This publication was made possible through support
provided by the Office of Agriculture and Food Security, Global Bureau,
United States Agency for International Development, under Grant No. DAN-1328-G-00-0046-00.
The opinions expressed herein are those of the authors and do not necessarily
reflect the views of the U.S. Agency for International Development.
PRESENTED TO THE SR/GL-CRSP PROJECT FOR IMPROVING
PASTORAL RISK MANAGEMENT ON EAST AFRICAN RANGELANDS
1Research assistant, International Livestock Research Institute (ILRI), Nairobi, Kenya.
2Post-graduate research fellow, Institute for Rural Development, University of Goettingen, Germany
3Research Scientist, International Livestock Research Institute (ILRI), Nairobi, Kenya.
4Post doctoral fellow, International Livestock Research Institute (ILRI), Nairobi, Kenya, and the International Food Policy Research Institute (IFPRI), Washington, D.C.
5Research scientist, Marburg University,
Germany.
Proper citation: Luseno, W.K., A.B. Kamara, B.M.
Swallow, N. McCarthy and M. Kirk 1998. Community Natural Resource Management
in Southern Ethiopia. SR/GL-CRSP Pastoral Risk Management Project Technical
Report No. 03/98. Utah State University, Logan. 17 pp.
Contents:
Introduction and Background Information
Overview of Pastoralism and Property Rights in Southern Ethiopia
Policy Issues Concerning Land-use and Property Rights in Rural Ethiopia
Development Interventions and Pastoralism in Borana
Socio-economic Characteristics
Adoption
of Dry-Season Feeding Strategies in the Borana Pastoral System
Background
Introduction and Background Information
Livestock production is one of few options available to millions of
impoverished people who live in arid and semi-arid areas of Africa and
Asia. Livestock are flexible and fungible: they can be moved in response
to variable rainfall conditions and can be purchased or sold in response
to variable market conditions. Livestock can supply animal traction and
play key roles in the transfer and cycling of nutrients for crop production.
At the same time, livestock are often associated with severe land degradation
and are the object of conflicts among pastoral groups and between pastoralists
and farmers.
Property rights - the rules that determine who can use resources and
how they can use them - help shape the way that people use land and manage
livestock. Currently, in much of sub-Saharan Africa (SSA) grazing lands
are primarily governed by common property regimes, which allow people to
pool and reduce the risks associated with variable forage production.
The International Livestock Research Institute (ILRI), the International
Food Policy Research Institute (IFPRI) and the Institute for Rural Development
at the University of Goettingen are the principle partners in a research
project that aims at providing information to improve the efficiency, equity
and environmental sustainability of livestock production and land use in
SSA. The project focuses on areas where mobile livestock production and
mixed crop-livestock farming are competing land uses. In SSA most such
areas receive average annual rainfall between 300 and 600mm.
The specific objectives of the project are to: (i) better understand
grazing management under different property rights regimes and with different
environmental and production risks; (ii) identify the conditions under
which different development pathways are followed; and (iii) identify how
policy and other external interventions can assist communities to achieve
preferred development pathways.
The project is comprised of five complementary activities: (1) a comprehensive
review of relevant literature; (2) a conceptual framework analyzing the
relationships between property regimes, risk, land allocation and livestock
production; (3) simulation models that depict the short and long-term consequences
of alternative policies; (4) field research in Niger and southern Ethiopia
to examine these relationships and test hypotheses; and (5) an international
symposium held in Feldafing, Germany, in September 1998.
The two study sites were selected to capture local concepts of grazing
management under alternative property rights systems and to document patterns
of land use and range condition. In the case of southern Ethiopia, the
project has been able to build on past research conducted by the International
Livestock Centre for Africa (ILCA) and by ILRI, in collaboration with CARE
- Ethiopia and with the Southern Rangelands Development Unit (SORDU) of
the Ethiopian Ministry of Agriculture. Though there has been limited cultivation
in this area for decades, increased cultivation and bush encroachment,
among other factors, are having dramatic effects on land-use patterns.
Thereby, this research site is particularly interesting because of the
many layers of institutions that interact over the governance of pasture
lands, cropped lands and maintenance of water sources. The fieldwork in
the Borana plateau was carried by Abdul B. Kamara, a Research Fellow at
the Institute for Rural Development, University of Goettingen, Germany,
under the supervision of Nancy McCarthy (ILRI / IFPRI post-doctoral scientist),
Brent Swallow (ILRI) and Michael Kirk (Marburg University in Germany).
The purpose of this report is to describe the field activities and to
profile the study area in terms of the dominant production and land tenure
systems (section II), key socio-economic variables and bio-physical characteristics
(section III). Preliminary results on the analysis of adoption of hay-making
among the Borana of southern Ethiopia are also presented (section IV).
Data collection
The field data was collected in forty communities(1)
chosen in the six districts of Yabello, Dirre (Mega), Arero, Negelle, Teltele
and Moyale. The major criteria for stratification were the level (mean)
and distribution of rainfall - coefficient of variation across communities.
The rainfall data used in the stratification consists of a six-year time
series data (1992 to 1997) collected by the Southern Rangelands Development
Unit (SORDU) from 11 weather stations on the plateau. In choosing the communities,
therefore, proximity to weather stations, dominance of pastoralism and
accessibility were the three major criteria. The distribution of the chosen
communities around weather stations is shown in the table below.
Table 1: Level and distribution of rainfall communities
| Station | chracteristicsa | Meanb | Std. Dev. | CoV | # of Comm. |
| Sarite | LR, LCV | 469 | 103 | 0.24 | 3 |
| Wachile | LR, HCV | 473 | 222 | 0.49 | 5 |
| Web | LR, HCV | 399 | 199 | 0.48 | 4 |
| D/Wachu | LR, HCV | 353 | 130 | 0.39 | 2 |
| Dillo | LR, HCV | 499 | 170 | 0.39 | 3 |
| Yabello | HR, HCV | 519 | 230 | 0.44 | 5 |
| Arero | HR, HCV | 873 | 374 | 0.43 | 4 |
| Negelle | HR, HCV | 739 | 241 | 0.32 | 3 |
| Moyale | HR, HCV | 869 | 588 | 0.67 | 2 |
| Hidilola | HR, LCV | 717 | 202 | 0.28 | 4 |
| Did Yabelo | HR, LCV | 496 | 141 | 0.28 | 1 |
| Teltele | HR,LCV | 634 | 135 | 0.23 | 4 |
| TOTAL | --- | --- | --- | --- | 40 |
From Table 1, it can be seen that the area is generally characterized
by low rainfall with an annual mean that fluctuates between 353 and 873
mm per annum; with a bimodal distribution that peaks around March to May
(long rains) and around September to November (short rains). The area is
thus semi-arid and precipitation poses the major livestock production constraint.
Phases of the data collection
Data collection was undertaken in three phases as follows:
i) Community Level Surveys: The community level survey consisted of a 12-page questionnaire that was completed using participatory appraisal methods (PRA). The first part consisted of a social mapping exercise that throws light on the various production activities undertaken within the communities. These include cultivation practices, land allocation patterns, hay-making strategies, transhumance routes etc. This was followed by field visits to different sites and geomorphological units (mountains, hills, valleys, etc) as well as the identification of community border points using a portable Geographical Positioning System (GPS) instrument. Two GIS packages were used as follows: the data were first entered into ArcInfo, where boundaries for each community were digitized and areas for each community calculated; ArcView was then used to prepare community maps. The remaining parts on the questionnaire consisted mainly of basic socioeconomic questions on production, land use, wealth, marketing, rules and regulations as well as social cohesion. A total of 40 pastoral communities (ardas) were surveyed, each of which exercise jurisdiction over some form of pasture and/or water resource. Each community consisted of two or more settlements (ollas).
ii) Marketing and Price Data: Marketing and price data were collected from the six marketing centers on the Borana Plateau. In each market, 30 bulls were measured (heart girth, wither height, condition score etc.) so as to give indirect measurements of body weight. Average prices will be estimated here while controlling for the effects of other factors that may affect price but vary from market to market. The method of hedonic price analysis will be used for the estimations. The estimated prices or coefficients across markets will then be used in the econometric estimations of the community level data.
iii) Range quality data: The range quality data consists of quadrants,
taken from two transects in each community: North-South-Transects (NST)
and East-West-Transects (EWT). From each transect, three quadrant samples
were taken. The plant species in each quadrant were identified and relative
abundance determined. The percentage of bush encroachment was also assessed
for each community followed by slope measurements. This data will be used
in assessing the range condition of the forty communities so that changes
that require urgent attention are brought to light. A range quality index
will also be generated for each community and this will be used in the
econometric estimations.
I. Overview of Pastoralism and Property
Rights in Southern Ethiopia
Livestock production plays a major role in Ethiopia's economy and accounts
for about 30 per cent of the output value of the agricultural sector (Central
Statistics Bureau, 1996). The country has a diverse agricultural environment:
the central massive highlands rise up to an altitude of 4000m while in
the lowlands altitudes may be far below 1000m. The highlands comprise about
40 per cent of the land area but support about 90 per cent of the rural
population and form the main seat of Ethiopia's agricultural activities.
Agropastoralism is the dominant production system in the highlands. Unlike
the highlands, the lowlands are characterized by relatively low human population
densities and highly variable and uncertain rainfall. These semi-arid lowlands
are thus characterized by pastoral systems, with traces of cultivation
around villages where population densities are relatively high and around
valleys where the agro-ecological conditions permit this activity. The
most important pastoral groups in Ethiopia are the Somali in the south-east,
the Afar in the north-east and the Borana in the southern rangelands. Though
livestock densities are much higher in the highlands, the highlands rely
very much on the lowlands for their livestock and products; about 20 per
cent of the draught animals used in the highlands come from the lowlands.
Lowland breeds are robust and generally regarded superior to indigenous
highland breeds both in terms of size, endurance, productivity, and in
terms of performance in cross breeding programs and consumer preferences
on the international market. Thus, Borana cattle may well comprise 90 per
cent of Ethiopia's cattle export, contributing a great deal to the country's
foreign exchange earnings (Coppock, 1994). Hence the southern rangelands
have received considerable research and development attention over the
past decades: including the Joint Ethiopian Pastoral Systems Study; ILCA's
eleven years of pastoral systems research (1980-1991); development activities
of the Southern Rangelands Development Unit (SORDU); CARE-Ethiopia; Deutsche
Gesellschat für Technische Zusammenarbeiz GTZ; Save the Children;
Norwegian Church Aid (NCA), etc.
The body of literature compiled here is intended to provide a general
guide to relevant literature on issues related to livestock production
and development in Ethiopia, with a specific focus on the Borana rangelands
of southern Ethiopia. The first section deals with policy related issues
that affect land use, property rights and the way in which these influence
pastoral livestock production. The second section discusses the dynamics
of the Borana pastoral system: constraints to extensive livestock production,
external interventions and the current status of the southern rangelands.
2.1 Policy Issues Concerning Land-use and Property Rights in Rural Ethiopia
Several changes in land use policy have been recorded in Ethiopia over
the past decades (Teka, 1983; Ghose, 1985; Tolosa et al. 1995). These changes
are believed to have had different impacts in diverse parts of the country,
based on the prevailing type of production systems. The changes are attributed
to the divergence of policy priorities by the different regimes, namely
- the Monarchy before 1974 (Pre Derg), the Ethiopian Socialist Government
(Derg) and the Transitional Government of Ethiopia (Post Derg).
Each of these regimes introduced and implemented different land policies
that have very important implications for different production systems
in the country. Before the fall of the Monarchy, land tenure in Ethiopia
was characterized by an intricate and hierarchical system that varied greatly
across different parts of the country (Teka, 1983; Ghose, 1985; Bruce et
al., 1994; Tolossa & Asfaw, 1995). In the northern region, access to
land was based on hereditary rights and community membership -with the
imposition of tithe, tributes and other services imposed on the peasants.
In the south, much of the land was held in large estates by landlords,
and farmed by tenants who were often the original inhabitants of the area.
Rents and other services were imposed or sharecropping arrangements made.
Tenure insecurity was high and eviction easy since most of the agreements
were merely verbal. Land holdings of peasant households hardly went above
5 hectares and this was often comprised of parcels in more than one community.
The area was thus characterized by an unequal distribution of land and
tenure insecurity. By the 1960s it had become obvious that land tenure
emerged as a great constraint to development in the country (Bruce et al.,
1994).
The emergence of the Derg Regime in 1974 was greeted by a series of
land reform programs (Tolossa & Asfaw, 1995). The regime was quick
to recognize the need for land reform and on March 5, 1974 all rural land
was nationalized and declared 'the collective property of the Ethiopian
people'. The reform was implemented almost everywhere; expropriation of
land from landlords, establishment of state farms, abolition of wage labor
and villagization programs were immediately initiated. All other existing
land rights were extinguished and all land became public property with
the aim of 'liberating the masses from oppression and exploitation' by
the land-owning classes. The implementation of this land redistribution
program began with the formation of Peasant Associations (Kebelles),
whose primary responsibility was the allocation of land to households.
The redistribution was based on a set of criteria that differed from one
region to another -family size, availability of land, productive potential
of the land- and guaranteed a maximum of ten hectares per household held
under usufructuary rights. Households could make claims only at residential
Kebelles. Under this system, all individual transfer of land, regardless
of the method involved, was prohibited. In pastoral areas, communities
that had commonly and independently managed their pasture in the past started
to become responsible to the central administration through the chairmen
of the peasant associations. This situation started to create disincentives
in land improvement -fencing, bush cleaning etc.- and constrained mobility
in pastoral areas. Before its eventual fall in 1991, the Derg regime had
already relaxed some of its rigid policies -lifting the ban on hired labor
and slowing down the resettlement program.
The post Derg regime has not been very drastic on reforming land tenure
policy. Despite its declaration that land continue to be nationally owned,
certain inheritance rights were once again reassured and the resettlement
programs stopped. This was followed by a call for the establishment of
commercial farms by private individuals with reassuring statements of guaranteeing
property rights, credit facilities, construction of feeder roads and tax
concessions to facilitate marketing activities (Bruce et al., 1994). Since
the fall of the Derg, a general improvement in marketing conditions both
in terms of better selling opportunities and in terms of more consumer
goods becoming available in local markets have been observed in the country.
On the Borana plateau, where pastoral livestock production has dominated
the production system for centuries, resource use decisions and rules had
been, and in many cases continue to be, made by traditional institutions
and implemented by groups of elders within the community. With the advent
of crop production in the recent decades, national level land policies
started to have an impact on traditional resource management systems. The
formation of the peasant associations in the 1970s to implement the land
redistribution programme conflicted a great deal with the role of traditional
elders and their institutions. Under the new system, individuals can secure
access to land through the peasant associations without having to go through
the traditional elders, who, according to tradition, will generally not
guarantee such private rights. Today, a considerable part of the plateau
is experiencing a growing trend of privatization; communal pastures falling
into the hands of cultivators and pretenders: those who acquire the land
with the pretext of cultivating, but who use it as pastures (Community
Surveys, 1997/98). It is believed that the spatial arrangement of these
private fields, if the trend continues, will soon become a major constraint
to mobility, one of the traditional risk management strategies of the Borana
pastoralists. The abolition of burning in the late 1970's and the concomitant
bush encroachment has also contributed a great deal to the spread of diseases,
restricted mobility and a reduction in the available grazing area each
year.
2.2 Development Interventions and Pastoralism in Borana
After many unfruitful efforts of intervention into pastoral production
systems -ranches, sedentarization programs, etc.- it is widely acknowledged
today that pastoralism itself is an adaptive feature of any group living
under conditions which make extensive livestock production an effective
way of utilizing natural vegetation of highly variable characteristics.
The Borana pastoral system, for instance, is characterized by extreme variability
and unreliability of rainfall both between different years and spatial
distribution within a given year and rainfall is bimodally distributed
(SORDU, unpublished data). The subsequent high spatial and temporal variability
of water and forage generation makes the southern rangelands vulnerable
to recurrent droughts. Though the Borana cattle have a remarkable ability
for endurance and tolerance in time of water scarcity, the impact of such
droughts has often been recorded to reach life threatening levels: the
droughts of 1983/84 and 1990/91 are a few examples. Coppock (1994) hypothesizes
that the impact of droughts on the Borana rangelands is exacerbated by
high stocking rates; so that even years with rainfall slightly below the
average could lead to great cattle losses. He described the system as one
which experiences phases of 'booms and bursts'. Once a high density phase
is reached (characterized by high stocking rates), the system becomes very
vulnerable to a 'crash' even by the slightest rainfall deficit that would
otherwise have been easy to overcome. This behavior has been ascribed to
the desire of the Boran people to persistently hold on to their cattle
and sell only under acute needs. At the beginning of a drought period,
the Borana pastoralists resort instead to reserve grazing areas, herd splitting
and long distance migrations in search of water and forage in fallback
areas and selling small ruminants to buy grains. The Boran people hold
on to their cattle in this way and simply hope that a miracle will happen
-the rains will soon come- until an inevitable situation of mass losses
is reached, paving the way for the influx of relief programs. Coppock concludes
that the system is subject to a simultaneous processes of 'long-term trends'
which result from a declining ratio of cattle to people; and 'short-term
cycles' characterized by variable stocking rates -low cattle density during
drought phases and high cattle density after drought recovery phases. This
unpredictable vulnerability to droughts, coupled with the fact that the
area is beset by poorly developed infrastructure and inadequate amenities
have made it possible for the Borana pastoral systems to attract a great
deal of development attention.
Development interventions in the areas of animal health, water technology
(procurement, use and conservation), roads, markets, ranches and fattening
programs etc. have been initiated and implemented on the southern rangelands,
since the 1960s. The impact of a number of such programmes has, however,
been far below expectation (SORDU-Manager, personal communication). It
is argued that most of these efforts were based on assumptions that conflicted
with the social values and traditional production rationale of the Borana
pastoralists; planners had erroneously assumed that animal health and market
outlets were the only constraint to market off-take both for domestic consumption
and export (Coppock, 1994). Cattle population over the years did in fact
grow due to animal health improvement but trade never took off as speculated,
leading to increased pressure on the limited range resource base.
II. Description of the Study Area
The Borana plateau is located in the southern lowlands of Ethiopia and
occupies a total land area of approximately 95, 000km2. The
study area occupies a 20 by 20 km region (between 36 and 42 east and 4
and 6 north), and is located about 650 km south of Addis Ababa. Figure
1 shows the location of the study area and the distribution of the forty
ardas in the study site.
As was mentioned earlier, the forty ardas were selected to represent
different rainfall patterns (level and distribution). Monthly rainfall
data from 1992 to 1997 for 12 weather stations located around the Borana
plateau were used to classify the communities into four rainfall categories:
high mean rainfall - high variation, high mean rainfall - low variation,
low mean rainfall - low variation, and low mean rainfall - high variation.
Figure 2 shows the distribution of the 40 communities on the basis of these
rainfall categories.
3.1 Biophysical Characteristics
In this section the research site is described in terms of the great
diversity of climatic and topographic features found in the area. The Spatial
Characterization Tool (SCT), a geographic information systems (GIS) application
which accesses gridded environmental data, point data and vector based
information (polygons), was used to investigate a variety of physical attributes
of the 40 communities. This review makes use of the information obtained
from applying the SCT to characterize the micro-environments of the 40
communities and to validate the stratification criteria employed in the
sampling strategy described above.
While the complete SCT data file incorporates data from Africa, Latin
America and parts of South East Asia, data availability varies from region
to region. Where available the full dataset includes information on climate,
topography, land cover, demographics, and soils as well as ancillary information
on towns, political units, waterbodies and major watersheds. The major
part of the data that is used in this analysis are in the form of gridded
ARC/INFO surfaces (Corbett and O'Brien, 1997).
For this report, the zone characterization facility of the SCT was employed.
This facility is useful when information about regions rather than individual
sites is required. The parameters defining the zone are entered as Geographic
Positioning System (GPS) coordinates. In the case of this report, the SCT
application was unable to identify individual communities as zones, thereby
larger regions which in most cases comprised of more than one arda were
defined. Figure 3 shows the results of the zone specification procedure.
A report on the characteristics of the site (climate, population density,
topography, soils) was generated. STATA, a statistical package, was used
to test the significance of various hypothesized relationships.
Long term annual normals for precipitation, potential evapotranspiration
(PE), precipitation / potential evapotranspiration (P/PE) and maximum,
minimum and mean temperature were accessed and are summarized below. The
study area is dominated by a semi-arid climate and the data show mean annual
rainfall in the 12 zones as varying between approximately 500mm (Zone 7)
and 700 mm (Zones 1, 2 and 11). The overall average annual rainfall is
648 mm. Mean annual rainfall varied significantly between the four rainfall
categories (P = 0.0000) and also with elevation (r2 = 0.8954;
P 0.05). Standard deviations also varied substantially between the four
stratification groups (P = 0.0114).
Table 2 : Mean Annual Precipitation (mm)
| Zones (Communities) | Mean | Maximum | Minimum | Std. Dev. |
| Zone 1 (1 to 6) | 711.68 | 787.00 | 660.00 | 33.05 |
| Zone 2 (7 to 8) | 702.76 | 787.00 | 642.00 | 38.35 |
| Zone 3 (9 to 12) | 650.13 | 658.00 | 640.00 | 5.75 |
| Zone 4 (13 to 15) | 659.50 | 665.00 | 656.00 | 3.40 |
| Zone 5 (16) | 663.33 | 673.00 | 653.00 | 8.18 |
| Zone 6 (17 to 18) | 663.00 | 663.00 | 663.00 | 0.00 |
| Zone 7 (19 to 21) | 497.10 | 553.00 | 462.00 | 28.24 |
| Zone 8 (22 to 25) | 654.00 | 671.00 | 632.00 | 14.54 |
| Zone 9 (26, 27, 29) | 668.50 | 739.00 | 619.00 | 41.27 |
| Zone 10 (28, 30, 31, 39, 40) | 554.18 | 573.00 | 536.00 | 10.91 |
| Zone 11 (32 to 35) | 714.44 | 814.00 | 626.00 | 63.51 |
| Zone 12 (36 to 38) | 637.91 | 744.00 | 558.00 | 53.38 |
The results from the SCT application show significant differences in
the maximum and the minimum temperatures between the four stratification
groups (P = 0.0073 and P = 0.0002, respectively). Mean temperatures were
also significantly correlated with elevation (r2 = -0.9673 and
r2 = - 0.9041, respectively; P 0.05), indicating that as elevation
increases temperatures are likely to decrease. Indeed, zones 7 and 10,
which are located at lower elevations, show the highest mean maximum temperatures;
while zones 1 and 2, which are situated at higher elevations, have the
lowest mean minimum temperatures.
Coppock (1994) notes that the principal issue regarding temperatures
in arid and semi-arid environments, is how it affects the effectiveness
of rainfall through its influence on evapotranspiration. The effectiveness
of rainfall in turn is determined by how it affects plant production and
the distribution of plant species. Regarding potential evapotranspiration,
the SCT results show the expected trend whereby the rates of evapotranspiration
increases as mean temperatures increase.
Table 3: Mean Maximum Temperatures (C)
| Zones (Communities) | Mean | Maximum | Minimum | Std. Dev. |
| Zone 1 (1 to 6) | 25.78 | 26.99 | 24.49 | 0.66 |
| Zone 2 (7 to 8) | 25.86 | 26.99 | 24.49 | 0.63 |
| Zone 3 (9 to 12) | 26.54 | 26.80 | 26.36 | 0.13 |
| Zone 4 (13 to 15) | 26.03 | 26.20 | 25.95 | 0.08 |
| Zone 5 (16) | 25.26 | 25.78 | 24.73 | 0.43 |
| Zone 6 (17 to 18) | 27.59 | 27.73 | 27.44 | 0.15 |
| Zone 7 (19 to 21) | 28.40 | 28.97 | 27.36 | 0.49 |
| Zone 8 (22 to 25) | 26.87 | 27.27 | 26.57 | 0.27 |
| Zone 9 (26, 27, 29) | 27.46 | 28.32 | 25.88 | 0.91 |
| Zone 10 (28, 30, 31, 39, 40) | 28.79 | 29.14 | 28.47 | 0.19 |
| Zone 11 (32 to 35) | 26.09 | 27.37 | 24.56 | 0.96 |
| Zone 12 (36 to 38) | 26.69 | 27.57 | 25.59 | 0.57 |
Table 4: Mean Minimum Temperatures (C)
| Communities | Mean | Maximum | Minimum | Std. Dev. |
| Zone 1 (1 to 6) | 14.19 | 15.17 | 12.98 | 0.59 |
| Zone 2 (7 to 8) | 14.36 | 15.52 | 12.98 | 0.69 |
| Zone 3 (9 to 12) | 15.39 | 15.64 | 15.13 | 0.16 |
| Zone 4 (13 to 15) | 15.92 | 16.01 | 15.88 | 0.04 |
| Zone 5 (16) | 15.39 | 15.80 | 14.96 | 0.34 |
| Zone 6 (17 to 18) | 17.35 | 17.49 | 17.20 | 0.15 |
| Zone 7 (19 to 21) | 18.11 | 18.71 | 17.13 | 0.49 |
| Zone 8 (22 to 25) | 15.22 | 15.70 | 14.82 | 0.32 |
| Zone 9 (26, 27, 29) | 14.62 | 15.56 | 13.02 | 0.91 |
| Zone 10 (28, 30, 31, 39, 40) | 17.35 | 17.57 | 17.08 | 0.15 |
| Zone 11 (32 to 35) | 15.08 | 16.43 | 13.59 | 0.97 |
| Zone 12 (36 to 38) | 15.56 | 16.84 | 14.10 | 0.78 |
The precipitation/potential evapotranspiration ratio (PPE) provides
a measure of the loss of moisture due to evapotranspiration. Thus, a relatively
low PPE is an indication of a high loss of moisture due to evapotranspiration.
It is expected that as one moves from high to low elevation, mean rainfall
will decrease while mean temperatures and thereby potential evapotranspiration
will increase. Thus, PPE will decrease as elevation decreases or conversely
will increase with elevation. Indeed, the SCT results strongly support
this theory (r2 = 0.9324; p 0.05). Further, the results indicate
that the high-risk groups have lower PPE ratios than the low-risk groups,
that is to say that moisture loss due to evapotranspiration is higher in
the former group. Indeed, the SCT results show that the high risk groups
are on average located at lower elevations and have on average higher mean
temperatures.
Table 5 : Total Evapotranspiration (mm)
| Zones (Communities) | Mean | Maximum | Minimum | Std. Dev. |
| Zone 1 (1 to 6) | 1419.89 | 1482.00 | 1361.00 | 32.74 |
| Zone 2 (7 to 8) | 1427.10 | 1482.00 | 1361.00 | 33.59 |
| Zone 3 (9 to 12) | 1463.75 | 1478.00 | 1452.00 | 7.48 |
| Zone 4 (13 to 15) | 1491.17 | 1494.00 | 1489.00 | 1.77 |
| Zone 5 (16) | 1467.67 | 1487.00 | 1446.00 | 16.82 |
| Zone 6 (17 to 18) | 1573.00 | 1582.00 | 1564.00 | 9.00 |
| Zone 7 (19 to 21) | 1683.00 | 1731.00 | 1613.00 | 36.39 |
| Zone 8 (22 to 25) | 1463.75 | 1487.00 | 1446.00 | 15.35 |
| Zone 9 (26, 27, 29) | 1469.88 | 1514.00 | 1400.00 | 39.36 |
| Zone 10 (28, 30, 31, 39, 40) | 1579.24 | 1595.00 | 1563.00 | 9.76 |
| Zone 11 (32 to 35) | 1537.56 | 1622.00 | 1449.00 | 58.85 |
| Zone 12 (36 to 38) | 1546.73 | 1620.00 | 1465.00 | 44.43 |
Table 6: Precipitation/Potential Evapotranspiration Ratio (PPE)
| Communities | Mean | Maximum | Minimum | Std. Dev. |
| Zone 1 (1 to 6) | 0.50 | 0.58 | 0.45 | 0.04 |
| Zone 2 (7 to 8) | 0.50 | 0.58 | 0.44 | 0.04 |
| Zone 3 (9 to 12) | 0.45 | 0.46 | 0.44 | 0.01 |
| Zone 4 (13 to 15) | 0.46 | 0.46 | 0.45 | 0.00 |
| Zone 5 (16) | 0.46 | 0.48 | 0.45 | 0.01 |
| Zone 6 (17 to 18) | 0.44 | 0.45 | 0.44 | 0.00 |
| Zone 7 (19 to 21) | 0.30 | 0.34 | 0.27 | 0.02 |
| Zone 8 (22 to 25) | 0.45 | 0.47 | 0.43 | 0.01 |
| Zone 9 (26, 27, 29) | 0.46 | 0.54 | 0.42 | 0.04 |
| Zone 10 (28, 30, 31, 39, 40) | 0.36 | 0.37 | 0.35 | 0.00 |
| Zone 11 (32 to 35) | 0.47 | 0.56 | 0.38 | 0.06 |
| Zone 12 (36 to 38) | 0.41 | 0.51 | 0.34 | 0.05 |
The landscape of the study area is gently sloping across an elevation
of 1000 to 1900m (Tables 7 and 8). Zones 1 and 2 show the most variable
landscapes with slopes ranging between 0 and 14%. Not surprisingly, these
zones also occupy the high elevations. Shrubland and savanna dominate the
vegetative cover in the study area (Corbett and O'Brien 1997).
Table 7: Elevation (m)
| Communities | Mean | Maximum | Minimum | Std. Dev. |
| Zone 1 (1 to 6) | 1669.89 | 1918.00 | 1423.00 | 134.91 |
| Zone 2 (7 to 8) | 1647.76 | 1918.00 | 1423.00 | 132.09 |
| Zone 3 (9 to 12) | 1498.00 | 1536.00 | 1449.00 | 24.82 |
| Zone 4 (13 to 15) | 1537.50 | 1549.00 | 1510.00 | 13.02 |
| Zone 5 (16) | 1675.67 | 1777.00 | 1579.00 | 80.90 |
| Zone 6 (17 to 18) | 1253.50 | 1279.00 | 1228.00 | 25.50 |
| Zone 7 (19 to 21) | 1054.36 | 1258.00 | 941.00 | 96.14 |
| Zone 8 (22 to 25) | 1455.00 | 1514.00 | 1377.00 | 52.31 |
| Zone 9 (26, 27, 29) | 1382.12 | 1645.00 | 1240.00 | 149.35 |
| Zone 10 (28, 30, 31, 39, 40) | 1097.41 | 1149.00 | 1042.00 | 30.18 |
| Zone 11 (32 to 35) | 1499.33 | 1852.00 | 1187.00 | 230.60 |
| Zone 12 (36 to 38) | 1382.36 | 1673.00 | 1161.00 | 149.54 |
Table 8: Percent Slope
| Zones (Communities) | Mean | Maximum | Minimum | Std. Dev. |
| Zone 1 (1 to 6) | 2.34 | 14.00 | 0.00 | 1.62 |
| Zone 2 (7 to 8) | 2.25 | 14.00 | 0.00 | 1.65 |
| Zone 3 (9 to 12) | 1.22 | 2.00 | 0.00 | 0.66 |
| Zone 4 (13 to 15) | 0.14 | 1.00 | 0.00 | 0.35 |
| Zone 5 (16) | 0.04 | 2.00 | 0.00 | 0.23 |
| Zone 6 (17 to 18) | 2.62 | 7.00 | 0.00 | 1.68 |
| Zone 7 (19 to 21) | 1.11 | 3.00 | 0.00 | 0.79 |
| Zone 8 (22 to 25) | 1.41 | 3.00 | 0.00 | 0.78 |
| Zone 9 (26, 27, 29) | 1.24 | 7.00 | 0.00 | 1.67 |
| Zone 10 (28, 30, 31, 39, 40) | 0.34 | 2.00 | 0.00 | 0.55 |
| Zone 11 (32 to 35) | 1.45 | 4.00 | 0.00 | 1.05 |
| Zone 12 (36 to 38) | 1.19 | 4.00 | 0.00 | 0.81 |
3.2 Socio-economic Characteristics
In this section, a profile of the study area is provided in terms of
key socio-economic variables. Some of the variables that are discussed
are community population, community herd-size, community stocking rates,
proportion of community land allocated to private pasture and proportion
of community land allocated to cultivation.
The basic unit of production and consumption in the Borana pastoral
system is the household. However, certain productive activities such as
herding, watering animals, marketing dairy products and constructing corrals
and fences are undertaken cooperatively by the community (Coppock 1994).
The tables below show descriptive statistics of the community level data
set.
Table 9: Household and Community Characteristics
| Settlement/Household | Total | % | Average/arda1 |
| No. of Settlements (Olla) | 199 | 100 | 4.99 |
| Number of Households | 3141 | 100 | 78.52 |
| Male Headed Household | 2314 | 73.68 | 57.85 |
| Female Headed Household | 825 | 26.32 | 20.67 |
| Wealthy Households | 385 | 12.57 | 9.63 |
| Middle class Households | 660 | 21.01 | 16.50 |
| Poor Households | 2098 | 66.47 | 52.45 |
settlement (Olla) consists of at least ten households.
In total, there are 199 ollas in the 40 communities comprising of 3,141
households. About 74% of these household are male headed. 13% of all households
are classified as wealthy, 21% as middle class and the rest as poor (Table
9). The classification was based on the wealth classification criteria
of the respective communities.
Table 10: Livestock Holdings (heads of animals per community)
| TYPE | (%)2 | Min. | Max. | Mean | Std Dev | Total |
| Cattle | 94 | 66 | 13,350 | 1,612 | 2,972 | 64,469 |
| Goats | 39.22 | 0(10) | 6,320 | 394 | 1,047 | 15,747 |
| Sheep | 23.17 | 0(2) | 1,640 | 125 | 349 | 5,005 |
| Horses | 0.57 | 0(2) | 10 | 0.57 | 2 | 23 |
| Donkeys | 14 | 0(1) | 153 | 24 | 35 | 942 |
| Camels | 16.3 | 0(2) | 606 | 48 | 98 | 1,928 |
Table 11: Livestock Holdings (TLU per community)
| TYPE | Min. | Max. | Mean | Std Dev | Total | %l |
| Cattle | 46 | 9,345 | 1,228 | 2,081 | 45,128 | 90.60* |
| Goats | 0(1) | 632 | 39 | 105 | 1,574 | 3.16 |
| Sheep | 0(1) | 164 | 13 | 35 | 501 | 1.01 |
| Horses | 0(2) | 8 | 0.5 | 1.62 | 18 | 0.04 |
| Donkeys | 0(1) | 107 | 17 | 24 | 659 | 1.32 |
| Camels | 0(2) | 606 | 48 | 98 | 1,928 | 3.87 |
| Total | 58 | 9,780 | 1,245 | 2,224 | 49,808 | 100 |
Cattle is the major type of livestock species kept by the Borana pastoralists.
A total of about 64,000 heads of cattle, 16,000 goats, 5,000 sheep, 900
donkeys and 23 horses constitute the livestock holdings of the forty communities.
The minimum number of cattle per community is 66 and the maximum of 1,600.
Cattle constitute about 91% of the total livestock (in TLU) of the sample
communities (Tables 10 and 11).
Table 12: Land Use - Land Allocation to Different Activities (%)
| LAND USE TYPE | Proportion (%) |
| Enclosure for lactating cattle etc. (Warra) | 49.00 |
| Grazing area for dry herds (forra) | 1.25 |
| Enclosure for calves, sick animals. (Kalo) | 11.00 |
| Area for settlements and small ruminants | 10.00 |
| Buffer area
SUM OF COMMUNAL LAND Private enclosure for calves Cultivated area |
4.00
75.25 3.00 16.75 |
| Area for draught animals | 5.00 |
| SUM OF PRIVATE LAND | 24.75 |
The results from the 40 community surveys provide an overall picture
of land allocation and property rights in the Borana plateau. The results
indicate a large increase in the area cultivated since 1986 (Table 12).
Seventeen percent of the area in the 40 ardas was reported to be within
cultivated fields in 1997. This includes pastures adjacent to cultivated
fields within the thorn fences used to protect and demarcate the fields;
this study was not able to distinguish between land actually cropped and
pasture land within crop enclosures. The community survey also indicates
that enclosures of pasture for grazing calves and for draught animals is
quite important, forming 14% and 5% of total area, respectively. More surprisingly,
the in-depth studies revealed that all calf enclosures in these two communities
were in fact enclosed by individual households. Ten percent of the land
in the ardas was used for housing, for grazing by sheep and goats, and
four percent was reserved as a buffer area. The remainder, about 64% of
the total area, was managed as communal pasture.
Table 13: Cultivation Practice
| Practice Duration | No. of Communities | Percentage |
| 0 years (no cultivation)
1-10 years 11-20 years >20 years Total Communities Cultivating |
8
12 9 11 32 |
20.0
30.0 22.5 27.5 80.0 |
Least number. of ha cultivated by a single household = 0.40
Average number. of ha cultivated by a single household = 1.41
Expansion of cultivation and enclosure of land around cultivated fields
are the most noticeable and important changes in land use in the Borana
plateau. Up to 17% of the land area in the 40 communities is now cultivated,
compared to 1.4% in 1986. About 80% of the communities in the sample now
include some households that cultivate; 30% of the 40 communities took
up cultivation within the last 10 years and 22% took up cultivation in
the last 20 years. Thirty years ago only four communities (10%) were cultivating.
Cultivation is still relatively less significant with an average of 1.4
hectares cultivated by a single household; a maximum of 2.3 hectares and
a minimum of 0.4 hectares.
The above data set will be utilized mainly in the estimation of the
econometric models, designed to test the relationships between stocking
rates, land use, property rights and management institutions. The community
level data is also supplemented by an in-depth data taken from two sets
of communities from among the forty communities covered in the earlier
survey. It is hoped that the in-depth data will generate information on
the impact of policy, external interventions and drought cycles on changes
in land use, property rights and management institutions.
III. Adoption of Dry-Season Feeding
Strategies in the Borana Pastoral System
The International Livestock Centre for Africa (now ILRI) conducted a
long-term systems study of Borana pastoralism between 1980 and 1991. During
that period, several experiments were conducted on hay making and, finally,
recommendations were made regarding the use by hay making in Borana. CARE-International,
a partner in the systems study, then incorporated hay making as part of
their extension message to farmers in the area. To date, however, no study
had been undertaken to consider the extent of adoption of hay making among
the Boran.
In the community survey, therefore, questions were posed to community leaders about the adoption of hay making in their communities. This section of the paper presents some preliminary results on the extent of hay making on the 40 communities, and factors affecting adoption at the community level.
The traditional feeding practice during the dry season consists of Borana
women collecting standing brown grass on a daily basis to hand feed calves.
This is quite a laborious task and indeed highly inefficient with regards
to time management (Mulugeta 1990; Coppock 1992). Further, this practice
has been shown to provide forage of very low nutritive value (Mulugeta
1990; Coppock 1993). In view of this, Coppock (1992) identifies two main
benefits to be gained from adopting haymaking as a dry-season feeding strategy:
(1) improved animal productivity; and (2) reduction in women's labor burden
in the dry season.
Inadequate nutrition due to deficits in forage quantity and/or quality,
particularly during the dry seasons and drought years, imposes serious
limitations on livestock production in pastoral systems. These limitations
include serious impediments on animal productivity such as low milk yields,
increased stock mortality (especially calf mortality), extended calving
intervals and low average daily animal weight gains (McIntire, Bourzat
and Pingali 1992). Coppock (1993) shows that hay made from local grasses
has a higher nutritive value than the standing brown grass that is available
in the dry season. Thereby, hay-making shows promise as a development intervention
aimed at solving problems associated with dry-season feeding among semi-settled
Borana pastoralists.
Married Borana women are responsible for managing calves during their
first year after birth. During the dry season, caring for calves is particularly
challenging because forage quantity and quality is low, water is scarce,
and demand for milk within the traditional homestead is increased. During
these dry periods, traditional calf management practices among the Borana
women include providing milk and forage daily, and watering every two to
three days (Coppock 1992). Mulugeta (1990) reports that in order for these
activities to be accomplished women need to budget approximately 20 hours
of their time every week. Thus, having hay available during such critical
periods, is expected to allow women more flexibility in budgeting time
and work schedules (Coppock 1991 and 1992).
In order that high quality hay be produced, the process begins after
the long rains after the vegetation has dried off sufficiently and before
dominant grasses flower. Prior to commencing the actual haymaking procedure,
an initial investment needs to be made in constructing a platform elevated
0.5m above the ground on which to store the hay. On commencement, time
budgets must take into consideration: searching for and cutting nutritious
grass species; drying the cut grass and guarding it from wandering livestock;
transporting the hay to the olla; and haystacking to protect it against
bleaching from the sun and leaching from rain.
Regarding hay making, community representatives were asked whether hay
making is practiced in the arda and if so what percentage of the arda members
are engaged in it. Representatives were also asked if people harvest crop
reside to feed to animals and in the cases where they did they were asked
to estimate the number of members engaged in the activity. The table below
shows the distribution of responses.
Table 14: Adoption of hay making in 40 communities in the Borana Platteau
| yes | No | |
| Q: Is hay making practiced in this arda?
(haymake) |
13 (32.50%) | 27 (67.50%) |
| Q: Do people harvest any crop residue to feed animals?
(cropfeed) |
18 (45.00%) | 22 (55.00%) |
Thirteen out of forty respondents said that hay was prepared in the arda. Results indicated that an average proportion of 68% of households within these ardas making hay were involved in the practice. Similarly, in the 18 ardas where crop residue is harvested to feed to animals, results show that an average of 72 households were involved in the practice. These suggests that some form of collective action or collective decision making affects hay making: hay making is not adopted by a few early adopters in each arda, but by a majority. One explanation for this phenomena is that hay-making is related to the structure of property rights in the community.
The results of the survey of 40 communities make it clear that the Borana
platteau of southern Ethiopia is becoming increasingly diverse. Cattle
numbers in the communities vary tremendously, with only 66 in one community
and over 13,000 in another. Cultivation is increasing, but at very different
rates in different communities. Of the 40 communities that we surveyed,
8 still have no land cultivated, while in other communities up to half
of the land is cultivated. Individualization of rangeland into private
and small group pastures is increasing, but again, at very different rates
in different communities. Adoption of hay making, an innovation first introduced
about 10 years ago, is increasing, but is still practiced in a minority
of communities.
More research needs to be done in order to understand the importance
of the many factors that have caused this increasing heterogeneity. Insights
from that research may have general applicability in Ethiopia and across
East Africa.
Bruce, J. W., Hoben, A., & Rahmato, D. (1994). After the Derg:
An Assessment of Rural Land Tenure Issues in Ethiopia. Land Tenure
Center, University of Wisconsin, Madison and Institute of Development Research,
Addis Ababa University (1994).
Coppock, D. L. (1994). The Borana Plateau of Southern Ethiopia: Synthesis
of pastoral research, development and change 1980 - 91. ILCA Systems
Study 5. Addis Ababa: International Livestock Centre for Africa.
Ghose, A. K. (1985). Transforming Feudal Agriculture: Agrarian Change
in Ethiopia since 1974. Journal of Development Studies, 22
(1985): 128-149
Tolossa, G. & Asfaw, Z. (1995). Land Tenure Structure and Development
in Ethiopia: A Case Study of Ten Peasant Associations in Wara Jarso Woreda.
Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ): Sector
Project: Assistance to Desertification Control, Relevance of Land Tenure
Development for Developing Countries (Division 425: 1995).
1 A community in this study refers to the Borana traditional administrative unit called Arda. An arda consists of two or more pastoral settlements (Ollas), each of which comprises of at least ten households. Each arda has jurisdiction over some form of grazing area, cultivation land and water resources.