the mid 1960s for livestock

keeping and dominated by the 'Wanyamkuya' who
moved to Pawaga from the Kilombero early in the
20th century(?). They are identified as being
responsible for having brought rice farming to
Pawaga. Patrilineage name of Matimbo identified.
28 7 58 91 83 Kimaling'oko The site of two large baobab trees, one of which has
since died, both of which marked the boundary
between Kayela and Ilolo jumbeates. This area and
that on the northern bank were especially frequented
by Maasai who had a long-established presence in the
area. Names of particular Maasai remembered are
Sikona, Kisota and 'Manuelo'.
29 7 54 91 80 Ikorongo Settlement fell derelict after the 1974 Uhamisho.

Identified as largely a Hehe and Kosisamba hamlet,
millet was farmed in the 'korongo'. About 30
households before the 1949 famine, thereafter, the
settlement declined to about 12 households. Cattle
were noted as being formerly kept by the
inhabitants.
30 7 48 91 74 Ilunda Prior to the 1949 foloboto famine, a hamlet of about

10 households mostly consisting of Kosisamba but
some Hehe. Thereafter it largely fell derelict. Maize
was farmed in riverine shambas with millet on the
upland areas. There also were a 'substantial' number
of cattle prior to 1949 but not thereafter. One old
mzee, Chambilele Chalo, described as hanging on
there right up until the 1974 Ujamaa uhamisho.
31 7 55 91 85 Mugaga Incorporated as part of the new game reserve in

1955, people moved to Mlowa and Pawaga. Maize was
farmed on the riverine inlet and millet on the
upland. Consisting of about 10 Kosisamba households
when it was abandoned, the hamlet suffered a long
term decline in its stock holdings - described as
arising from the combined effect of tsetse incursion
and re-occurring drought. Nevertheless, 200 head of
stock estimated in 1949. Homesteads were spread
out towards Nyamapalula and Nytalele.
248
Settlement
Number
32 7 56 91 86 Nytalele A small Kosisamba settlement of 4 households, On

the boundary between the Kayela and Ilolo
jumbeates. Largely an agricultural community,
people farmed maize along the river inlet (Ruaha)
and kept livestock. Evicted in 1955 - Msengesi moved
to Ilolo and Mkomola to Ikolongo.
33 7 56 91 87 Ihwani Not included in the 1955 gazettement of the Ruaha

Game Reserve, nevertheless the settlement declined
largely as a result of the 1949 famine when many
families moved away to Ilolo. Hamlet's cattle
declined and were replaced by goats. Patrilineages
remembered include Muhegele (Kosisamba), Wasasa
(Sangu) and Ndondole. Post 1949 the settlement is
described as consisting of about 20 households.
34 7 53 91 92 Makutupa A settlement that post 1949 was gradually abandoned

with people moving to Mpangile due to a shortage of
potable water - either having to be carried from the
Ruaha River or fetched from seasonal wells dug in the
River Igembe during the dry seaon. The fields lay in a
nearby area called Ilale. Unlike many other hamlets,
very few people are described as leaving during the
1949 famine as they managed to survive on famine
foods as well as food aid carried in by foot from
Kimande. Cattle also described as having helped food
security considerably. The Ilale shambas were
primarily cultivated for millet together with peanuts
but with very little maize. Pre 1949, the hamlet is
described as consisting of about 30 households, and
notable patrilineages are remembered as being:
Ngiliama, Wasasa and Waliongole. These families
were described as being of recent (one generation)
origin from the (U)sangu - old men of the 1950s
having been born in the Usangu.
35 7 59 91 92 Ilolo The hamlet from which the Ilolo jumbeate was

administered by (at one point) Jumbe 'Chai' (sic). In
1964, many surrounding hamlets were compelled to
move to Ilolo in an effort to provide better services.
Thus Ilolo became a larger more eclectic collection
of people. A school and hospital (clinic) were built,
the former at Talangwe. Up to 1 in 3 households are
described as having kept livestock. Millet was the
major crop and even eucalypti are reported as having
been planted at one point. The whole settlement was
moved in 1974 and taken over by Ruaha National
Park. Formerly, a mixture of Kosisamba (eg Mbwati)
and Hehe (eg Kayugwa, Ndindile).
36 7 59 91 95 Talangwe A small hamlet to the north of Ilolo consisting of

about 6 households, which farmed finger millet and
kept some cattle. Nyutupa (Kosisamba) Mtemle
(Hehe).
37 7 56 91 96 Udindamisi Close to the River Igembe, about 30 households

which farmed finger millet and held 'substantial'
numbers of livestock. A mixture of peoples eg
Kayugwa and Nyamoga (Kosisamba); Msingalata
(Safwa) and; Chavala (Hehe).
38 7 63 91 96 Ny'angai A hamlet of about 30 households where finger millet

was farmed together with some cattle. Abandoned in
the 1974 uhamisho. Perhaps largely Gogo eg Nzinzile
249
Settlement
Number
39 7 69 91 96 Mkombilenga A Gogo settlement which is reported as growing some

finger millet but relying predominantly on hunting
using spears and dogs, although some bow/arrow
hunting too. The grasslands in the Mpangile area
were hunted in. Livestock were also kept. NFD
40 7 71 91 95 Unynyamala Described as a settlement in which many livestock

were kept pre1949, subsequently the herd declined.
During the 1949 famine, the inhabitants moved to
hamlets further south but returned in 1961. Maize
was grown along the river banks with finger millet
planted in fields further upland. Largely a Kosisamba
settlement but some Hehe too. Patrilineages
remembered include: Kinula, Bamba and Ngendo (all
Kosisamba).
41 7 50 91 95 Mpangile Situated on the Igembe River, received drought

refugees from Makutupa. Water supply described as
being relatively predictable as dry season river bed
wells yielded sufficient water. The settlement was
spread along both the north and south banks and was
about 50 households in all The settlement was a
mixture of Kosisamba - Chai and Muhegele - and
Sangu - Wasasa and Wangaza
42 7 14 91 62 Unylinga A very small hamlet which was abandoned before

1949. The only artefacts remaining are grind stones.
NFD
43 7 7 91 59 Ibagi Described as a very small hamlet which fell derelict

sometime before 1949 (possibly pre 1940). Riverine
maize and upland millet grown. Cattle were also
apparently kept. Inhabitants all moved to either
Makuluga or Igawa. Remembered as a Kosisamba
settlement eg Mgesi
44 7 19 91 55 Mtumbulikwakwa Fell derelict sometime pre 1940 but previously the

hamlet had grown riverine maize, some millet and
herded some cattle. Kosisamba - Chaka and
Mbwikuwangi. NFD
45a 7 18 91 52 Nyamakala An area with large tamarind and Misagati trees on the
upland (Nyamakala) and fields on the river plain
45b 7 19 91 52 Fihwawi below. There were also fields at Fihwawi on the river
banks. People came from Kiganga to farm these
46 7 17 91 51 Kiganga A large village in which Jumbe Kayela (d.~1940) is
buried. An area of maize farming with previously, a
large number of cattle latterly replaced by goats. At
its largest, perhaps 30 households all told. Other
patrilineages remembered Mwandole (Kimbu),
Kadelega and Nyongo (Kosisamba).
47 7 13 91 48 Matinga A hamlet which overlooked a grassland and which

depended largely on riverine maize fields and goats -
perhaps around 15 households. Mbugeni (Kosisamba)
and Kihwele (Hehe) both from Msembe. NFD
48 7 13 91 50 Msembe Formerly quite a large hamlet of about 30

households, riverine maize was grown together with
some millet. The settlement was evicted in 1955.
250
Settlement
Number
49 7 2 91 50 Mdonya A largely Hehe settlement with some Sangu and

NB The location Gogo. Settled by a substantial number (relatively) of
people from highland Uhehe who were attracted by
reports of easy hunting and good harvests. Described
as being heavily wooded in the 1940s/1950s
(indicating a local absence of elephant?) but being
infested with tsetse where not cleared for agriculture
and therefore precluding livestock keeping except
that of goats. The south bank of the Ruaha was much
more open and suitable for livestock keeping eg
Kiganga, Igawa and Makuluga. The north bank was
wetter with more predictable rainfall than the south.
A system of trade between the north and south bank
villages for livestock for the former and grain for the
latter - especially during years of drought. Also a
trade with the highlands - dried (game/fish -
increasingly?) meat for highland grain. Agriculture
consisted of mainly maize and peanut farming with
small amounts of banana and cassava. The
settlement was evicted in 1956, a year after that of
Njongomeru. Names of families remembered include:
Madomi, Ngaila, Mukwaga and Kabonyela.
50 7 22 90 39 Nyamtupa The area now known as 'Mudweka' lying to the south

of the Tungamalenga-Msembe road. Previously
unsettled, a number of Bena families moved there
from Makifu in the early 1980s (?) to farm but
subsequently evicted by TANAPA in the late 1980s.
51 7 14 90 21 Nyaluhanga A hamlet that primarily relied on honey gathering but

farmed peanuts and some maize. Was moved to
Makifu in the Uhamisho of 1974. Patrilineages
remembered are: Makoga, Kahomba and Mpulule.
52 7 12 90 11 Waga A hamlet that received substantial number of

evictees from Mdonya post 1955, it's economy was
based on maize and peanut farming, together with
seasonal fishing. Rice was initially farmed in 1971
opportunistically. The settlement was evicted in 1974
but by 1980 had been successfully re-established.
(Reputedly a centre of witchcraft - more info
required and the guardian of the tambiko site at
Mdonya, an elderly woman, resides here)
53 6 98 90 21 Kitagasa The site of a large valley reputedly excellent for

farming and rice was at one stage grown here. The
economy was largely dependent on maize farming
with peanuts and honey grown as something akin to
cash crops. To the north were the forests of
Nyamkola and Nyamgungu.
54 6 94 90 20 Sanzala Economy based on honey collecting, fishing and

peanut growing. The hamlet was relatively little
impacted in terms of in-migration from Mdonya.
Families remembered are: Mbeti (Kosisamba) Mtulu
(Gogo) and Njiwa (Sangu).
55 6 93 90 33 Kinyangesi The boundary point between Iringa and Mbeya

Regions, lying on the Ruaha River. Kinyangesi itself
paid taxes to Kiponzelo and was subject to
substantial in-migration post 1955 from Game
Reserve evictees. During the uhamisho, people
largely moved to Tungamalenga and Makifu with
some going to Madibira.
251
Settlement
Number
56 6 73 90 23 Mkupule A hamlet that was much enlarged post 1955 with

evictees. A similar economy to the surrounding
hamlets of fishing, honey gathering and peanut and
maize growing. During the uhamisho, evictees moved
to Mapogoro and others to Madibira. Families
included Nyahulo (Kimbu), Mwaiyege, Mjola and
Ngolotima (Kosisamba) the latter of who lived in Ilolo
a little hamlet further downstream.
NB: Grid refs estimated (About a 10km east-west up- An extended area of settlements of up to about 400
stream spread of settlements) largely Sangu and some Kimbu people (guesstimate)
6 73 90 27 stretching along the banks of the river flowing from
57a Mahango (11)
the west. Small shambas farmed but no livestock due
57b Muvilinge (7) to the presence of tsetse fly. A limited number of
57c Ufikiro (9) specialist hunters who owed their own weaponary
57d Wimbwa (Njongomeru) (15) and hunted as a major part of their livelihood.
57e Kibidimilla (22) Others, less specialist, hunted when the need arose -
largely thru pit-hunting. There was some trade with
57f Ihombelwa (6)
other villages. People were evicted in 1955 (after the
rains when a vehicle could get thru), and the
settlements burnt. Many elected to move to the
57g 6 73 89 27 Masapi (13) Usangu but others moved to the Mkupule hamlets. A
rough indication of the size of each hamlet in the
late 1940s/early 1950s) is given in adjacent brackets.
58 6 77 90 19 Mpama A hamlet established by evictees post 1955, primarily
consisting of one family (Lungwa - Sangu) which
moved to Tungamalenga in 1974.
59 6 86 90 20 Lwani Many people described as arriving here after 1955

from Mdonya. Major families remembered are
Mwambete, Mbepwa and Nganylika (all Kosisamba)
60 7 15 90 14 Wota Primarily a farming area with some hunting and

fishing carried out. The settlement was moved in
1974 people going to Mahuninga and Nyamakuyu
(Mbeya). People passing through this area post
eviction noticed that tsetse fly densities had
increased markedly. Key families remembered are
Lukova and Lyangoya.
61 7 16 90 24 Mudweka This hamlet was situated next to what was

considered an extremely fertile cultivation bonde.
However, due to the relatively low numbers of
people living in this settlement, in 1970/71 they
decided to move to Makifu before the 1974
uhamisho. Families remembered include Chafuwawi,
Malagi (Sangu) and Kayoka (Bena).
252
Settlement
Number
The following settlements have been taken from a colonial regional map and remain to be followed up and their
details investigated.
62 - Mpanga
63 - Lungwa
64 - Igula
65 - Kangaro
66 - Mandwa
67 - Mtovelisusi
68 - Makutupa
69 - Ilusi (N)
70 - Mukombe
71 - Itunundu
72 - Igohungula
73 - Mbagi
75 - Kimande
76 - Ndolela
77 - Mlengi
78 - Chimamba
79 - Magangamatitu
80 - Magoya
81 - Sasamambo
82 - Mloa
83 - Msimbi
84 - Idodi
85 - Nyangano
Further details are available for these settlements, but for the moment are not provided:
86 - Kitisi
87 - Mapogoro
88 - Kidangwe
89 - Tungamalenga
90 - Isanga
91 - Makifu
92 - Ikwavila
93 - Nykapembe
94 - Mwitikira
253
A4: The reported incidence of disease in the Idodi livestock herd
The data below are provided to illustrate the relative incidence of disease that is reported to have
occurred in the past in the Idodi livestock herd.
Table A3: The reported incidence of different diseases in the Idodi livestock herd in 1991
[Source: Annual veterinary report to Divisional Secretary 1991 File RA/LDC/MLOWA/01]
Percent infected with

Number
Checked East Coast
Trypanosomaisis Liver Fluke Brucellosis
Fever
Cows 5,678 37.0% 0.5% 30.9% 1.8%
Goats 3,400 - - 11.9% -
Sheep 2963 - - 13.2% -
Note: These data were collected by the divisional veterinary officer based in Malinzanga village for Idodi
Division.
254
A5: Age grades and age sets of the Ilparakuyo in Idodi
The following age-sets are still in existence in Idodi. Although the age-sets correspond to the northern
sections of Maasai, the opening and closing of age sets and divisions is delayed by a varying number
of years.
Age Grade Age Set Name Date Opened/Closed Division Date Opened/Closed
Junior Moran 2002 - Ikelimboti 2002 -
Senior Moran Ilandisi 1987 - 2000 Ikelimboti 1994 - 2000

(Emuatua e tateene) Ipalingotwa 1991 - 1993
Inkereyani 1987 - 1990
Elders Ildaleto 1972 - 1987
Senior Elders Ilmedoti 1956 - 1974/5
Retired Elders Ilkiduto 1940 - 1956
255
Bibliography
Adams, J. and T. McShane 1992. The Myth of Wild Africa: Conservation without Illusion. New York:
W.W. Norton.
Agrawal, A. 2001 Common property institutions and sustainable governance of resources. World
Development 29 (10):1649-1672.
Agrawal, A. 2003. Sustainable governance of common pool resources. Annual Review of Anthropology
32: 243-262.
Agrawal, A. and C. Gibson 1999. Enchantment and disenchantment: the role of community in natural
resource conservation. World Development 27 (4): 629-49.
Agrawal, A. and G. N. Yadama 1997. How do local institutions mediate market and population
pressures on resources? Forest Panchayats in Kumaon, India. Development and Change 28: 435-
465.
Amanor, K. S. 1999. Global Restructuring and Land Rights In Ghana: Forest Food Chains, Timber and
Rural Livelihoods. Research Report No. 108, Uppsala: Nordiska Afrikainstitutet.
Amanor, K. S. 2001. Land, Labour and the Family in Southern Ghana. Research Report No. 116,
Uppsala: Nordiska Afrikainstitutet.
Anderson, D. A. 1984. Depression, dust-bowl, demography and drought: the colonial state and soil
conservation in East Africa during the 1930s. African Affairs 83 (332): 321-343.
Anderson, D. and R. Grove 1987. Conservation in Africa: People, Policies and Practice. Cambridge:
Cambridge University Press.
Arhem, K. 1985a. The Maasai and the State: The Impact of Rural Development Policies on a Pastoral
People in Tanzania. IWGIA Document.
Attwood, D. A. 1990. Land degradation in Africa: the impact of agricultural production. World
Development 21 (4): 663-639.
Baldus, R.D. and A.E. Cauldwell 2004. Tourist hunting and its role in the development of wildlife
management areas in Tanzania. Paper presented to: Sixth International Game Ranching
Symposium, Paris July 6th to 9th, 2004
Barrows, R. and M. Roth 1990. Land tenure and investment in African agriculture: Theory and
evidence. Journal of Modern African Studies 28 (2): 265-297.
Bassett, T. J. 1993 Introduction: The land question and agricultural transformation in sub-Saharan
Africa. In: Land in African Agrarian Systems, T. J. Bassett and D. Crummey, eds. Pp. 3-34.
Madison: University of Wisconsin.
Bassett, T. J. and D. E. Crummey eds. 1993. Land in African Agrarian Systems. Madison: University of
Wisconsin Press.
Bassett, T. J. and K. B. Zuéli 2000. Environmental discourses and the Ivorian savanna. Annals of the
Association of American Geographers 90 (1): 67-95.
Baxter, P. and R. Hogg, eds. 1990. Property, Poverty, and People: Changing Rights in Property and
Problems of Pastoral Development. Manchester: University of Manchester, Department of Social
Anthropology and International Development Centre.
Behnke, R. H. 1992 New Directions in African Range Management Policy. London: Overseas
Development Institute.
Behnke, R. H. Jr, I. Scoones, and C. Kerven, eds, 1993. Range Ecology at Disequilibrium: New Models
of Natural Variability and Pastoral Adaptation in African Savannas. London: Overseas
Development Initiative.
Beidelman, T. O. 1960. The Baraguyu: Introduction. Tanganyika Notes and Records 55: 245-278.
Beidelman, T. O. 1968. Some Hypotheses Regarding Nilo-Hamitic Symbolism and Social Structure:
Baraguyu Folklore. Anthropological Quarterly 41 (2): 78-89.
256
Beinart, W. 1994. Soil erosion, conservationism and ideas about development: a southern African
exploration 1900-1963. Journal of Southern African Studies 11: 52-83.
Beinart, W. 1996. Soil Erosion, Animals and Pasture over the Longer Term. In: The Lie of the Land:
Challenging Received Wisdom on the African Environment, M. Leach and R. Mearns, eds. Pp. 54-
72. London: The International African Institute in Association with James Currey and Heinmann.
Beinart, W. 2000. African history and environmental history. African Affairs 99: 269-302.
Beinart, W. and J. McGregor 2003. Introduction. In: Social history and African environments. W.
Beinart and J. McGregor eds. Pp. 1-24. Oxford: James Currey.
Bekure, S., P. N. de Leeuw, B. E. Grandin and P. J. H. Neate 1991. Maasai Herding. An Analysis of the
Livestock Production System of Maasai in eastern Kajiado District, Kenya. ILCA Systems Study 4.
Addis Ababa: International Livestock Centre for Africa.
Berry, S. 1989. Social institutions and access to resources in Africa. Africa 59 (1): 41-55.
Berry, S. 2002 Debating the land question in Africa. Comparative Studies in Society and History 44
(4): 638-668.
Besteman, C. 1994. Individualisation and the assault on customary tenure in Africa: title registration
programmes and the case of Somalia. Africa 64 (4): 484-515.
Biebuyck, D.1963. Introduction. In: African Agrarian Systems, D Biebuyck, ed. Oxford: Oxford
University Press for the International African Institute.
Bigsten, A. and A. Danielson 2001. Tanzania: Is the Ugly Duckling Finally Growing Up? Report No 120,
Uppsala: Nordiska Afrikainstitutet Research.
Biodiversity Support Programme 1993. African Biodiversity: Foundation for the Future. Washington
DC: Wildlife Fund, the Nature Conservancy and World Resources Institute with the US Agency
for International Development.
Birley, M. H. 1982. Resource Management in Sukumaland. Africa 52 (2): 3-29.
Bjørnstad, A. 1976. The vegetation of Ruaha National Park Tanzania: 1. Annotated checklist of the
plant species. Serengeti Research Institute Publication No. 215
Blaike, P. 1995. Understanding environmental issues. In: People and Environment, S. Morse and M.
Stocking, eds. Pp. 1-30. Norwich: School of Development Studies, University of East Anglia.
Blaikie, P. and H. Brookfield 1987. Land Degradation and Society. London: Methuen.
Bonner, R. 1993. At the hand of man: peril and hope for Africa’s wildlife. New York: Vintage Books.
Bourdieu, P. 1977. Outline of a Theory of Practice. (Trans. R. Nice) Cambridge: Cambridge University
Press.
Bourn D, R. B. and R. Blench 1999. Can Livestock and Wildlife Co-exist? An Interdisciplinary Approach.
Livestock, Wildlife and People in the Semi-arid Rangeland of East Africa. London: Overseas
Boserup, E. 1965. The Conditions of Agricultural Growth: The Economics of Agrarian Change under
Population Pressure. London: Allen & Unwin.
Bradbury et al. 1994 Working With Pastoralist NGOs and Land Conflicts in Tanzania: A Report on a
Workshop in Terrat, Tanzania. 11-15th December, 1994. Pastoral Land Tenure Series No. 7.
Brandström, P. 1985. The Agro-pastoral Dilemma: Underutilisation or Overexploitation of Land Among
the Sukuma of Tanzania. Uppsala, African Studies Programme, Department of Cultural
Anthropology, University of Uppsala. Working papers in African studies, University of
Uppsala. African Studies Programme, No.8.
Brock, K., A. Cornwall, and J. Gaventa 2001. Power, knowledge and political spaces in the framing
of poverty policy’, IDS Working Paper 143. Brighton: Institute of Development Studies. Online
at: http://www.ids.ac.uk/ids/bookshop/wp/wp143.pdf/
Brockington, D. 1998. Land Loss and Livelihoods. The Effects of Eviction on Pastoralists Moved from
Mkomazi Game Reserve, Tanzania. PhD Thesis, University College London.
Brockington, D., 2002. Fortress Conservation. The Preservation of the Mkomazi Game Reserve. Oxford:
James Currey, African Issues Series.
257
Brockington, D. 2003. The Contingency of Community Conservation. In: Rural Resources and Local
Livelihoods, K. M. Homewood, ed. Oxford: James Currey.
Brockington, D. 2004. Communal property and degradation narratives: debating the Sukuma
immigration into Rukwa Region, Tanzania. Les Cahiers No. 20: 1-22.
Brockington D, Homewood K. 1996. Wildlife, pastoralists and science. Debates concerning Mkomazi
game reserve, Tanzania. In: The Lie of the Land. Challenging Received Wisdom in African
Environmental Change and Policy, M. Leach and R. Mearns, eds. Pp. 91-104. International
African Institute in association with James Currey and Heinmann.
Brockington, D. and K. M. Homewood 2001. Degradation debates and data deficiencies: The
Mkomazi Game Reserve, Tanzania. Africa 71 (3): 449-480.
Bromley, D. and M. Cernea 1989. The Management of Common Property Natural Resources: Some
Conceptual and Operational Fallacies. World Bank Discussion Paper No. 57.
Brown, G. G. and A. M. Bruce-Hutt 1935. Anthropology in Action: An Experiment in the Iringa District
of the Iringa Province, Tanganyika Territory. Oxford: Oxford University Press.
Brosius, P. 1997. Analysis and interventions: anthropological engagements with environmentalism.
Manuscript.
Brown, C.W. ed. n.d. Contested domains: political ecology and social practice. Manuscript.
Bruce, J. W. 1986. Land tenure issues in project design and strategies for agricultural development in
sub-Saharan Africa. Land tenure Centre Paper 128. Madison: University of Wisconsin.
Bruce, J. W. 1994. and Migot-Adholla, S. E. eds. 1994. Searching for Land Tenure and Security in
Africa. Dubuque IO: Kendall/Hunt Publishing Co..
Bryceson, D. F. 2002. Multiplex livelihoods in rural Africa: Recasting the terms and conditions of
gainful employment. Journal of Modern African Studies 40 (1):1-28
Bryceson, D. F. and L. Bank, 2001. End of an era: Africa’s development policy parallax. Journal of
Contemporary African Studies 19 (1): 5-23.
Campbell, B., A. Mandondo, N. S. B. Nemarundwe, W. de Jong, M. Luckert, and F. Matose 2001.
Challenges to proponents common property resource systems: Despairing voices from the social
forests of Zimbabwe. World Development 29 (4): 589-600.
Carney, D. and J. Farington 1998. Natural Resource Management and Institutional Change. London:
Overseas Development Institute and Routledge.
Chanock, M. 1991. Paradigms, policies, and property: a review of the customary law of land tenure.
In: Law in colonial Africa. K. Mann and R. Roberts eds. Pp. 61-84. Portsmouth, NH: Heinemann.
Charnley, S. 1994. Cattle, Commons and Culture: The Political Ecology of Environmental Change on a
Tanzanian Rangeland. PhD Thesis, University of Stanford.
Charnley, S. 1997. Environmentally-displaced peoples and the cascade effect: Lessons from
Tanzania. Human Ecology 25 (4): 593-618.
Cleaver, F. 2000. Moral ecological rationality, insitutions and the management of common property
resources. Development and Change 31: 361-383.
Cleaver, F. 2003. Reinventing institutions: bricolage and the social embeddedness of natural resource
management. In: Securing Land Rights in Africa. T.A. Benjaminsen and C. Lund. Pp. 11-31.
London: Frank Cass.
Cleaver, F. and Franks, T. 2003a. How institutions elude design: river basin management and
sustainable livelihoods. Paper presented to: The Alternative Water Forum. Bradford Centre for
International Development, University of Bradford, 1-2 May. Online at:
http://www.bradford.ac.uk/acad/bcid/seminar/alternative_water/ papers/
Cliffe, L. and G.L. Cunningham 1973. Ideology organisation and the settlement experience in
Tanzania. In: Socialism in Tanzania 2. L. Cliffe and J. Saul eds. Nairobi: East Africa Publishing
House.
Coast, E. 2002. Maasai Socioeconomic Conditions: A cross border comparison. Human Ecology 30 (1):
79-105.
Coldham, S.F.R., 1978. The effect of registration of title upon customary land rights in Kenya. Journal
of African Law 22: 91-111.
258
Cole, J. S. R. and W.N. Denison 1964. Tanganyika: The Development of its Laws and Constitution.
London: Stevens.
Colson, E. 1971. The impact of the colonial period on the definition of land rights. In: Colonialism in
Africa. V. Turner ed. Pp. 193-215. Cambridge: Cambridge University Press.
Coulson, A. 1982. Tanzania: a Political Economy. Oxford: Oxford University Press.
Cousins, B. 1993 A political economy model of common property regimes and the case of grazing
management in Zimbabwe. ODI Pastoral Development Network Paper 34b. London: Overseas
Creighton, C. and C.K. Omari eds. 1995. Gender, Family and Household in Tanzania. Aldershot:
Avebury.
Davis, S., V. Heywood and A. Hamilton eds. 1994. Centres of plant diversity. WWF/IUCN, Gland,
Switzerland.
Davison, J. ed. 1988. Agriculture, Women and the Land: The African Experience. Boulder CO:
Westview Press.
Demeritt, D. 2002. What is the 'social construction of nature'? A typology and sympathetic critique.
Progress in Human Geography 26 (6): 767-790.
Douglas, M. 1987. How Institutions Think. London: Routledge and Kegan Paul.
Dublin, H., 1995. Vegetation dynamics in the Serengeti-Mara ecosystem: the role of elephants, fire
and other factors. In: Serengeti II: Dynamics, Management and Conservation of an Ecosystem. A.
Sinclair and P. Arcese eds. Pp. 71-90. Chicago: Chicago University Press.
Drinkwater, M. 1992. Cows eat grass don’t they? Evaluating conflict over pastoral management in
Zimbabwe. In: Bush Base: Forest Farm: Culture, Environment and Development. Croll, E. and D.
Parkin eds. Pp.169-86. London: Routledge.
Ellis, F. 1998. Household strategies and rural livelihood diversification. Journal of Rural Development
Studies 35 (1): 1-38.
Ellis, J. and D. Swift 1988. Stability of African pastoral ecosystems: alternate paradigms and
implications for development. Journal of Range Management 41 (6): 450-459.
Epstein, A. L. 1967. The Case Method in the Field of Law. In: The Craft of Social Anthropology, A. L.
Epstein, ed. Pp. 205-230. London: Tavistock.
Escobar, A. 1996. Construction nature: Elements for a post-structuralist political ecology. Futures 28
(4): 325-343.
Escobar, A. 1999. After nature: Steps to an anti-essentialist political ecology. Current Anthropology
40 (1): 1-30.
Fairhead, J. and M. Leach 1996. Misreading the African Landscape: Society and Ecology in a Forest-
Savanna Mosaic. Cambridge: Cambridge University Press.
FAO 1984. Land, Food and People. FAO Economic and Social Development Series 30. Rome: Food
and Agricultural Organisation of the United Nations.
Ford, J. 1971. The Role of Trypanosomiases in African Ecology: A Study of the Tsetse Fly Problem.
Oxford: Clarendon Press.
Ford, J. and K.M. Katondo 1977. The distribution of tsetse flies in Africa. OAU, Nairobi: Cook,
Hammond & Kell.
Fortes, M. 1971. Introduction. In: The Development Cycle in Domestic Groups. J. Goody ed. Pp. 1-14.
Cambridge: Cambridge University Press.
Fosbrooke, H. A. 1948. An administrative survey of the Maasai social system. Tanganyika Notes and
Records 26: 1-50.
Fosbrooke, H. A. 1956. The Masai age group system as a guide to tribal chronology. African Studies
15 (4): 188-206.
Foucault, M. 1971. The order of discourse. In: Untying the Text: A Poststructuralist Reader. London:
Routledge and Kegan Paul.
Foucault, M. 1980. Power/Knowledge: Selected Interviews and Other Writings 1972-1977. C. Gordon
ed. Brighton: Harvester Press.
259
Franke, R. and B. Chasin. 1980. Seeds of Famine: Ecological Destruction and the Development Dilemma
in the West African Sahel. Tototowa NJ: Rowman & Allenheld.
Fratkin, E., K. A. Galvin, and E. A. Roth 1994. African Pastoralist Systems: An Integrated Approach.
London & Boulder: Lynne Rienner.
Funtowicz, S. O. and J. R. Ravetz 1992. Three types of risk assessment and the emergence of post-
normal science. In: Social Theories of Risk. S. Krimsky and D. Golding eds. Pp. 251-73.
Westport CT: Praeger.
Galaty J.G. 1981. Land and livestock among Kenyan Maasai: symbolic perspectives on pastoral
exchange, change and inequality. Journal of Asian and African Studies XVI: 1-2.
Galaty J.G. 1988. Pastoral and agro-pastoral migration in Tanzania: factors of economy, ecology
and demography in cultural perspective. In: Production and Autonomy: Anthropological Studies
and Critiques of Development, J. W. Bennett and J. R. Bowen eds. Pp. 163-184.Lanham:
University Press of America.
Galaty, J. G. 1993. Maasai expansion and the new East African pastoralism. In: Being Maasai:
Ethnicity and Identity in East Africa. T. Spear and R. Waller eds. Pp. 61-86. London: James
Currey.
Gandy, M. 1996. Crumbling land: the postmodernity debate and the analysis of environment
problems. Progress in Human Geography 20 (1): 23-40.
Ghai, D. 1992. Conservation, Livelihood and Democracy: Social Dynamics of Environmental Changes in
Africa. UNRISD Discussion Paper 33. Geneva: United nations Research Institute Social
Development.
Gibson, C. C. 1999. Politicians and Poachers: The Political Economy of Wildlife Policy in Africa.
Giddens, A. 1984. The Constitution of Society: Outline of the Theory of Structuration. Cambridge:
Polity Press.
Giddens, A. 1989. Sociology. Cambridge: Polity Press.
Gilson, L, M Sheridan, and D. Brockington 2003. Representing environments in flux: case studies from
East Africa. Area 35 (4): 371-389.
Gluckman, M. 1969. Ideas and Procedures in African Customary Law: Studies Presented and Discussed.
Addis Ababa. 8th International Africa Seminar. 19-6-1974.
Goheen M. and Shipton, P. 1992 Understanding African land-holding: power, wealth and meaning.
Africa 62 (3): 307-326.
Goldman, M. ed. 1998. Privatizing Nature: Political Struggles for the Global Commons. New
Brunswick, NJ: Rutgers University Press.
Granovetter, D. 1992. Economic action and social structure: the problem of embeddedness. In: The
Sociology of Economic Life. M. Granovetter and R. Swedburg eds. Pp. 53-81. Boulder CO and
Oxford: Westview Press.
Green, J.K. 1987. Evaluating the impact of consolidation of holdings, individualisation of tenure, and
registration of title: lessons from Kenya. Land Tenure Centre Paper 129. Madison: University of
Wisconsin.
Grove, R. 1987. Early themes in African conservation: the Cape in the nineteenth century. In:
Conservation in Africa: People, Policies and Practice. D. Anderson and R. Grove Pp. 21-40.
Gulliver, P. H. 1958. Land Tenure and Social Change among the Nyakusa. London: Routledge and
Kegan Paul.
Gulliver, P. H. 1961. Land shortage, social change and social conflict in East Africa. Journal of Conflict
Resolution 5:16-26.
Gulliver, P. H. 1963. Social Control in an African Society: A Study of the Arusha, Agricultural Maasai of
Northern Tanganyika. London: Routledge and Kegan Paul.
Gulliver, P. H. 1979 Disputes and Negotiations. London: Routledge and Kegan Paul.
Hagberg, S. 1998. Between Peace and Justice. Dispute settlement between Karaboro Agriculturalists and
Fulbe Agropastoralists in Burkina Faso. Uppsala: Acta Universitatis Upsaliensis.
260
Halimshauri ya Wilaya ya Iringa. 2000. Taarifa ya Hali Halisi ya Wilaya Hadi Machi 2000. Wizara
ya Tawala za Mikoa na Serikali za Mitaa, Jamuhuri ya Muuangano ya Tanzania.
Hall, H.T.B. 1977. Diseases and Parasites of Livestock in the Tropics. London: Longman
Hankins, T.D. 1974. Response to drought in Sukumaland, Tanzania. In: Natural Hazards: Local,
National Global. G.F. White ed. London: Oxford University Press.
Hardin, G. 1968. The tragedy of the commons. Science 162: 1243-1248.
Haugerud, A.1989. Land tenure and agrarian change in Kenya. Africa 59 (1): 61-90.
Havnevik, K. J. 1993. Tanzania. The Limits to Development From Above. Motala: Motala Grafiska AB.
Hjort af Ornäs, A. and M. Salih eds. 1989. Ecology and politics: environmental stress and security in
Africa. Uppsala: Nordiska Afrikainstitutet.
Hoben, A. 1996. The cultural construction of environmental policy: paradigms and politics in Ethiopia.
In: In: The Lie of the Land. Challenging Received Wisdom in African Environmental Change and
Policy, M. Leach and R. Mearns, eds. Pp. 186-208. International African Institute in association
with James Currey and Heinmann.
Hodgson, D. ed. 2000. Rethinking Pastoralism in Africa: Gender, Culture and the Myth of the Patriarchal
Pastoralist. Oxford: James Currey
Hodgson, D. L. 2001. Once Intrepid Warriors: Gender, Ethnicity and the Cultural Politics of Maasai
Development. Indiana: Indiana University Press.
Homewood, K. 2004. Policy, environment and development in African rangelands. Environmental
Science and Policy 7: 125-143.
Homewood, K., E.F. Lambin, E. Coast, A. Kariuki, I. Kikula, J. Kivelia, M. Said, S. Serneels, M. Thompson
2001. Long-term changes in Serengeti-Mara wildebeest and land cover: pastoralism,
population or policies? Proceedings of the National Academy of Sciences. U.S.A. 98 (22),
12544-12549.
Homewood, K. M. and A. Rodgers 1987. Pastoralism, conservation and the overgrazing controversy.
In: Conservation in Africa: People, Policies and Practice, D. Anderson and R. Grove, eds. Pp.
111-128. Cambridge: Cambridge University Press.
Homewood, K. M. and A. Rodgers 1991. Maasailand Ecology: Pastoralist Development and Wildlife
Conservation in Ngorongoro. Cambridge: Cambridge University Press.
Homewood, K. M. and D. Brockington 1999. Biodiversity, conservation and development in Mkomazi
Game reserve, Tanzania. Global Ecology and Biogeography 8: 301-313.
Horowitz, M. M. 1986. Ideology, policy, and praxis in pastoral livestock development. In: The
Anthropology of Rural Development in West Africa. eds. M.M. Horowitz and T. Painter Pp. 251-
72. Boulder CO: Westview Press.
Huntingford, G. W. 1953. The Southern Nilo-Hamitics, East-Central Africa. Ethnographic Survey of
Africa, Part VIII. London: International African Institute
Hussein, K. 1998. Conflict between farmers and herders in the semi-arid Sahel and East Africa: A
review. IIED Pastoral Land Tenure Series No. 10.
Hyden, G. 1980. Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry.
London : Heinemann Educational
IIED 1994. Whose Eden? An Overview of Community Approaches to Wildlife Management. London:
International Institute for Environment and Development.
ILCA 1981. Introduction to the East African Range Livestock Study, International Livestock Centre for
Africa (ILCA)/Republic of Kenya, Nairobi.
Iliffe, J. 1979. A Modern History of Tanganyika. Cambridge: Cambridge University Press.
International Bank of Reconstruction and Development 1961. The Economic Development of
Tanganyika. Baltimore: John Hopkins Press.
Jacobs, A. H. 1965. Traditional Political Organisation of the Pastoral Maasai. PhD Thesis, University of
Oxford.
Jacobs, A. H. 1980. Pastoral development in Tanzanian Maasailand. Rural Africana 7: 1-14.
261
Jambiya, G. 1998. The dynamics of population, land scarcity, agriculture and non-agricultural
activities: West Usambara Mountains, Lushoto District, Tanzania. Working Paper, Vol. 28. Dar
es Salaam: Institute of Resource Assessment and Leiden: African Studies Center.
James, R.W. and G.M. Fimbo 1973. Customary Land Law of Tanzania. A Source Book. Nairobi: East
African Literature Bureau.
Jarosz, I. 1996. Defining deforestation in Madagascar. In: Liberation ecologies: environment,
development, social movements. R. Peet and M. Watts eds. Pp. 248-64. London: Routledge.
Jennings, S. L. 1994. Village Oral Histories Mini-study. Msembe, Ruaha Ecosystem and Wildlife
Management Project, ODA.
Johnston, C. 2004. Uncommon ground: The 'poverty of history' in common property discourse.
Development and Change 35 (3): 407-433.
Keeley, J. and Scoones, I. 1999. Understanding environmental policy processes: a review. IDS
Working Paper 89. Brighton: Institute of Development Studies. Online at:
http://www.ids.ac.uk/ids/bookshop/wp/wp89.pdf/
Keenja, C.N. 2003. Hotuba ya Waziri wa Kilimo na Chakula Mheshimiwa Charles N. Keenja (Mb.)
kuhusu Makadirio ya Matumizi ya Fedha Kwa Mwaka Wa 2003/2004. Serikali ya Muungano
wa Tanzania. Online at: http://www.agriculture.go.tz/About MAFS/Ministers and
Speeches/Budget Speeches/BSpeech Kilimo 2003-2004.pdf
Kelsall, T. 2004. Contentious Politics, Local Governance and the Self: A Tanzanian Case Study. Report
No. 129, Uppsala: Nordiska Afrikainstitutet Research.
Keohane, R. O. and E. Ostrom 1994. Local Commons and Global Interdependence: Heterogeneity and
Cooperation in Two Domains. London: Sage.
King, J. M., A. R. Sayers, C. P. Peacock, and E. Kontrohur 1984. Maasai herd and flock structures in
relation to livestock wealth, climate and development. Agricultural Systems 13: 21-56.
Kjekshus, H. 1977, 1995. Ecology, Control and Economic Development in East African History. The case
of Tanganyika 1850-1950. London: James Currey.
Klima, G. J. 1970. The Barabaig: East African Cattle Herders. London: Holt, Reinhart and Winston.
de Leeuw, P. N. and J. C. Tothill 1994. The concept of carrying capacity in sub-Saharan Africa - myth
or reality? In: Range Ecology at Disequilibrium: New Models of Natural Variability and Pastoral
Adaptation in African Savannahs, R. Behnke, I. Scoones, and C. Kerven, eds. Pp. 77-87. London:
Overseas Development Institute, International Institute for Environment and Development and
the Commonwealth Secretariat.
Lambin, E. F., B. L. Turner, H. J. Geist, S. B. Agbola, A. Anglesen, J. W. Bruce, O. T. Coomes, R. Dirzo,
G. Fischer, C. Folke, P. S. George, K. M. Homewood, J. Imberon, R. Leemans, L Xiubin, E. F.
Moran, M. Mortimore, P. S. Ramakrishnan, J. F. Richards, H. Skånes, W. Steffen, G. D. Stone, U
Svedin, T. A. Veldkamp, C. Vogel, and J. Xu 2001. The causes of land-use and land-cover
change: moving beyond the myths. Global Environmental Change 11: 261-269.
Lane, C. R. 1991. Alienation of Barabaig Pastureland: Policy Implications for pastoral Development in
Tanzania. PhD Thesis, University of Sussex, Surrey.
Lane, C. R. 1996. Pastures Lost: Barabaig Economy, Resource Tenure, and the Alienation of their Land in
Tanzania. Nairobi: Initiatives Publishers.
Lane, C. R. ed. 1998. Custodians of the Commons: Pastoral Land Tenure in East and West Africa.
London: Earthscan.
Lane, C. and J. Petty 1990. Displaced pastoralists and transferred wheat technology in Tanzania.
IIED Gatekeeper Series No. SA20.
Laris, P. 2002a. Burning the seasonal mosaic: preventative burning strategies in the wooded savanna
of Southern Mali. Human Ecology 30 (2): 155-186.
Laris P. 2002b. Grounding environmental narratives: the impact of a century of fighting against fire
in Mali. In: African Environment and Development: Rhetoric, Programs, Realities. eds. W.G.
Moseley and B.I. Logan. Ashgate Publishing.
Le Houérou, H.N. 1989. The Grazing Land Ecosystems of the African Sahel. Ecological Studies 75. New
York, Berlin & Heidelberg: Springer-Verlag.
262
Leach, M. and R. Mearns 1996. Environmental change and policy: Challenging received wisdom in
Africa. In: The Lie of the Land. Challenging Received Wisdom in African Environmental Change
and Policy, M. Leach and R. Mearns, eds. Pp. 1-33. International African Institute in association
with James Currey and Heinmann.
Leach, M., R. Mearns and I. Scoones. 1997. Environmental entitlements: a framework for
understanding the institutional dynamics of environmental change. IDS Discussion Paper 359.
Sussex: Institute of Development Studies.
Leach, M., R. Mearns and I. Scoones. 1999. Environmental entitlements: dynamics and institutions in
community-based natural resource management. World Development 27 (2): 225-247.
Lengisugi, ole N. 2000. Livestock Movements From Northern Tanzania Southwards and their Impact to
Poverty Alleviation. Paper presented at the 18th Tanzania Veterinary Association (TVA) Annual
Scientific Conference, 5-7th December, 2000, Arusha, Tanzania.
Linblade, K.A., G. Carswell and J.K Tumuhairwe 1998. Mitigating the relationship between
population growth and land degradation: land-use change and farm management in
southwestern Uganda. Ambio 27:565-571
Little, P. D., K. Smith, B. Cellarius, L. D. Coppock, and R. Barrows 2003. Avoiding Disaster:
Diversification and Risk Management among East African Herders. Development and Change
32: 401-433.
Long, N 1992. From paradigm lost to paradigm regained? The case for an actor-oriented sociology
of development. In: Battlefields of Knowledge: The Interlocking of Theory and Practice in Social
Research and Development. N. Long and A. Long eds. Pp. 16-43. London: Routledge.
Maddox, G., J. Giblin and J. Kimambo eds. 1996. Custodians of the Land: Ecology and Culture in the
History of Tanzania. London: James Currey.
Mackenzie, J. M. 1988. The Empire of Nature. Hunting, Conservation and British Imperialism.
Manchester: Manchester University Press.
Maganga, F.P. 2003. The interplay between formal and informal systems of managing resource
conflicts: some evidence from Tanzania. In: Securing Land Rights in Africa. T.A. Benjaminsen and
C. Lund. Pp. 51-70. London: Frank Cass.
Malthus, T. 1960. On Population (First Essay on Population, 1798, and Second Essay on Population,
1803). New York: Random House.
Massaro, R. 1993. Beyond participation: empowerment for environmental action in Tanzania’s West
Usambara Mountains. In: In Defence of Livelihood: Comparative Studies on Environmental Action.
J. Friedmarr and H. Rangan eds. Pp. 23-51. West Hartford CT: Kumarian Press
Max, J. A. O. 1991. The Development of Local Government in Tanzania. Dar es Salaam: Educational
Publishers and Distributors.
McAuslan, P. 1998. Making law work: restructuring land relations in Africa. Development and Change
29: 525-552.
McAuslan, P. 2000. Only the name of the country changes: the diaspora of “European” land law in
Commonwealth Africa. In: Evolving land rights and tenure in Africa. C. Toulmin and J. Quan eds.
Pp. 75-96. London: Department for International Development / International Institute for
Environment and Development / Natural Resources Institute
McCann, J. 1999. Green Land, Brown Land, Black Land. An Environmental History of Africa, 1800 -
1900. Portsmouth, New Hampshire / Oxford: Heinemann / James Currey.
Migot-Adholla, S. E., P. Hazell, B. Blarel and F. Place 1991. Indigenous land rights systems in sub-
Saharan Africa: a constraint on policy? World Bank Economic Review 5 (1): 155-175.
Migot-Adholla, S. E., G. Benneh, F. Place and S. Atsu 1994. Land security of tenure, and productivity
in Ghana. In: Searching for Land Tenure and Security in Africa. J. W. Bruce and S. E. Migot-
Adholla Pp. 97-118. eds. Dubuque IO: Kendall/Hunt Publishing Co..
Mnyampala, M. E. 1954. Historia, Mila, na Desturi za Wagogo wa Tanganyika. Desturi na masimulizi
ya Afrika ya Mashariki Kampala: Eagle Press.
Moore, D. S. 1998. Subaltern struggles and the politics of place: remapping resistance in Zimbabwe’s
eastern highlands. Cultural Anthropology 13(3): 344-81.
263
Moore, D. S. 1999. The crucible of cultural politics: reworking “development” in Zimbabwe’s eastern
highlands. American Ethnologist 26(3): 654-89.
Moore, S. F. 1986. Social Facts and Fabrications: "Customary" Law on Kilimanjaro, 1880-1980
Cambridge and New York: Cambridge University Press.
Moore, S. F. 1996. Post-socialist micro-politics, Kilimanjaro, 1993. Africa 66 (4): 587-605.
Moris, J. R. 1981. Managing induced rural development. International Development Institute.
Bloomington IN: Indiana University Press.
Mortimore, M. 1998. Roots in the African Dust: Sustaining the Sub-Saharan Drylands. Cambridge:
Mosse, D. 1997. The symbolic making of a common property resource: History, ecology and locality in
a tank-irrigated landscape in south India. Development and Change 28: 467-508.
Mung’ong’o, C. G. 1998. Coming full circle: agriculture, non-farm activities and the resurgence of out-
migration in Njombe District, Tanzania. Working Paper, Vol. 26. Dar es Salaam: Institute of
Resource Assessment and Leiden: African Studies Center.
Murton, J. 1999. Population growth and poverty in Machakos. Geographical Journal 165 (1): 37-46.
Musso, M. 1968. Mukwava na Kabila lake. Dar es Salaam: Eastern Africa Publications Ltd.
Mwakipesile, J. S. 1976. Peasants and Migrants: A Case Study from the Usangu Plains. MA thesis,
University of Dar es Salaam.
Mwinasa, E and V. Shauri 2001. Review of the Decentralisation Process and its Impact on Environmental
and Natural Resources Management in Tanzania. Dar es Salaam: Lawyers Environmental Action
Team.
Nahonyo, C. L., L. Mwasumbi, and D. G. Bayona 1998. A Survey of the Vegetation Communities and
Woody Plant Species in the MBOMIPA Project Area. Ministry of Natural Resources and Tourism
in co-operation with the Department for International Development, Tanzania.
NHPC 2002. National Housing and Population Census. United Republic of Tanzania. Online at:
http://www.tanzania.go.tz/census
Nader, L. and H. J. Todd 1978. Introduction. In: The Disputing Process: Law in Ten Societies. Ed. P.
Caplan. Oxford PR: Berg.
Ndagala, D. K. 1974. Social and Economic Change among the Pastoral Wakwavi and Its Impact on
Rural Development. Masters thesis, University of Dar es Salaam.
Ndagala, D. K. 1982 'Operation imparnati': The sedentarisation of the pastoral Maasai in Tanzania.
Nomadic Peoples No. 10: 28-39.
Ndagala, D. K. 1986. The Ilparakuyo livestock keepers of Bagamoyo: persistent fighters but ultimate
losers. Working Papers in African Studies No. 32, Department of Cultural Anthropology.
Uppsala: University of Uppsala.
Nelson, F., Nshala, and W. A. Rodgers. 2003. The evolution of land reform in Tanzania. Unpublished
paper.
Netting R. 1968. Hill Farmers of Nigeria. Cultural Ecology of the Kofyar of the Jos Plateau. Seattle:
University of Washington Press
Netting, R. M. 1993. Smallholders, Householders: Farm Families and the Ecology of Intensive,
Sustainable Agriculture. Stanford: Stanford University Press.
Neumann, R. P. 1992. Political ecology of wildlife conservation in the Mt. Meru area of northeast
Tanzania. Land Degradation and Society 3: 85-98.
Neumann, R. P. 1995b. Local challenges to global agendas: Conservation, economic liberalisation and
the pastoralists' rights movement in Tanzania. Antipode 27 (4): 363-382.
Neumann, R. P. 1996 Dukes, earls and ersatz edens: aristocratic nature preservationists in colonial
Africa. Society and Space 14 (1): 79-98.
Neumann, R. P. 1997. Primitive ideas: protected area buffer zones and the politics of land in Africa.
Development and Change 27: 559-582.
Neumann, R. P. 1998. Imposing Wilderness: Struggles over Livelihood and Nature Preservation in
Africa. University of California Press, Berkley.
264
Neumann, R. P. 2001. Africa's 'last wilderness': Reordering space for political and economic control in
colonial Tanzania. Africa 71 (4): 641-665.
Ness, G., W. Drake and S. Brechin eds. 1993. Population-Environment Dynamics: Ideas and
Observations. Ann Arbor, MI: University of Michigan Press.
Niamir-Fuller, M., S. Lugando, and T. Kundy 1994. Barabaig displacement from Hanang District to the
Usangu Plains: changes in natural resource management and pastoral production in Tanzania.
FAO Forests, Trees and People Programme - Phase II working paper.
Nigmann, E. 1908. Die Wahehe. Ihre Geschichte, Kult-, Rechts-, Kriegs- und Jagdbräuche. Berlin.
Njiforti, H. and N. Tchamba 1993. Conflict in Cameroon: parks for or against people. In: Indigenous
Peoples and Protected Areas: The Law of Mother Earth. E. Kemf ed. Pp. 173-78. London:
Earthscan.
North, D. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge
University Press.
Nyerere, J. 1968. Ujamaa - the basis of African Socialism. In: Ujamaa - Essays on Socialism. ed. J.
Nyerere. Oxford: Oxford University Press.
Odgaard, R. 1986. Tea - does it do the peasant women in Rungwe any good? In: Tanzania - Crisis
and Struggle for Survival. J. Boesen, K. Havnevik, J. Koponen and R. Odgaard eds. Pp. 207-
224. Uppsala: Scandinavian Institute of African Studies.
Odgaard, R. 2002. Scrambling for land in Tanzania: process of formalisation and legitimisation of
land rights. The European Journal of Development Research 14 (2): 71-88.
Odgaard, R. 2003. Scrambling for Land in Tanzania: Processes of Formalisation and Legitimisation of
Land Rights. In: Securing Land Rights in Africa. T. A. Benjaminsen and C. Lund eds. Pp. 71-88.
London: Frank Cass.
Ofkansky, T.P. 2002. Paradise Lost: A History of Game Preservation in East Africa. Morgantown: West
Virginia University Press.
Okoth-Ogendo, H. W. O. 1976. African land tenure reform. In: Agricultural Development in Kenya.
eds. J. Heyer, J.K. Maitha and W.M. Senga. Nairobi: Oxford University Press.
Okoth-Ogendo, H. W. O. 1989. Some issues of theory in the study of tenure relations in African
agriculture. Africa 59(1):6-17.
Okoth-Ogendo, H. W. O. 1993. Agrarian reform in sub-Saharan Africa: an assessment of state
responses to the African agrarian crisis and their implications for agricultural development. In:
Land in African Agrarian Systems, T. J. Bassett and D. Crummey, eds. Pp. 247-273. Madison:
University of Wisconsin.
Okoth-Ogendo, H. W. O. 2000. Legislative approaches to customary tenure and tenure reform in
east Africa. In: Evolving land rights and tenure in Africa. C. Toulmin and J. Quan eds. Pp. 123-
134. London: Department for International Development / International Institute for
Environment and Development / Natural Resources Institute.
Olson, M. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. London:
Harvard University Press.
Ostrom, E. 1990. Governing the Commons: The Evolution of Institutions for Collective Action.
Ostrom, E. 1994. Neither market nor state: governance of common-pool resources in the twenty-first
century. IFPRI Lecture Series. Washington DC: International Food Policy Research Institute.
Ostrom, E. 1998. A behavioural approach to the rational choice theory of collective action. American
Political Science Review 92 (1): 1-22.
Pala, A.O. 1980. The Joluo Equation: Land Reform = Lower Status for Women. Ceres May-June: 37-
42.
Parkipuny, M. S. ole 1975. Maasai Predicament Beyond Pastoralism. M.A. thesis, Institute of
Department Studies, University of Dar es Salaam.
Parkipuny, M. S. ole 1979. Some Crucial Aspects of the Maasai Predicament. In: African Socialism in
Practice: The Tanzanian Experience. A. Coulson ed. Nottingham: Spokesman
265
Peters, P. 1987. Embedded systems and rooted models: the grazing lands in Botswana and the
‘commons’ debate. In: The Question of the Commons: The Culture and Ecology of Communal
Resources. B. McCay and J. Acheson eds. Pp. 171-94. Tuscon AZ: University of Arizona Press.
Peters, P. E. 1994. Dividing the Commons: Politics, Policy, and Culture in Botswana. Charlottesville:
University of Virginia Press.
Peters, P. E. 2004. Inequality and social conflict over land in Africa. Journal of Agrarian Change 4 (3):
269-314.
Pimental, D., R. Harman, M. Pacenza, J. Pecarsky and M. Pimental 1994. Natural Resources and an
optimal human population. Population and Environment 15 (5): 347-369.
Pinckney, T. C. and P. K. Kimuyu 1994. Land tenure reform in East Africa: Good, bad or unimportant?
Journal of African Economies 3 (1): 1-28.
Platteau, J-P. 2000. Does Africa need land reform? In: Evolving land rights and tenure in Africa. C.
Toulmin and J. Quan eds. Pp. 51-74. London: Department for International Development /
International Institute for Environment and Development / Natural Resources Institute.
Pratt, R.C. 1971. The cabinet and presidential leadership in Tanzania, 1960-1966. In: The State of
the Nations: Constraints on Development in Independent Africa. M. Lofchie. Pp. 226-240.
University of California Press.
Popkin, S. L. 1979. The Rational Peasant: The Political Economy of Rural Society in Vietnam. Berkeley
and Los Angeles, CA: University of California Press.
Quan, J. 2000. Land tenure, economic growth and poverty in sub-Saharan Africa. In: Evolving land
rights and tenure in Africa. C. Toulmin and J. Quan eds. Pp. 31-50. London: Department for
International Development / International Institute for Environment and Development / Natural
Resources Institute.
Redmayne, A. 1964. The Wahehe People of Tanganyika. PhD Thesis, University of Oxford.
Ranger, Terence, 1993. The invention of tradition Revisited: the case of colonial Africa. In: Legitimacy
and the State in Twentieth-Century Africa. T. Ranger and O. Vaughan, eds. Pp. 62-111. London:
Macmillan.
Rekdal, O.B., and Blystad, A. 1999. "We are as sheep and goats": Iraqw and Datooga discourses on
fortune, failure, and the future. In. "The poor are not us": Poverty and pastoralism in Eastern
Africa. D.M. Anderson, & V. Broch-Due eds. Pp. 125-146.. Oxford: James Currey.
Redmayne, A. 1968. Mkwawa and the Hehe wars. Journal of African History IX: 409-436.
Reij, C., I. Scoones and C. Toulmin 1996. Sustaining the soil: indigenous soil and water conservation in
Africa. In: Sustaining the soil. C. Reij, I. Scoones and C. Toulmin eds. Pp. 1-27. London: Earthscan
Ribot, J. C. 2000. Decentralization, participation, and representation: administrative apartheid in
Sahelian forestry’. In: Development Encounters: Sites of Participation and Knowledge. ed. P.E.
Peters Pp. 29-60. Cambridge: Harvard University for HIID.
Richards, P. 1983. Ecological change and the politics of African land use. African Studies Review 26
(2): 1-71.
Rigby, P. 1969. Cattle and Kinship among the Gogo: A semi-Pastoral Society of central Tanzania.
Cornell University Press, Ithaca.
Rigby, P. 1983. Time and historical consciousness: the case of Ilparakuyo Maasai. Comparative Studies
in Society and History 25 (3): 428-456.
Rigby, P. 1985. Persistent Pastoralists: Nomadic Societies in Transition. London: Zed Press.
Rigby, P. 1992. Cattle Capitalism and Class: Il Parakuyo Maasai Transformations. Philadelphia: Temple
University Press.
Roe, E., L. Huntsinger, and K. Labnow 1998. High reliability pastoralism. Journal of Arid Environments
39: 39-55.
Runge, C.F. 1981. Common property externalities: isolation, assurance, and resource depletion in
traditional grazing context. American Journal of Agricultural Economics 63 (4): 595-606.
Runge, C.F. 1984. Institutions and the free rider: the assurance problem in collective action. Journal of
Politics 46: 154-81.
266
Runge, C. F. 1986. Common property and collective action in economic development. World
Development 14 (5): 623-635.
Sandford, S. 1983. Management of Pastoral Development in the Third World. Wiley, Chichester.
Sandford, S. 1995. Improving the efficiency of opportunism: new directions for pastoral development.
In: Living with Uncertainty: New Directions in Pastoral Development in Africa. I. Scoones ed. Pp.
174-182. London: Intermediate Technology Publications.
Scholes, R. J. and B. H. Walker 1993. An African Savanna: Synthesis of the Nylsvley study. Cambridge:
Schroeder, R. A. 1999. Geographies of environmental intervention in Africa. Progress in Human
Geography 23 (3): 359-378.
Scoones, I. 1992. Wetlands in drylands: Key resources for agricultural and pastoral production in
Africa. Ambio 20: 366-371.
Scoones, I. 1995. Living with Uncertainty: New Directions in Pastoral Development in Africa. London:
Intermediate Technology Publications.
Scoones, I. 1995. New directions in pastoral development in Africa. In: Living with Uncertainty: New
Directions in Pastoral Development in Africa. I. Scoones ed. Pp. 1-36. London: Intermediate
Technology Publications.
Scoones, I. 1996. Range Management Science and Policy. In: The Lie of the Land: Challenging Received
Wisdom on the African Environment, M. Leach and R. Mearns, eds. London: International African
Institute in Association with James Currey and Heinmann.
Scoones, I. and C. Toulmin 1998. Soil nutrient budgets and balances: What use for policy? Agriculture,
Ecosystems & Environment 71: 255-267.
Scoones, I. and C. Toulmin 2001. Transforming soils: the dynamics of soil-fertility management in
Africa. In: Dynamics and Diversity: Soil Fertility and Farming Livelihoods in Africa, I. Scoones, ed.
Pp. 1-44. London: Earthscan.
Schmidt, P. R. 1989. Early exploitation and settlement in the Usambara Mountains. In: Forest
Conservation in the East Usambara Mountains, Tanzania. Hamilton A C. and R. Bensted-Smith Pp.
75-78. Gland: IUCN
Sellen, D. W. 2003 Nutritional consequences of wealth differentials in East African pastoralists: the
case of the Datoga of Tanzania. Human Ecology 31 (4): 529-570.
Sen, A. 1984. Rights and capabilities. In: Resources, Values and Development. A. Sen ed. Pp. 307-324.
Oxford: Blackwell.
Shackley, S., B. Wynne, and C. Waterton 1996. Imagine complexity: the past, present and future
potential of complex thinking. Futures 28 (3): 201-225.
Shipton, P. 1988. The Kenyan land tenure reform: misunderstandings in the public creation of private
property. In: Land and Society in Contemporary Africa, eds. R.E. Downs and S.P. Reyna Pp. 91-
135. Hanover, NH: University Press of New England.
Shipton, P. 1994. Land and culture in tropical Africa: soils, symbols, and the metaphysics of the
mundane’. Annual Review of Anthropology 23: 347-77.
Shivji, I. G. 1998. Not Yet democracy: Reforming Land Tenure in Tanzania. Institute of Environment and
Development, London and Land Rights Research and Resources Institute, Dar es Salaam.
Shivji, I.G. and W.B.L. Kapinga. 1998. Maasai Rights in Ngorongoro, Tanzania. London: International
Institute for Environment and Development; Dar es Salaam: HAKI-ARDHI.
Shivji, I. G. and C. Maina Peter. 2000. The Village Democracy Initiative: A Review of the Legal and
Institutional Framework of Governance at Sub-District Level in the Context of Local Government
Reform. Report for the Ministry of Regional Administration and Local Government and the
United Nations Development Programme
Shorter, A. 1972 Chiefship in Western Tanzania: A Political History of the Kimbu. Clarendon Press,
Oxford.
Sieff, D. 1995. The Effects of Resource Availability on the Subsistence Strategies of Datoga Pastoralists
of North West Tanzania Institute of Biological Anthropology, Oxford: University of Oxford.
267
Solbrig, O., 1993. Ecological constraints to savanna land use. In: The World’s Savannas: Economic
Driving Forces, Ecological Constraints and Policy Options. Young, M., Solbrig, O. eds. Pp. 21-
48. For: Sustainable Land Use. Man and Biosphere Series, vol. 12. Paris: UNESCO and
Parthenon Publishing Group.
Sommer, G. and R. Vossen. 1993. Dialects, sectiolects or simply lects? The Maa language in time
perspective. In: Being Maasai: Ethnicity and Identity in East Africa. T. Spear and R. Waller, eds.
Pp. 25-37. London: James Currey.
Soper K. 1996. Nature/nature. In: Future Natural: Nature/Science/Culture. G. Roberson, M. Mash, L.
Tickner, J. Bird, B. Curtis and T. Putnam eds. Pp. 22-34. London: Routledge.
Stattersfield, A., M. Crosby, A. Long and D. Wedge 1998. Endemic Bird Areas of the World. Priorities
for Biodiversity Conservation. Birdlife Conservation Series No. 7. Cambridge: Birdlife
International.
Stenning, D. 1971. Household viability among the pastoral Fulani. In: The Development Cycle in
Domestic Groups. J. Goody ed. Pp. 92-119. Cambridge: Cambridge University Press.
Sullivan, S. 1999b. The impacts of people and livestock on topographically diverse open wood- and
shrub-lands in arid north-west Namibia. Global Ecology and Biogeography 8:257-277.
Sundet, G. 1997. The Politics of Land in Tanzania. PhD Thesis, University of Oxford.
Sundet, G. 2005. The 1999 Land Act and Village Land Act: a technical of the practical implications of
the acts. Draft manuscript.
Swai, I. 1996. The effect of villagisation and other policies on wildlife conservation in Tanzania. In
Community-based Conservation in Tanzania. N. Leader-Williams, J.A. Kayera, and G.L. Overton
eds. Pp. 51-54. Occasional Paper of the IUCN Species Survival Commission No. 15. Gland and
Cambridge: IUCN.
Swift, J. 1981. The Economics of Traditional Nomadic pastoralism: Twareg of Adrar N Iforas (Mali).
PhD Thesis, University of Sussex.
Swift, J. 1995. Dynamic ecological ecosystems and the administration of pastoral development. In:
Living with Uncertainty: New Directions in Pastoral Development in Africa. I. Scoones ed. Pp.
153-173. London: Intermediate Technology Publications.
Swift, J. 1996. Desertification: narratives, winners & losers. In: The Lie of the Land: Challenging
Received Wisdom on the African Environment M. Leach and R. Mearns, eds. Pp. 73-90. Oxford:
James Currey.
Sylla, I. 1995. Pastoral organisations for uncertain environments. In: Living with Uncertainty: New
Directions in Pastoral Development in Africa. I. Scoones ed. Pp. 134-152. London: Intermediate
Technology Publications.
Tax, S. and V. Hauck 2003. Sector-wide Approaches and Decentralisation Towards Greater Policy
Coherence: The Case of Tanzania. Study prepared in the context of a five-country review on the
relationship between Sector-Wide Approaches and Decentralisation coordinated by the
European Centre for Development Policy Management Maastricht, The Netherlands.
Tenga, R. W. 1987. Land law and the peasantry in Tanzania: a review of the post-Arusha period.
East Africa Social Science Review III (1): 38-57.
Thompson, J. 1881. To the Central African Lakes and Back: The Narrative of The Royal Geographical
Society's East Central African Expedition, 1878-1880. London: Sampson, Low, Marston, Searle
& Rivington.
Tiffen, M., Mortimore, M. and Gichuki, F. 1994. More People, Less Erosion: Environmental Recovery in
Kenya. Chichester: Wiley
Tomikawa, M. 1970. The distribution and migrations of the Datoga tribe. The sociological distinction
of the Datoga society in the Mangola area. Kyoto University African Studies 5: 1-46.
Tomikawa, M. 1972. Cattle brands of the Datoga tribe. Human relations in the Datoga pastoral
society in East Africa. Kyoto University African Studies 7: 1-35.
Toulmin, C. and J. Quan eds. 2000. Evolving land rights and tenure in Africa. London: Department for
International Development / International Institute for Environment and Development / Natural
Resources Institute.
268
Turner, M. 1998a. Long term effects of daily grazing orbits on nutrient availability in Sahelian West
Africa I. Journal of Biogeography 25: 669-682.
Turner, M. 1998b. Long term effects of daily grazing orbits on nutrient availability in Sahelian West
Africa II. Journal of Biogeography 25: 683-694.
URT 1994. Report of the Presidential Commission of Inquiry into Land Matters Vol. 1: Land Policy and
Land Tenure Structure. The Ministry of Lands, Housing and Urban Development, The United
Republic of Tanzania in co-operation with the Scandinavian Institute of African Studies.
URT 1995. A Review of the Wildlife Sector in Tanzania. Volume 1: Assessment of the Current Situation
and Volume 2: Possible Future Options. Wildlife Sector Review Taskforce. Dar es Salaam,
Ministry of Natural Resources and Tourism, United Republic of Tanzania.
URT 1997. National Irrigation Development Policy. Ministry of Agriculture and Cooperatives. Dar es
Salaam: United Republic of Tanzania.
URT 1998. The Wildlife Policy of Tanzania. Ministry of Natural Resources and Tourism, United
Republic of Tanzania.
URT 2001. Rural Development Strategy for Tanzania: Final Draft. Prime Minister's Office, The United
Republic of Tanzania.
URT 2002. Initial Environmental Examination, Mapogoro Irrigation Scheme. Environmental Cell Unit,
River Basin management and Smallholder Irrigation Improvement Project. Ministry of
Agriculture and Food Security, United Republic of Tanzania.
URT 2005 Preparation of a Strategic Plan for the Implementation of the Land Laws (SPILL) – Draft
Strategic Plan. Ministry of Lands and Human Settlements Development, United Republic of
Tanzania.
Turner, M. 1993: Overstocking the range: critical analysis of the environmental science of Sahelian
pastoralism. Economic Geography 69: 402-421.
UNEP 1992. World Atlas of Desertification. Nairobi: United Nations Environmental Programme.
van Velsen, J. 1967. The extended-case method and situational analysis. In: The Craft of Social
Anthropology, A. L. Epstein, ed. Pp. 129-149. Tavistock.
von Mitzlaff, U. 1988 Maasai Women. Life in a Patriachal Society. Field Research among the Parakuyo,
Tanzania. Munchen: Trickster Verlag.
Walsh, M.T. 1984. The Mis-interpretation of Chiefly Power in Usangu, South-west Tanzania. PhD Thesis,
University of Cambridge.
Walsh, M.T. 2002. Summary notes on the Kosisamba. Unpublished Manuscript.
Walsh, M.T. 2004. Against consensus? Anthropological critique and the deconstruction of international
water policy. Paper prepared for the ESRC-funded seminar on The Water Consensus -
Identifying the Gaps, Bradford Centre for International Development, Bradford University, 18-
19th November, 2004. Online at: http://www.brad.ac.uk/acad/bcid/seminar/water/
seminar_1/papers/ Walsh_Against_Consensus_Nov_04.pdf
Waller, R. 1979. The Lords of East Africa: The Maasai in the mid-Nineteenth Century, c. 1840-1885.
PhD Thesis, University of Cambrige.
Wanitzek, U. & Sippel, H. 1998. Land rights in conservation areas in Tanzania. GeoJournal 46: 113-
128.
Watson, D. and M. Baek 2001. Joint Government / Donor Review of the Local Government Reform
Programme. Dar es Salaam: Local Government Reform Programme.
Watts, M. 1983. Silent Violence: Food, Famine and Peasantry in northern Nigeria. Berkley CA:
University of California Press.
Whitmore, T. 1990. Long term population change. In: The Earth as Transformed by Human Actions:
Global and Regional Changes in the Biosphere over the Past 300 Years. B. L. Turner et al. eds.
pp. 25-39. New York: Cambridge University Press.
Williams, R. 1983. Keywords: a Vocabulary of Culture and Society. Second edition. London: Flamingo.
Wilson, G. 1952 The Tatoga of Tanganyika (Part I). Tanganyika Notes and Records 33: 34-47.
Wilson, G. 1953. The Tatoga of Tanganyika (Part II). Tanganyika Notes and Records 34:35-56.
269
Wily, E. 1988. The Political Economy of African Land Tenure: A Case Study from Tanzania. PhD Thesis,
University of East Anglia.
Wily, E. 2003. Community-based land tenure management: Questions and answers about Tanzania's
new Village Land Act, 1999. Issue Paper No. 120, International Institute for Environment and
Development. Drylands Programme.
Woodhouse, P. and H. Bernstein 2000. Whose environments? Whose livelihoods? In: African
enclosures? The Social Dynamics of Wetlands in Drylands. P. Woodhouse, H. Bernstein and D.
Hulme eds. Pp. 195-214. Oxford: James Currey.
World Bank Group 1996. River Basin Management and Smallholder Irrigation Improvement Project.
Staff Appraisal Report No. 15122-TA. Washington DC: Agriculture and Environment
Operations, Eastern Africa n Department, Africa Region.
World Bank Group 2001. Tanzania at the Turn of the Century. From Reforms to Sustained Growth and
Poverty Reduction. A World bank Country Study No. 22136. Washington DC: The World Bank
Group and United Republic of Tanzania.
World Bank Group/MIGA. 2002. Tourism in Tanzania: Investment for Growth and Diversification.
Washington DC: MIGA (Multilateral Investment Guarantee Agency) and United Republic of
Tanzania in cooperation with The Promote Africa Program.
Yeager, R. and N. Miller 1986. Wildlife, Wild Death: Land Use and Survival in Eastern Africa. Albany:
State University of New York Press.
Archival sources
Tanzania National Archives 1953-1955. Veterinary Monthly Reports, Accession 24, File V1/8/III.
Tanzania National Archives 1943-1955. Tsetse Fly General. Accession 24, File 70/1.
Tanzania National Archives 1930-1957. Game, National Park Policy and Instructions. Accession 24,
File G1/1, 7.
Tanzania National Archives 1935-1945. Aviation – Emergency Landing Ground – Idodi Area.
Accession 24, File A/8/8
Tanzania National Archives 1933-1955. Provincial Administration – Safari Reports District Officers,
File P4/1/III).
270

AWilliams PHD Thesis

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

AWilliams PHD Thesis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AWilliams PHD Thesis

Uploaded by

Copyright:

Available Formats

People cascades, land and livelihoods: Farmer and

herder land-use relations in the Idodi rangelands,

1.1 A study overview ............................................................................................................................................ 14

2 ‘Safeguarding’ environment and productivity ...................................................................... 18

2.1 Received knowledge, past policy and new understandings ................................................................. 18

3 Tanzanian semi-arid rangelands in perspective................................................................... 44

3.1 Introduction ...................................................................................................................................................... 44

4.1 Introduction ...................................................................................................................................................... 71

5 The peoples of ‘Kwigongo’: the old and the new ................................................................ 95

6 The Ilparakuyo of Idodi: Contending with change ............................................................... 143

7 Negotiating the political ecology of landed resources ....................................................... 199

7.1 Introduction ................................................................................................................................................... 199

8 Conclusion ......................................................................................................................... 231

8.1 Disaggregating trends in land-use relations in Idodi........................................................................... 231

A1: A short overview of Local government in Tanzania ...................................................... 239

Bibliography .............................................................................................................................. 256

Table 4.1: The population of Idodi by village in 2000 .................................................................................. 75

1.1 A study overview

1.2 The thesis structure

2.1 Received knowledge, past policy and new understandings

2.1.1 Received knowledge and past policy

2.1.2 New understandings

Arguments rooted in an anti-colonial and sometimes populist or anti-modernist

Common pool resource theory

The moral entitlement school

Firstly, socio-economic equality and poverty reduction, as opposed to the efficiency

2.2.2 Political ecology

The second form, informed by post-structuralist social theory, is characterised by a recognition of

2.3 Contending with new paradigms of complexity in environment and society

In substantial part, neo-Malthusian and science-based understandings of society and environment

2.4 Understanding conflict: environment, power and violence

The ‘environmental security’ paradigm is strongly underpinned by a neo-Malthusian understanding of

2.4.1 A summary critique of the ‘environmental security’ model

In summary, as reviewed by Hartmann (2001, 40-42), the model asserts that:

2.4.2 Political ecology: power, environmental discourse and conflict

2.5 Land and institutions: social embeddedness, inequality and conflict

2.5.1 The evolutionary model of land tenure

The problems of defining families, households, units of production, consumption and

2.5.2 Communitarian understandings of land tenure

… the importance of situating land (and resource) struggles in specific historical

Thus by deconstructing contemporary processes of socio-economic differentiation and inequality in

2.6 Mainstreaming new understandings of society and environment in policy

2.7 The study approach, analytical framework and methodology

2.7.1 The study context

2.7.2 The research questions and analytical framework

The analytical framework

2.7.3 The study methodology

3.1.1 Defining the semi-arid rangelands

Water Field site

Source: Institute of Resource

3.2.1 Colonial partitioning and landscape re-organisation

3.2.2 Post independence state control: Ujamaa, deconcentration and decentralisation

The Arusha Declaration and the onset of Ujamaa

Deconcentration of state power

The end of Ujamaa: decentralisation, liberalisation and local government reform

3.2.3 Control and space: state and nature

Western Tanzania 6 Msima GCA 12 Mlele South GCA 18 Wembere OA North

3.2.5 People cascades: landlessness, insecurity and wanderings