Much of Africa is suitable only for livestock production and wildlife conservation. Climatic patterns make cultivation difficult or impossible in the arid and semi-arid regions of East Africa. Droughts occur with high frequency and the East African bimodal rainfall regime is unfavorable for crop production (Pratt and Gwynne 1977, Ellis and Galvin 1994). Consequently, pastoralism or extensive livestock herding is the most prevalent form of land use in East Africa in terms of land area utilized. These pastoral regions also support some of the largest and most viable wildlife populations in Africa. Traditional pastoral livestock production has been highly compatible with wildlife conservation in the past, but this compatible interaction is showing signs of disintegration. Spatial components of pastoral ecosystems have been disrupted by competing forms of land use, with negative implications for ecosystem persistence (Coughenour 1991). Pastoralists who have lived with wildlife for centuries often reap little benefit from the income which wildlife generates through ecotourism (Norton-Griffiths 1995). Although the idea that livestock production and wildlife conservation are compatible is shared by many, and there are examples of successful integration (Western 1982, Cumming 1991), there are examples where compatible associations of pastoralism and wildlife are deteriorating (eg. Arhem 1985, Galvin 1995).

Kenyan rangelands support over 25% of the human population, and over half of the livestock population of that country. The livestock sector produces 10% of the gross domestic product. About 80% of Kenyan large wildlife are found on these areas. Income from associated tourism has grossed over $500 million per year, and has become an important and reliable source of revenue for the national government and local authorities (Ottichilo et al. 1997). According to one source tourism is the primary source of foreign exchange, and wildlife-based tourism is 50% of the total (Byrne et al. no date, Grootenhuis et al. 1991).

Recently released wildlife monitoring data of the Kenya Department of Resource Surveys and Remote Sensing (DRSRS) show that during the 1970's-1990's there were significant declines in 13 wildlife species, while only 4 species showed no significant decline (Ottichilo et al. 1997). The majority of the observed declines occurred during 1970-80. The main causes of the decline were poaching and landuse change, but much of the decline could be attributed to efforts to commercialize livestock production through the creation of group ranches (Ottichilo et al. 1997). Wildlife were systematically excluded from these areas to minimize livestock-wildlife competition. Indeed, there seems to have been little doubt that livestock production would be reduced by wildlife. One study showed that grazing competition may reduce net ranch income by 50% or more (Byrne et al. no date, Grootenhuis 1991). Costs of additional disease control may reduce ranch income by another 15%. Furthermore, wildlife provided no direct benefit to the land holders. Wildlife also declined significantly in areas where the government promoted arable land use in the arid and semi-arid ecoclimatic zones (eg. Narok, Laikipia) (Ottichilo et al. 1997).

Ecologically unsound livestock development schemes, coupled with increased human population densities have often led to overgrazing and environmental degradation (Coughenour 1991). It has been suggested that a common cause of livestock and wildlife declines (Otichello et al. 1997) may be a reduction in rangeland carrying capacity brought about by progressive rangeland degradation (Rainy and Worden 1997). The degradation would presumably be a result of excessive livestock densities and restricted livestock movements. Unfortunately, there are little data to show that range production has declined throughout the region. Livestock declines have probably contributed to the dramatic increase in livestock raiding in northern Kenya in the last two years, however (Rainy and Worden 1997).

Thus, livestock-wildlife and livestock-environment interactions have been major constraints to livestock development in East Africa. The approach of simply attempting to maximize livestock production through ranching, banking, or other schemes, has backfired because of a lack of understanding of livestock-wildlife interactions, failure to assess the direct and indirect effects of livestock development on wildlife and the environment, failure to recognize the ecological adaptive features of traditional pastoralism, and failure to recognize the importance of wildlife for economic development and long-term ecological viability in this region. Livestock development in this region cannot occur without integrated assessments of livestock-wildlife interactions and resultant effects on human welfare. There is clearly a need to establish a more appropriate and sustainable balance between food security and natural resource conservation in the pastoral regions of East Africa.

Landscape-level integrated assessments are essential since that is the scale at which livestock, wildlife and environment interact, and at which land use problems can be solved through community-based conservation, and integrated conservation and development programs. Regional-level integrated assessment are needed for policy formulation at national and international levels. For example, regional conservation strategies are being formulated through methodologies like GAP analyses, i.e. the identification of biodiversity hotspots and their protection status, and the sizes and connectivities of wildlife reserves (Scott et al. 1993); and the identification of large-scale metapopulations, migration corridors, and related opportunities for preserving genetic diversity (Harris 1984, Hansen et al. 1993).

During the first half of our research we revised the problem model as follows.

Patterns of land use in pastoral areas of East Africa are rapidly changing, mainly towards systems of decreasing spatial scale, and reduced options for large-scale livestock movement. In Kenya, there was conversion of former Maasai communal grazing lands into group ranches in the 1960's. Now, these group ranches, and other communal grazing lands in Maasai land and elsewhere are being subdivided and privatized. In many cases land-tenure rights have been granted in small parcels to individual families. They may attempt to use the parcel as a private ranch, or increasingly, they are using the value of the land holding as collateral to take out loans. As spatial scale decreases, intensity of use usually increases. Above a certain livestock density, wildlife apparently cannot exist. A second trend in land use is the increasing adoption of small-scale farming by pastoral people. In some cases the pastoralists are being encouraged to do so. Increasingly, pastoral systems are mixed systems of livestock and small agricultural plots.

Although wildlife dispersal areas have been occupied by pastoralists for many decades, agricultural development in the dispersal areas now threatens to destroy season grazing, and close critical game migration corridors (J. Kinyamario, pers. commun.). The remaining parks and reserves are becoming isolated islands with smaller, less diverse, and genetically poorer wildlife populations. The current government policy in Kenya is to subdivide the large group ranches, but the consequences of this subdivision for wildlife and plant biodiversity are unknown. Pastoralists and wildlife both need to move seasonally, for survival and production. Inappropriate programs to regulate pastoral movements may decrease livestock production and survival. There is actually little capability now, to be able to predict the effects of disrupted movements on the welfare of either the wildlife, or the pastoral livestock.

The original problem model emphasized spatially extensive pastoral ecosystems. We broadened the problem model to include a wide range of land tenure arrangements ranging in spatial scale from: (a) private ranches (eg. Kajiado, Kenya, Laikipia), to (b) group ranches (eg. Kajiado, Loita), c) restricted communal land use (eg. Serengeti-Ngrongoro), to (d) traditional communal land use (eg. Loliondo, Turkana), to (e) parks which exclude people entirely.

Many of the formerly wildlife rich pastoral areas of East Africa are now depauperate or wildlife poor, due to warfare and increased access to advanced weaponry. J. Else pointed out that this is especially true in Uganda, and may also be true in Ethiopia, Somalia, Sudan, and Eritrea. For example, game was essentially extirpated from NW Uganda during the Amin and Obote years. Now there is continued rebel activity coming from southern Sudan. In Karamoja, most of the game has been eliminated due to the arming of the people with automatic weapons left behind by Amin's exiled forces. A few "community" game reserves do exist. Game declines in Turkana can also be attributed to access to weapons, and a weak conservation ethic among these people. Other reasons for wildlife decline include increasing human populations, and increased use of marginal lands by agriculturalists rather than by pastoralists.

The original problem model was aimed primarily at wildlife rich areas such as Ngorongoro Conservation Area. The problem model was revised to allow for the fact that many in many cases the wildlife-pastoral interaction is more of an issue of rehabilitation, than of conservation. Indeed, according to J. Else, rebuilding the protected area system in Uganda will be a prerequisite to developing tourism in that country. We did not definitively decide to pursue research in wildlife poor areas, however. Security issues are a concern, as are the possibilities for accomplishing anything in such areas in a reasonable amount of time. The IMAS could be used to provide input and guidance into rehabilitation efforts.

Comments made by a representative of a Maasai NGO (F. ole Ikayo) suggested more attention should be given to assessing the value of indigenous knowledge, both in livestock husbandry and in natural resources conservation. The Maasai have coexisted with wildlife for centuries, and are good stewards of the land. Accordingly, we expanded the problem model to recognize the value of assessing ecosystem responses to indigenous natural resource conservation practices. The Maasai also feel that "capacity building" is not the issue since they already have "capacity". Similarly, "participatory research" is insufficient. What is needed is increased "empowerment". This exemplifies the political problems that lie at the core of land-use conflicts. Pastoralists are losing traditional grazing rights and land to other forms of land use including conservation areas and agriculture. The integrated modeling and assessment system (IMAS) we are aiming to develop will not give political power to pastoralists directly, but instead, the goal is to "empower with information" thereby leveling the playing field of the political arena. We believe that the IMAS will be useful for clearly illustrating the causes and effects of land use changes, according to the goals and objectives of the pastoralists, as well as other land use interests. Lack of information about these causes and effects has led to no-win situations, as competing land users have not been capable of expressing their goals in objective terms, much less coming to a successful resolution. At present there are either no data to support the viewpoints of either side, or the data have not been provided in a useful format.

The problem model can be summarized at this point as shown in Figure 1. Changes in land-use and land management appear to be central to the entire problem. Land-use interacts with ecosystem structure and dynamics through such processes as primary and secondary production, which are in turn driven by climate. Development and conservation policy both influence land-use, with subsequent impacts on the ecosystem, pastoral welfare, livestock production, wildlife conservation efforts, and ecosystem integrity. Although traditional culture influences land-use, land management, modernization and increasing interaction with society at local through regional scales is also affecting land-use, with subsequent impacts on the variables mentioned above.

The problem model evolved further through Workshop II, in which we specifically solicited input from regional representatives of government and non-government organizations.

A common theme expressed by the representatives was that information by itself has little impact, even when it is readily available. For example the Kenya DRSRS data showing long-term trends in wildlife abundance and distribution has been little used for policy formulation. Only recently has the data been summarized (Ottichilo et al. 1997, Rainy and Worden 1997). A similar comment was made by S. Chama at the first workshop, that information is already abundant, and it has had little impact. In fact, he pointed out there was a moratorium on information gathering in Maasai-land. H. Cheruiyot of KARI also pointed out that there is already lots of information available. The indigenous knowledge of the pastoralists has even been taken into account. The problem now is what to do with the information, especially how to use it for effective policy formulation. For example there currently are no policies on competition for forage or disease transmission between wildlife and livestock. According to N. Menzies of the Ford Foundation, there is little evidence that information by itself can sway entrenched socio-political and socio-economic interests which influence changes in land use (and thus pastoralists and wildlife). We need to do more than just inform management. We need to find solutions that take into account and resolve the wants and needs of stakeholders at different levels. H. Dublin of WWF pointed out that despite high quality science, political interests may drive processes in other directions, thus constraining the effectiveness of purely science-based analyses. There is a need to understand how policies originate, and what their effects are. There is a need to improve the use of available information for policy formulation. Policy assessments are imperative.

This might be summarized by making the important distinction between simply providing information, and making the information useful. Information is only useful if it is transformed into knowledge which explain the causes of the observations or when it can be used to develop potential solutions. Many of the causes of change are external to the local ecological and social systems. Consequently, despite excellent information about local ecological and social systems, the information may have little impact. In a sense, local ecosystems have become decoupled from processes at broader scales. As a result, assessments must do a better job of integrating local and regional scale processes. Strictly local-level assessments cannot succeed. Conversely, the broad-brush approaches that have been taken at only the regional levels have not been effective in producing tangible results and there is a critical need for solutions which produce tangible results at the local level (H. Dublin).

For information to be useful, it must inform people about the implications of changes, and show linkages between cause and effect. There is ample information and widespread recognition that changes are occurring. What people need is information about the trade-offs of taking alternative courses of action. Why have the observed changes occurred? What can be done? People need to be able to make informed choices about alternate land-use strategies to select the one that is most viable (N. Menzies). This is exactly what our proposed integrated modeling and assessment system aims to accomplish. We will not simply provide information about change, we will provide both retrospective and prognostic information about the causes and implications of change under alternative scenarios of policy and management. We will develop a tool which integrates and synthesizes information from different sources and disciplines. The integrated modeling and assessment system (IMAS) will be used to evaluate alternative solutions, and search for the best solution. It will involve stakeholders and scientists in an integrative assessment process, which is markedly different from developing data sets that will be archived in obscure documents or computer files.

The original problem model emphasized the potential conflicts encountered when wildlife and pastoralists share the same habitat. However, E. Barrow of AWF suggested that wildlife conservation and pastoralism are potentially complementary land uses. Pastoralists and wildlife have similar requirements for large ranges, opportunistic movements, water sources, dry season grazing areas, and diversity of browsers and grazers. They also have several common enemies including land and water peremption by cultivation and irrigation, lack of rights to access land, and lack of economic returns from wildlife conservation. Pastoralists and wildlife have three main conflicts - competition for resources, disease, and damage. However, the benefits arising from compatibilities of the two land uses are likely to offset the costs of conflicts if ways can be found to accrue benefits from conservation. Pastoralism has considerable economic value which has not been factored into integrated assessments thus far. In contrast to the previous emphasis on conflicts between wildlife and pastoralists, the combination of pastoralism and wildlife may prove to be an economically successful land use, from both a local and national perspective. We caution, however, that the ecological characteristics of intact pastoral ecosystems, such as ability to move over large areas and access key seasonal grazing areas, would have to be fully recognized, and either conserved or emulated for successful combined use. Pastoralism was indeed a sustainable land use under free access to grazing lands, but the reality is that now, land use is constrained. Having recognized that fact, we propose that these constraints can be alleviated through strategic management and policy. Our integrated modeling and assessment system is aimed at the identification of these effective strategies.

Progress is being made in Kenya towards solving the problems of conflicts between wildlife and other forms of land use, including pastoralism. A new national wildlife policy (1996) has been put into place. The policy calls for developing partnerships among government agencies, local authorities, and private landowners, and an integrated approach to conservation and development based on coordinated ventures and interagency coordination (Ottichilo et al. 1997). In implementing the policy, KWS has created the Partnership Department whose main objective is to encourage land owners to accept wildlife on their land and by assisting land owners to obtain tangible benefits from wildlife, including benefit sharing, tourism, and certain forms of utilization (Ottichilo et al. 1997). Where the program has been implemented, land owners are eager to keep wildlife on their land. Ranchers in Laikipia and Nakuru are now conserving wildlife on their lands.

If the solution to wildlife-livestock conflicts is being implemented already through such policies, then how useful will the proposed integrated modeling and assessment system (IMAS) be? Firstly, these schemes have not been widely implemented throughout East Africa. They have been implemented only in the more modernized districts of Kenya, which is the most highly developed and privatized country in the region. Secondly, the schemes have not been implemented long enough, or over a large enough area, to demonstrate that they are universally successful. Thirdly, the IMAS will be designed so that it can and will be used to conduct assessments of areas where these schemes are implemented, such as Kajiado and Laikipia. The system will be useful for monitoring and assessing the success of the schemes, for identifying why they do or do not work under the existing constraints, and for providing useful information to managers to further refine management strategies and policies. Livestock production and wildlife conservation could be even more highly optimized through the use of the proposed IMAS in such environments.
 

Pastoralism or extensive livestock herding is the most prevalent form of land use in East Africa in terms of land area utilized. These pastoral regions also support some of the largest and most viable wildlife populations in Africa. Pastoral livestock production has been highly compatible with wildlife conservation in the past, but this compatible interaction is showing signs of disintegration. Wildlife declined markedly during the 1970's-1990's. Although population growth, poaching, and land use change were major causes, much of the decline in Kenya, at least, could be attributed to efforts to commercialize livestock production, create group ranches, and exclude wildlife to minimize livestock-wildlife competition. In widespread areas of East Africa, traditional pastoral grazing lands have been lost to cultivation and wildlife reserves. These losses have reduced pastoral ecosystem integrity, and have probably contributed to wildlife declines outside protected areas. Thus, livestock-wildlife and livestock-environment interactions have been major constraints to both livestock development and wildlife conservation.

We are proposing to develop an integrated modeling and assessment system (IMAS) to assess livestock-wildlife interactions in pastoral ecosystems of East Africa. The assessment system will integrate computer modeling, geographic information systems, remote sensing, and field studies. The system will enable alternative policy and management strategies to be objectively explored, debated, implemented, and reassessed. Stakeholder involvement will be elicited from the outset. Regional level analyses for East Africa will eventually be conducted using GIS, modeling, networking, and cross-site comparisons. The regional GIS and modeling analyses would be used to identify and prioritize areas of strong and problematic interactions between pastoralists and wildlife.

Our ultimate goal is to improve prospects for increasing food security, conserving wildlife, and sustaining ecosystem integrity throughout the region. Pastoralism and wildlife conservation can be compatible enterprises.The combination of land-uses can be economically beneficial, support the welfare of pastoral people, and also conserve wildlife and the environment. Compatibility requires careful management and a policy framework which is sensitive to the needs of both pastoralists and wildlife. However, quantitative methodologies for impacts of alternative policy and managment solutions on pastoralists, wildlife, and ecosystems are lacking.

Two workshops were held in East Africa to develop the assessment approach and to identify appropriate sites and establish a communication network. An interdisciplinary team was formed, comprised of U.S. and East African scientists, pastoralists, resource managers, and other stakeholders. Input from the region was obtained through problem statements given by regional representatives, and group identification of scientific priorities, policy and management issues, and potential benefits of the IMAS at individual sites. The overall design and specific components of the integrated modeling and assessment system were developed in working groups. Potential implementation sites were identified representing the wide range of problems encountered in livestock-wildlife interactions throughout East Africa.

Overview of Proposed Activities

Our ultimate goal is to improve prospects for increasing food security, conserving wildlife, and sustaining ecosystem integrity. We share the conviction that pastoralism and wildlife conservation can be compatible enterprises. The combination of land-uses can be economically beneficial, support the welfare of pastoral people, and also conserve wildlife and the environment.

The objective of our assessment team process was to develop a protocol which combines integrated modeling, geographic information systems (GIS), remote sensing (RS) technologies, field methods for assessing wildlife, natural resources, pastoral land use, nutrition, and household economics (Figure 2). The proposed approach would start with the SAVANNA model, a GIS-based ecological simulation model originally developed for the Turkana pastoral ecosystem in northern Kenya (Coughenour 1991,1992,1993, Ellis et al. 1993). This model builds upon concepts and methods used to assess energy flow through a spatially extensive pastoral ecosystem (Coughenour et al. 1985). The spatially explicit model simulates plant growth, animal foraging, animal production, animal population or herd dynamics, and animal spatial distributions (Figure 3).

During the initial one-year assessment we aimed to create an interdisciplinary team of U.S. and East African scientists, managers and other stakeholders which would develop an analytical system through which alternative policy and management strategies can be objectively explored and debated (see Assessment Team Members). The team, with American and East African members, would assess and modify the integrated modeling and assessment system (IMAS) to fit selected regions in Tanzania, Kenya, Uganda, and Ethiopia. The assessment team research would focus on a review of the SAVANNA system, its potential application, data requirements, and necessary modifications to be useful as an integrative management tool.

In our assessment team research we proposed to identify and focus on study sites where there has been a long history of research, with involvement and familiarity by assessment team members. We suggested the Ngorongoro Conservation Area (NCA)/Loliondo Game Reserve/Serengeti National Park Region as a primary intensive study area for the development of the IMAS.

The approach for selecting team members would be to make contacts in writing, fax, or email, invite representatives to the workshops, or arrange to meet prior to or after the workshops at their sites. The team members will be selected on the basis of their level of interest and commitment to collaborative research, and the needs of the assessment process.

The assessment team would, through two workshops in East Africa, further design the research agenda, design the IMAS and decide how to provide analysis and assessment procedures for policy and land use analysis for optimizing human welfare and biodiversity in ecosystems which support wildlife, livestock, and biotic diversity.

The workshops would focus on problem assessment, the identification of team members who have the complement of skills to contribute to problem model development and implementation. The current assessment implementation model would be presented, along with a review of the SAVANNA-GIS model and its inputs, outputs, and proposed means of implementation. The model discussions would serve to integrate and codify various components of the overall assessment. The IMAS would evolve through each workshop. The problem analysis would include identification of critical information needs, how component studies would fit into the broader IMAS, methods of transferring information among collaborators, and methods of building capacity through education, outreach and stakeholder participation. Feedback from other team members, and from invited African stakeholders and collaborators would be elicited throughout the workshops.

The assessment team formation process would produce several results. By the end of the process we anticipated that the problem model and the IMAS would have received considerable thought, input from stakeholders. The IMAS would have evolved to a stage where it could be conceptually implemented at selected study sites.

Progress and Possible Future Activities

Model Framework

The conceptual framework for the IMAS evolved considerably during the assessment process. The principle objective of the IMAS is to inform stakeholders of the consequences of their activities. It should guide decision making with accurate and reliable information on interactions between livestock and wildlife. To accomplish this, the model must quantify the costs and benefits of alternative management scenarios to each stakeholder. The IMAS should identify areas of misunderstanding, and empower stakeholders with information. A user interface between the model and stakeholders must be developed, scenarios of each of the stakeholder's preferred actions would be simulated, and alternatives could then be provided for stakeholders to meet their needs and resolve their conflicts.

The IMAS will be comprised of a process of empirical and data-based assessment procedures, linked to computer-based procedures, and landuse planning and analyses. Research will be conducted both within the scope of our project, and by other parallel projects funded externally. Assessments will be made based upon modeling, and participatory involvement from stakeholders at the community level. The results of the assessment would be used in an implementation by the end users such as land managers, and policy analysts. The results of the implementation would be then evaluated at the community level, and the assessment would be modified accordingly. Inventory and monitoring studies will be vital to the IMAS, both as information inputs, and as measurements of results.

The interaction between stakeholders and the IMAS will inform stakeholders of impacts scenarios consistent with their goals and objectives. Scenarios will be created and encoded as both spatial and aspatial data that can be evaluated by the IMAS, model, or other analyses tools. The scenarios will be played out in the IMAS and the results returned to the stakeholders. Multiple scenarios can be compared and evaluated relative to different objectives.

The IMAS will be used in an iterative process of conflict resolution and risk analysis, the goal of which is to converge on a solution that all stakeholders can accept. The IMAS will be used to show the conflicts, risks, costs, and benefits to each stakeholder of multiple versions of their proposed solutions. The solutions would then be revised with the aim of converging on a solution that is most acceptable to all stakeholders.

The benefits of the proposed IMAS for each of 7 potential study sites were evaluated by regional participants at Workshop II. For all 7 of the sites, it was agreed that the IMAS would be beneficial by 1) improving the policy making process by serving as a focal point for problem evaluation and expanding the range of choices, 2) providing a capability to predict responses to drought and disease, 3) helping to distinguish effective short and long-term management strategies. At six of the sites it would be beneficial to 4) suggesting best spatial-temporal animal movement patterns to minimize disease risk, and 5) putting pastoralists on an equal footing with policy makers. At four sites, the IMAS would 6) be useful to a land planning agency, and 7) assist individual land managers to evaluate alternative land management strategies.

Ecology

A range ecology working group (Workshop I) decided that it's primary goals would be to: 1) develop a response data base for management practices, 2) use the IMAS model to predict and evaluate impacts of drought, 3) develop a process for determining forage balance, 4) facilitate agreement for range site descriptions, 5) develop a monitoring system and, 6) determine desired future condition for each site. Outputs of the range ecology submodel would include maps of resources, resource utilization, resource users, nutrient concentrations, and vegetation change. Additionally, the model would provide diagnostic and prognostic output on progress toward future change.

Biodiversity assessments were also recommended. The goal is to provide increased monitoring capabilities for plant and animal species in relationship to site factors and levels of livestock and wildlife utilization. This would require field work, data analysis, and synthesis.

The range ecology and wildlife working group of Workshop II also identified the following desirable outputs: 1) maintain a viable ecosystem, 2) maintain a balance between browsers, grazers and predators, 3) reduce the probability and extent of disease transmission, and 4) predict land use change and its impact on wildlife and livestock distribution and movements, and therefore numbers and composition. The model would need to consider distribution, composition and abundance of plants, animals, and human land use, and how these respond to rainfall, ground water, soil, and climate change.

Disease

Wildlife and livestock utilize the same pastures and migrate extensively based on the availability of feed resources, breeding patterns and market forces. Co-existence of wildlife and livestock populations provide conditions that are favorable for transmission of infectious agents from some species of wildlife to cattle, sheep, goats and camels. Some species of wildlife, such as the wildebeest and the African buffalo, can remain unapparently infected with viruses and protozoa and are reservoirs of disease agents that cause diseases with high morbidity and mortality in cattle. The major diseases of wildlife that constrain livestock productivity in the areas adjacent to the designated game parks and reserves include wildebeest-derived malignant catarrhal fever, corridor disease (theileriosis), trypanosomosis, foot-and-mouth disease and rinderpest. Though rinderpest is principally maintained in cattle, epidemics with high mortality have been observed to occur in different species of wildlife. The interactions of livestock and wildlife responsible for cycles of rinderpest outbreak need to be assessed.

The preliminary components of a disease submodel and assessment system were identified. The model would represent factors affecting disease prevalence, including presence of the agent, host resistance, animal movement, and transmission. Weather would influence each of these three factors, either directly or indirectly. Social-cultural policy affects host resistance via control measures, and animal movement. A participatory rapid appraisal would be conducted in the selected areas to assess the community perceptions of wildlife as a source of diseases to livestock and also to prioritize the importance of such diseases. Existing data on wildlife associated diseases of cattle would be collated (from PRA, published data, KWS and from veterinary services) and determination would be made of the applicability of such data to develop the output. Epidemiological studies would be conducted to determine important disease-host-vector dynamics within the wildlife and livestock populations.

Human Ecology and Economics

It was clear throughout the assessment process that a human ecology and economics component needs to be developed, which would provide indicators of human welfare at the individual (household), and community levels. These indicators would be measured in energetic and monetary units. Individual-level variables would include income, food energy, population density, and proximity of human settlements. Household and community-level economics submodels should be developed to link to the other components of the IMAS.

Gender issues must be addressed explicitly in these studies. For example, studies would be conducted of gender-related benefits from household cash production from gardens, livestock, hunting, and backflow of tourist revenues.

Human welfare and livestock production have been identified as two of four major objective functions for the IMAS. Both of these relate to the more global objective of using the IMAS to increase food security, as opposed to food production. Food security might be defined as the ability of households to access a reliable and adequate quantity and quality of food, produced by themselves or obtained from markets or social support systems. Indicators of food security include household income, market access, access to land, water, and grazing resources, access to services (health, education, infrastructure), and health and nutritional status. The latter would be monitored through measurements of demography, disease incidence, and energy and nutrient consumption rates.

Sites

It was decided that the IMAS will be applied to a select few (3-5) intensive study sites during the first phase of the project. Potential sites were suggested at Workshop I. A list of site selection criteria was generated, to be used for ranking. The criteria included: presence of pastoralists; wildlife populations; conservation value; representativeness of larger areas; contrasting treatments, environments, policies; data availability; potential for partnerships; likelihood of success, impacts; financial and time constraints; and others. The top sites (in no particular order) included Laikipia District (Uaso Ngiro River Basin, Kenya), the Greater Serengeti Ecosystem (Tanzania), Kajiado District (Kenya), Karamoja District (Uganda), Turkana District (Kenya), Tsavo N.P. - Mkomazi N.P. (Kenya-Tanzania), Lake Mburo N.P. (Uganda), and Tarangiri N.P. -Simanjiro Plains (Tanzania). Regional inputs on scientific priorities, policy and management issues, IMAS benefits, and human welfare and development for each site were obtained at Workshop II.

A study design was identified, in which sites would be selected from across the range of land use intensity and spatial scale - from no livestock use, to spatially extensive pastoralism, to group ranches, and to finally small private land holdings. Site selection will depend upon demand for the proposed IMAS by end users.

Based upon positive support received from local representatives, three sites were chosen as our initial intensive study sites. All three of the sites have good data bases and high likelihood of successful IMAS implementation. It was decided that the Ngorongoro Conservation Area portion of the Greater Serengeti would constitute the best site at the more traditional end of the spectrum. Kajiado would contain a range of situations from large group ranches to small holdings. Lake Mburo would be a good example of a developing ranching and dairying scheme adjacent to a national park. Other sites of high promise include Laikipia (Kenya), Tarangire (Tanzania), Loliondo (Tanzania), Karamoja-Turkana (Uganda-Kenya), Awash (Ethiopia), and Katavi-Rukwa Valley (Tanzania). Selection among these sites will require further problem analysis, and indications of demand from the potential end users.

The Greater Serengeti Ecosystem (GSE) is a large region comprised of Serengeti National Park, Ngorongoro Conservation Area (NCA), Loliondo and Maswa Game Reserves, and Maasai Mara National Park (Kenya). Serengeti National Park and Masai Mara are parks with a vast wildlife population, well known for the migratory wildebeest herds. Land use pressures in the Serengeti Regional Ecosystem of Tanzania are growing rapidly (Mbano et al. 1995). Pastoralists in the NCA share grazing areas with Serengeti migratory wildebeest and zebra. Pastoral population densities in the NCA nearly tripled 1966-1988 (Perkin 1995). Many have settled, and are cultivating small plots. In the adjacent Maswa Game Reserve, increases in agro-pastoral populations have led to increased poaching, unplanned fires, and illegal tree-cutting (Campbell and Hofer 1995, Mbano et al. 1995). There is increasing conversion of pastoral rangelands to large-scale cultivation, because the economic incentives of leasing out to farmers exceed those accruing from conservation (Norton-Griffiths 1995). Since 1987 when the Tanzania-Kenya border reopened, tourist numbers have increased nearly 3-fold and many new lodges and camps have been built (Sinclair 1995). Vehicular traffic has increased markedly (Perkin 1995, NCAA 1995).

Kajiado District, Kenya is the site of one of the great experiments in international livestock development. In the late 1960's, the Government of Kenya requested, and the World Bank implemented the Kenya Livestock Development Program (KDLP), a district-wide project aimed at promoting commercial livestock production among the Maasai herders of Kajiado. The principal instrument was land adjudication; providing freehold title to groups of Maasai who organized themselves into group ranches. The most prominent effect of group ranch formation, reduction of the spatial scale of grazing land exploitation. Though group ranch formation has achieved some of the original program objectives, the unanticipated impacts in Kajiado probably outweigh those envisioned by the KLDP planners. On the positive side, livestock production has increased somewhat over the past 30 years, but not as much as anticipated. Likewise, range degradation is thought to be limited in scope and patchy, while wildlife abundance has not plummeted as has occurred in much of Kenya. On the negative side, cattle/ human ratios have gone from about 14/person in the early 60's to about 3/person in the late 1980's. This is mainly a result of human population growth while livestock numbers have fluctuated or expanded rather slowly. Recent government policies are encouraging subdivision and privatization of the group ranches.

Lake Mburo, location in southwest Uganda, is a National Park adjoining group ranches. It is biologically diverse and wildlife are common. Originally, pastoralists and wildlife inhabited the area in and around Lake Mburo National Park. During the Colonial Era, controlled hunting was instituted. During the 1960s, the what is today designated as the National Park was designated as a game reserve, and a livestock improvement program and ranching scheme were introduced in the surrounding areas. In 1983, Lake Mburu was designated as a National Park, and the pastoral inhabitants were removed from the land, to be "resettled" on the adjacent ranches. There has been a steady decline in wildlife numbers, especially in the areas immediately outside the National Park since the 1980s. This has been attributed largely to hunting (and poaching) and loss of habitat. Given the Park boundaries, problems have been created with regard to access to land and importantly, access to water for both people and livestock has become limited. The movement of wildlife from the Park to ranch areas has caused problems. There is an issue of disease transmission between humans, livestock, and wildlife, specifically foot and mouth, brucellosis, tuberculosis, and tick borne diseases. Finally, there is pasture degradation due to overstocking of livestock and a lack of mobility inducing overgrazing.

Training, Communication, and Information

Training will be essential to the success of this program. At least three American students should receive graduate training, providing increased assessment capacity both during this project, and after it ends. East African students and technicians will also be trained. The goal would be to develop regional capacity to implement the IMAS after the project ends.

We envision a core level of technicians and scientists operating the DSS at universities and government agencies, as well as within our project. These experts would train technicians to run lower level analyses on the IMAS at field offices, national park offices, and NGO's. Results would be communicated to pastoralists and landowners through workshops, seminars, and community information centers.

Objective Matrices
 

Objective I: Develop a computerized decision-support system that will improve prospects for balancing the needs of pastoralists and wildlife conservation in East Africa.
 

Objective II: Develop and implement a demand-driven, integrated assessment and monitoring system for improving pastoral-wildlife interactions at site and policy levels.

Outputs	Impacts	End Users	Actions Required	Team Members	Time to Completion	Developmental Relevance
Implementation at 3 intensive sites (Ngorongoro, Kajiado, Lake Mburo). Possible later implementation at 3-6 additional sites (Laikipia, Tarangire, Karamoja, Turkana, Rukwa Valley, Awash N.P.).    Assessments, and useful and accessible information.	1) Use by organizations which influence, or aim to influence natural resources.  2) Increased capacity to assess alternative management and planning scenarios 3) Increased sense of empowerment arising from better information 4) Improved relationships between conflicting land uses 5) Increased food security for pastoralists  6) Wildlife conserved, ecosystem integrity maintained	1) Site management and planning entities. 2) Government organizations and non-government agencies - eg. pastoral, wildlife, livestock, parks. 3) Pastoralists, or their representatives - traditional, private, commercial, ranching, grazing associations. 4) Tourism industry. 5) Educational institutions. 6) Researchers 7) Media	1) Stakeholder inputs via workshops and forums.  2) Conduct field appraisals of ecology, wildlife, disease, livestock production, human welfare. 3) Assemble GIS data bases 4) Assemble non-spatial data bases 5) Establish information centers 6) Implement monitoring scheme	All U.S. and ILRI.    Host-country team members at corresponding sites.   U.S. site leaders: Kajiado- Ellis NCA - Galvin Mburo - Magennis   In-country site coordinators: Kajiado: Rainey NCA: Moehlman Mburu: Mugishu	2-3 years per 3 sites. Six years total.	Implementations accross sites with a diverse array of developmental problems, representing those encountered in pastoral development throughout the region.
Uniform Method for Classifying Ecological Range Sites in Region	1) Adoption by government and non-government entities  2) Improved ability to conduct local through regional level assessments 3) Increased coordination among planning entities	1) Government and non-government entities. 2) Researchers	1) Background research 2) Regional workshop 3) Document preparation	D. Child  L. Rittenhouse R. Kidunda J. Kinyamario, F. Banyikwa	3 years	Provides a widely accepted set of criteria for measuring state, and progress to management goals throughout region
Recommendations for management of livestock and wildlife to minimize disease. Recommendations for disease control	Decreased losses due to disease.  Increased livestock productivity.	Pastoralists Farm managers, and owners. National Park personnel Gov't agencies NGOs.	Literature survey Obtain databases Identify questions and conditions Run Savanna model with disease component. Analyze and interpret results	J. Ellis M. Coughenour J. DeMartini P. Rwambo J. Grootenhuis (Others)	5 years	Enhances wildlife conservation and improves human welfare in pastoral development efforts throughout the region.

 
 

United States
 

Name	Title	Affiliation
Child, Dennis	Department Chair  Professor	Colorado State University, Rangeland Ecosystem Science Department
Coughenour, Michael	Senior Research Scientist,  Associate Professor (Affiliate) Advising Faculty	Colorado State University, Natural Resource Ecology Lab, Rangeland Ecosystem Science Department, Graduate Degree Program in Ecology
Davis, Bob	Senior Associate	Colorado University  Institute of Behavioral Science
DeMartini, James	Professor	Colorado State University, Pathology Department
Ellis, James	Senior Research Scientist  Associate Professor (Affiliate) Advising Faculty	Colorado State University, Natural Resource Ecology Lab, Rangeland Ecosystem Science Department, Graduate Degree Program in Ecology
Galvin, Kathleen	Senior Research Scientist  Assistant Professor Advising Faculty	Colorado State University, Natural Resource Ecology Lab, Department of Anthropology, Graduate Degree Program in Ecology
Magennis, Ann	Assistant Professor	Colorado State University, Department of Anthropology
O'Malley, E.B.	PhD. Candidate	Colorado University, Anthropology Department
McCabe, Terry	Associate Director,  Assistant Professor	Colorado University Institute of Behavioral Science
Rittenhouse, Larry	Professor	Colorado State University, Rangeland Ecosystem Science Department

 

Kenya
 

Name	Title	Affiliation
Barrow, Ed	Community Conservation Advisor	African Wildlife Foundation
Grootenhuis, Dr. Jan	DVM	Consultant
Kinyamario, Jenesio	Professor	University of Nairobi
Kruska, Russ	Geographic Information Systems Specialist	International Livestock Research Institute
Mukhebi, Adrian	Senior Agricultural Economist	International Livestock Research Institute
Rainy, Mike	Ecotourism and Education  Instructor and Consultant	Bush Homes of East Africa
Reid, Robin	Senior Ecologist	International Livestock Research  Institute
Rwambo, Paul	DVM, Resident Scientist	Kenya Agricultural Research Institute
Thornton, Phillip	Agricultural Economist	International Livestock Research Institute

 

Tanzania
 

Name	Title	Affiliation
Banyikwa, Feetham	Adjunct Faculty  Research Associate	University of Dar es Salaam Syracuse University
Kidunda, Rashid	Assistant Professor	Sokoine University
Moehlman, Patricia	PhD., Research Scientist  Equid Specialist Group Leader	Serengeti Wildlife Research Institute, IUCN
ole Ikayo, Francis	Director	Inuyat e-Maa (Maasai Pastoralists)

 

Uganda
 

Name	Title	Affiliation
Acen, Joyce	Management Systems Officer	Uganda Ministry of Tourism, Wildlife, Antiquities
Else, Dr. James	DVM, Institutional Development Advisor	Uganda Ministry of Tourism, Wildlife, Antiquities
Mugisho, Arthur	Community Conservation Coordinator	Uganda Ministry of Tourism, Wildlife, Antiquities

University of Colorado
Environment and Behavior Program
Institute for Behavioral Science
Campus Box 468
Boulder, Colorado 80309
Phone: 303-492-0443
Fax: 303-492-1231

International Livestock Research Institute
P.O. Box 30709
Nairobi, Kenya
Phone: 254-2-632311
Fax: 254-2-631499

Kenya Agricultural Research Institute
P.O. Box 57811
Nairobi, Kenya
Phone: 254-2-583304
Fax: 254-2-583344

University of Nairobi
Department of Botany
P.O. Box 30197
Nairobi, Kenya
Phone: 254-2-447865
Fax: 254-2-446141

University of Dar es Salaam
Institute of Resource Assessment
P.O.Box 35097
Dar es Salaam, Tanzania
Phone: 255-51-43393
Fax: 255-51-49039

Sokoine University
Faculty of Agriculture
Department of Animal Science and Production
P.O. Box 3004
Morogoro, Tanzania
Phone: 255-56-4617
Fax: 255-56-4562/4088/4988

Uganda Ministry of Wildlife and Antiquities
P.O. Box 4241
Kampala, Uganda
Phone: 256-41-232971/346288
Fax: 256-41-234054
 
 

An international team was formed to oversee development of a system to assess livestock-wildlife interactions in pastoral ecosystems of East Africa. We introduced our project goals, explained how we intend to meet these goals, and elicited insight and advice from participants on what the assessment system must consider to be relevant to the needs of pastoralist, resource managers, and policy makers. Participants shared their experiences with pastoral-wildlife systems, and specified the types of information that would be useful from their perspectives. We then focused on the conceptual development of the decision support system. Other activities included the identification and evaluation of research sites, and breakout groups which focused on the identification of goals, objectives, activities, constraints, inputs, and assumptions for modeling, biodiversity, range ecology, human ecology, and disease. Participants included range and wildlife ecologists, wildlife disease experts, anthropologists and social scientists, and experts and stakeholders in pastoral-wildlife interactions.
 
 

Representatives from governments, NGO's, donor agencies and universities from Kenya, Tanzania, Ethiopia, Uganda joined the assessment team to set regional priorities for the project. First, there were several demonstrations of how technology (Savanna, ILRI GIS analyses, SCOP project) has been used to address various science and management questions and how the results have (or could) influence policy. This demonstrated that technology is available to address the objectives of this USAID-CRSP project: to provide an integrated assessment of the implications of alternative management and policy decisions for pastoral development and wildlife conservation. Second, people from several organizations in the region presented their policy and management perspectives within the context of the integrated assessment system. Within the context of regional perspectives, workshop attendees identified scientific priorities, policy priorities, and priorities for the needs and welfare of pastoral peoples. Working groups were formed to discuss specific sub-problems of the integrated assessment. They included: human ecology and economics; disease interactions, range ecology, livestock production and pastoral needs, range ecology and wildlife; and regional analyses and policy. Study sites were selected, and site-specific problems, data needs, and team members were identified.
 
 

Arhem, K. 1985. Pastoral Man in the Garden of Eden. The Maasai of Ngorongoro Conservation Area, Tanzania. Uppsala Research Report in Cultural Anthropology. Uppsala Sweden.

Byrne, P.V., J.G. Grootenhuis, J.C. Staubo. The economics of living with wildlife in Kenya. World Bank, Washington D.C. (Grootenhuis et al. 1991, op. cit.).

Cambell, K., and H. Hofer. People and wildlife: spatial dynamics and zones of interaction. Chap. 25 in Sinclair and Arcese, 1995.

Coughenour, M. B. 1991. Spatial components of plant-herbivore interactions in pastoral, ranching, and native ungulate ecosystems. J. Range Manage. 44:530-542.

Coughenour, M. B. 1992. Spatial modeling and landscape characterization of an African pastoral ecosystem: a prototype model and its potential use for monitoring drought. pp. 787-810 in: D.H. McKenzie , D.E. Hyatt and V.J. McDonald (eds.). Ecological Indicators, Vol. I. Elsevier Applied Science, London and New York.

Coughenour, M.B. 1993. The SAVANNA Landscape Model - Documentation and Users Guide. Natural Resource Ecology Laboratory, Colorado State University, Ft. Collins CO.

Coughenour, M. B., J. E. Ellis, D. M. Swift, D. L. Coppock, K. Galvin, J. T. McCabe, and T. C. Hart. 1985. Energy extraction and use in a nomadic pastoral ecosystem. Science 230:619-624.

Cumming, D.H.M. 1991. Wildlife conservation in African parks: progress, problems, and prescriptions. In: D.M. Lewis and L.A. Carter (eds). A Heritage to Share: African Approaches Towards Conservation. WWF, Washington D.C.

Ellis, J. and K. Galvin. 1994. Climate patterns and land use practices in the dry zones of Africa. Bioscience 44: 340-349.

Ellis, J. E., M. B. Coughenour and D. M. Swift. 1993. Climate variability, ecosystem stability, and the implications for range and livestock development. pp. 31-41 in R.H. Behnke Jr., I. Scoones and C. Kaven (eds.), Range Ecology at Disequilibrium: New Models of Natural Variability and Pastoral Adaptation in African Savannas. Overseas Development Institute, London.

Galvin, K.A. 1995. Conservation policy and human nutrition in Ngorongoro Conservation Area, Tanzania. Invited paper presented at the session on Natural Resource Management in Eastern and Southern Africa: Issues of Sustainability and Conservation at the annual meetings of the Society of Applied Anthropology (Abstract).

Grootenhuis, J.G., S.G. Njuguna, and P.W. Kat. 1991. Wildlife Research for Sustainable Development. Proceedings of an International Conference Held in Nairobi, April 22-26, 1990. KARI, KWS, NMK, Nairobi, and Pride Publishing, Nairobi.

Hansen, A. J., S. L. Garman, B. Marks, and D. L. Urban. 1993. An approach for managing vertebrate diversity across multiple-use landscapes. Ecological Applications 3:481-496.

Harris, L. D. 1984. The fragmented forest: island biogeographic theory and the preservation of biotic diversity. Univ. of Chicago Press, Chicago, IL. 211 pp.

Mbano, B.N.N., R.C. Malpas, M.K.S. Maige, P.A.K. Symonds, and D.M. Thompson. 1996. The Serengeti Regional Conservation Strategy. Pp. 605-616 In: Serengeti II: Dynamics, Management, and Conservation of an Ecoystem, A.R.E. Sinclair and P. Arcese (eds.). University of Chicago Press. Chicago.

Ottichilo, W.K., J. Grunblatt, M.Y. Said, and P.W. Wargute. 1997. A national perspective - Kenya case study: Wildlife/livestock population trends and protected areas. (unpublished report).

Norton-Griffiths, M. 1995. Economic incentives to develop the rangelands of the Serengeti: implecations for wildlife conservation. pp. 588-604 In: Serengeti II: Dynamics, Management, and Conservation of an Ecoystem, A.R.E. Sinclair and P. Arcese (eds.). University of Chicago Press. Chicago.

NCAA, 1996. Ngorongoro Conservation Area Authority Draft General Management Plan. Tanzania Ministry of Tourism, Natural Resources, and Environment.

Perkin, S. 1995. Multiple land use in the Serengeti Region: The Ngorongoro Conservation Area. pp. 571-587 In: Serengeti II: Dynamics, Management, and Conservation of an Ecoystem, A.R.E. Sinclair and P. Arcese (eds.). University of Chicago Press. Chicago.

Pratt, D.J. and M.D. Gwynne. 1977. Rangeland Management and Ecology in East Africa. Robert E. Krieger Publishing Company, Huntington, N.Y.

Rainy, M.E. and J.S. Worden. 1997. Summary of wildlife and domestic herbivore population trends in Kenya's rangeland districts. (unpublished).

Scott, J. M., F. Davis, B. Csuti, R. Noss, B. Butterfield, C. Groves, H. Anderson, S. Caicco, F. Derchia, T. C. Edwards, J. Ulliman, and R. G. Wright. 1993. Gap analysis - a geographic approach to protection of biological diversity. Wildlife Monographs 1-41.

Sinclair, A.R.E. 1995. Serengeti past and present. Chap. 1, in Sinclair, A.R.E. and P. Arcese. 1995. Serengti II - Dynamics, Management, and Conservation of and Ecosystem. University of Chicago Press, Chicago. 665 pp.

Western, D. 1982. Amboseli National Park: enlisting landowners to conserve migratory wildlife. Ambio. 11: 302-8.