- Open Access
Applying an ecosystem services approach to support land-use planning: a case study in Koboko district, Uganda
© Speziale and Geneletti; licensee Springer. 2014
Received: 23 December 2013
Accepted: 16 March 2014
Published: 4 May 2014
This paper presents a case study of the application of an ecosystem services framework to support land-use planning. The district of Koboko, in northwestern Uganda, formed the study area. The overall purpose was to improve the understanding of links between human actions, their impact on ecosystems and the services they provide, and, ultimately, consequences for human wellbeing. The framework developed by the Millennium Ecosystem Assessment was used as a main conceptual reference.
The analysis comprised four main stages: (1) identification of the main ecosystems and their services; (2) analysis of human wellbeing components, focusing on poverty issues; (3) assessment of the interrelations between ecosystem services and human wellbeing; (4) identification of the main drivers of change and development of future spatial development scenarios.
The scenarios were developed starting from the main drivers of change and show the dependency of human wellbeing on ecosystems services, identifying how spatial development strategies can affect that dependency. Quantitative information was not easily available, therefore scenario development made use of both quantitative and qualitative analysis. The results of qualitative and quantitative approaches were consistent, proving the strength and flexibility of the method.
Scenario results aim to provide guidance for local government land-use planning, focusing on the promotion of sustainability through ecosystem services preservation. The spatially explicit analysis illustrated how different policies affect urban development; it was clearly shown that even with very different demographic outcomes, the impact of sectorial policies guaranteed a good level of suitability for new residential areas.
The importance of ecosystem services for human wellbeing is well established. In the last ten years there has been a significant increase in scientific papers focusing on ecosystem services ( Fisher et al.2009), fuelled by the Millennium Ecosystem Assessment, which pointed out the many interrelations between nature and society. Thanks to the Millennium Ecosystem Assessment, many studies were carried out to explore these issues in more detail, with the overall aim of promoting sustainable development. However, how the analysis on ecosystems and human wellbeing can be directly used for land-use planning purposes has not been developed to the same extent ( Daily et al.2009). A comprehensive approach that includes impacts on ecosystem services would be a significant improvement for promoting sustainability in spatial planning decisions. This is especially true considering that sectorial policies and strategies should be well integrated within a spatial plan and that an ecosystem approach provides the comprehensive framework needed to reach this objective.
This paper presents a case study related to the application of an ecosystem services approach to supporting land-use planning. The study aimed at improving the understanding of the links between human actions, the impacts on the state of ecosystems and the services they provide, and ultimately the consequences for human wellbeing. The case study area is the district of Koboko, a poor yet fast-developing region in Uganda.
The district has a mixture of ethnic groups, with inhabitants also coming from neighbouring countries, such as the Democratic Republic of Congo, Southern Sudan and even Somalia. Furthermore, a massive number of refugees, who settled in 1995 in the northeastern area of the district, caused an extreme acceleration in the depletion of the ecosystems, especially in forest areas (United Nations Environment Programme, 2007). Annual population growth rate has been estimated at 5.6%, which is higher than the national average (Uganda Bureau of Statistics, 2010). Fertility rates in Koboko are among the highest in the country (the average number of children per woman is 7, according to UBS 2010).
Key environmental concerns for the district include (National Environmental Management Authority, 2006; Uganda Bureau of Statistics, 2002. ; Rogers et al2006; Kibreab, 1997; Byakola and Mukheibr, 2009):
Poor health and sanitation,
Poor waste management,
Wetland and river bank degradation,
Poor land management,
Conflict with protected area management.
Over the last 15 years, land use has changed significantly, owing to deforestation processes and growth of the urban area. In addition, croplands have expanded into areas previously covered by forests. The nearly complete conversion of wetlands, forests and savannah woodlands into agricultural farms has negatively affected the supply of provisioning ecosystem services, such as water and firewood.
The town of Koboko has become a very important player in the district dynamics, particularly because the current population is 51,300 people. The average household size is 7.3 members and the average household density has been calculated as about 211 people per hectare. The high population is also due to the immigration of refugees and returnees from Sudan and neighbouring districts (Ministry of Local Government, 2007). District-scale dynamics have proven to be deeply connected to urban processes, especially because urban features are still very close to rural lifestyles. Consistently with district trends, agriculture is also a major activity in the town. The most common crops are tobacco and cassava; their production is concentrated in the villages, away from the town centre (Ministry of Local Government, 2007).
The work developed for the Millennium Ecosystem Assessment was taken as an inspiration for the study: the interrelations between ecosystem services and human wellbeing were analyzed and spatial development scenarios were obtained, thanks to the drivers of change identified. The purpose was to point out the closed link between human wellbeing and ecosystem services, in order to provide guidelines for policy-makers.
Assessment of the interrelations between ecosystem services and human wellbeing;
Identification and evaluation of the main drivers of change;
Development of future spatial scenarios, combining of qualitative and quantitative information;
Comparison of scenarios through social and environmental indicators.
Assessment of the interrelations between ecosystem services and human wellbeing
The links between ecosystems and human wellbeing were identified following the framework proposed by the Millennium Ecosystem Assessment. The main ecosystems were identified with a land-use map of the study area generated by satellite image processing. The image analyzed was a Landsat frame from 2005 (the most recent available) with a pixel resolution of 60 m. Image processing was performed using supervised classification, with GRASS GIS software. The result of the classification was then compared with data collected by field surveys using a global positioning satellite. The following land-use types were classified: woodland, degraded woodland, grassland, farmland and urban areas (Speziale, 2012).
Socio-economic information about human wellbeing and poverty levels was collected by reviewing available reports and other literature (Uganda Bureau of Statistics, 2010; International Institute for Sustainable Development, 2005; Bahiigwa and Muramira, 2001; Emwanu et al.2003; Robinson and Pozzi, 2011; Rogers et al.2006), as well as by field surveys. A total of 100 households were sampled randomly throughout the four wards of the town of Koboko and data on housing equipment, water and sanitation, and demographics were collected. The collection was carried out together with the town council staff in order to facilitate communication and clarify doubts about the questions asked. The main vulnerability factors were represented by food security, limited livelihood opportunities and firewood shortage. Other critical factors, particularly considering the high rate of population growth, were poor farming methods, land fragmentation and degradation of forest land (International Institute for Sustainable Development, 2005; National Environmental Management Authority, 2006).
Identification and evaluation of the main drivers of change
Once the socio-ecological system was studied, it was possible to select the main drivers of change using expert opinions, a literature review (Uganda Bureau of Statistics, 2010; United Nations Environment Programme, 2007; Arua DLG, 2004) and field work. The main indirect drivers considered are demographic change and economic change, whereas direct drivers are land-use change, climate change and overexploitation of natural resources. To develop a solid basis for the elaboration of future scenarios, the selected drivers of change were analyzed on a quantitative and/or a qualitative level (depending on the specific driver). For example, land-use change was analysed using GIS software, based on supervised classification of Landsat images (from 1995 and 2005). Demographic trend was projected by analyzing the census data available for the last 20 years.
(The evaluation of the main drivers of change was used to classify their degree of impact on ecosystems and human wellbeing and their current trend, in order to identify main concerns for future development. Consistent with the findings of Ash et al.2010), the impact of each driver on the ecosystems included in the analysis and the components of human wellbeing was assessed using a low-medium-high scale, while the trend was described as decreasing, continuing, increasing or rapidly increasing.
Development of future spatial scenarios
(Scenarios were developed starting from the main drivers of change and showing the dependency of human wellbeing on ecosystems services in order to identify how spatial development strategies can have an effect on that dependency. Following the method described in Ash et al.2010), within the main scenarios developed, a quali-quantitative approach was adopted: the qualitative assessment was carried out at the district scale, while quantitative information was applied at the urban scale. This combination also helped in checking the consistency of qualitative outcomes against quantitative ones in the same scenario.
Among the drivers taken into account for this analysis, particular attention was paid to policies, because in spatial planning they have a significant influence. In particular they have been divided in spatial policies, meaning all the actions that can lead to a spatial development, and sectorial policies, or the political choices that can help finding substitutes for ecosystems services. For each driver identified as a critical uncertainty, the possible outcomes were unfolded and scenarios were obtained by combining the drivers’ outcomes. Once the scenario set was completed, the most significant were selected: scenarios with outcomes that conflicted or contradicted each other were removed, because they were considered less likely to happen than the others. Moreover, some led to similar futures, thus making it harder to appreciate the different impacts: in this case, one scenario was chosen for analysis and considered representative of the others. Overall from the 24 possible combinations, four were chosen for further analysis and these were compared against each other using a series of indicators. The selected scenarios were named Far West, Eating Broth with a Fork, Full Steam Ahead and Atlantis. The following key quantitative indicators were selected to summarize the impacts of the different scenarios: deforestation rates, demographic trends and farmland demand.
The quantitative assessment proposed for the urban scale included assumptions on household density, making the work twofold: on one hand, the focus was on urban growth (with the assumption that all nonurban land is available for growth), while simultaneously suitability factors were identified, weighed and merged, to obtain suitability maps. To estimate urban expansion, the outcomes of demographic drivers of change considered for the qualitative analysis (without considering refugee flows) were taken into account, so that the resulting outcomes were consistent with the scenario analysis undertaken at a district scale. Furthermore, the impacts of an increase in household density were compared with those linked to scenarios where the density does not change.
The suitability model was based on critical factors that influence the likelihood of a certain area to be urbanized: these factors were implemented using GIS software, to perform a spatial multicriteria evaluation. In this case, the choice was based on literature data about which factors are more attractive in contexts such as the one studied (Geneletti, 2012; Mendoza et al.2009):
Distance from city centre,
Distance from rivers and streams,
Distance from roads,
Distance from water points,
Distance from woodlands,
A number of factors were not considered in this analysis, such as industrial land expansion, road and bridge construction projects, ownership and land tenure, succession (for example, commercial replacing residential land use), affordability, political and commercial factors and urban development in the proximity of the area. The basic assumption considered was that all nonurban land is available for expansion.
The suitability factors were standardized and then weighted, according to priorities consistent with the critical uncertainties’ outcomes in the selected scenarios. Concerning the weighting phase, not only was it taken into considerations that factors do not have the same importance in one scenario, but also that the same factor might weigh differently in different scenarios. An example is the weight assigned to distance from the forest: when wood is supplied following an organized pattern, people do not have to gather their own firewood from woodlands, and therefore living close to woodlands is not an influencing parameter. The suitability maps thus obtained were then applied to urban expansion, combining the two aspects of the evaluation.
Comparison of scenarios through social and environmental indicators
To analyse impacts of the different district-scale scenarios, social and environmental indicators were chosen based on the assessment of ecosystems and human vulnerabilities in the study area and generally used in scenario evaluations (Alcamo and Henrichs 2008, Jaeger et al.2007, National Environmental Management Authority, 2008). Scenarios were then classified based on their performance according to the single indicator, using ‘traffic light’ options: red represented a bad performance, green a good one and orange a medium one. Once scenarios were ranked, a quantitative evaluation of the most critical indicators was undertaken: the impacts on land demand for farming purposes and firewood and charcoal demand (Knöpfle, 2004) showed, for each scenario, whether the natural resources of the area were sufficient to support the population’s needs.
From the urban-scale analysis, a series of maps showing, for each scenario, the demand for new urban area and the associated suitability was obtained. Hence, for each map, it was possible to extract suitability thresholds associated to each scenario considered and provide conclusions based on scenario assumptions.
Results and discussion
Interrelations between ecosystem services and human wellbeing and drivers of change
(Overexploitation and pollution were analyzed together, because they presented similar outcomes for human wellbeing: with reference to the constituents of wellbeing underlined in Ash et al.2010), security is the most affected part (because it undermines resources access) together with basic materials for health and a good life and (because they put access to clean water and food productivity at risk). However, their behaviour differs: analysis of ecosystems reveals that pollution mainly affects fresh water, while overexploitation impacts mainly on agro-ecosystems. This difference could be explained by considering that human wellbeing is a complex concept where many factors act together and therefore repercussions of ecosystems degradation are not linear or easily predictable.
Development of future spatial scenarios
This is the case of land-use change, urbanization, pollution and economic growth: if considered as stand-alone parameters, these drivers have a remarkable impact on future outcomes, but their impact also depends highly on policies’ directions. If governments do not improve their sustainable actions, some drivers will grow out of control and will therefore affect ecosystems and human wellbeing heavily. Proactive actions could be undertaken to promote the development of formal settlements and waste treatment facilities and to avoid overexploitation, so that the impacts of those drivers would be reduced.
As explained in the Methods section., to develop future scenarios, it was necessary to identify critical uncertainties; that is, drivers that have the biggest impacts on future development and at the same time have highly unpredictable outcomes. The trend of migration and the sudden flow of refugees (Kanyamibwa, 2007), the sectorial policies adopted to meet ecosystem needs and empower them and the governance aiming at improving human wellbeing meet the requirements, as shown in Figure 6. Therefore, these three are the drivers taken into account to develop future scenarios, with the addition of the demographic trend,a chosen because in fast-developing areas it greatly influences future outcomes despite its low uncertainty.
Outcomes of critical uncertainties selected for scenario development
Annual growth rate settles to 3.4% (national value) and migration fluxes are relevant.
Annual growth rate settles to 2.6% and migration fluxes are relevant.
Annual growth rate settles to 3.4% (national value).
Annual growth rate settles to 2.6%.
No spatial plan is developed.
No specific infrastructures are built.
No spatial plan is developed.
Electricity is provided through the national grid and the main road of Koboko town centre is tarred.
Spatial plan is developed.
Electricity is provided through the national grid, the main road of Koboko town centre is tarred and waste treatment facilities are built.
No special policies are developed.
Sectorial policies are developed to provide replacement for ecosystem services: wood plantations and support for development for better farming methods.
Outcomes for urban expansion analysis
Residential areas (ha)
Residential area increase (ha)
For the multicriteria analysis, however, suitability factors were standardized and then weighted according to the outcomes of special policies and sectorial policies considered for Far West (G1P1) and Atlantis (G3P2) (see Table 1). For this reason, two different standardization and weighting methods were carried out and the two suitability maps obtained were applied to the four urban expansion outcomes, thus providing six different scenarios. Nevertheless, following the same rationale as for district-scale scenarios, combinations between conflicting assumptions (for example, farsighted policies and uncontrolled increased in household density) were excluded from the analysis. Table 1 summarizes the analysis at the urban scale, showing the suitability threshold for each resulting scenario.
Considering also that the highest suitability threshold belongs to scenario C and that the same scenario matches the other requirements explained previously, it can be concluded that scenario C brings together the best conditions for human and environmental wellbeing.
Comparison of scenarios through social and environmental indicators
Overall, Full Steam Ahead improved the lifestyle of people involved in commercial and industrial activities, but for small farmers who practised subsistence agriculture there would be no positive impact. Atlantis empowers poor and marginalized groups, such as women, tends to increase equity regarding access to natural resources and, in general, improves human wellbeing.
The performance of Eating Broth with a Fork shows how farsighted policies can impact positively in the worst demographic picture. For a long-term perspective, where Atlantis is the scenario to aim at, Eating Broth with a Fork shows the effectiveness of human responses in the medium term. The indicators that do not perform very well are the ones heavily influenced by demography: change in population growth happens slowly and the effects of good governance appear only in the medium or long term.
Suitability thresholds for the six urban scenarios
A (Atlantis in urban area with D3a)
B (Far West in urban area with D3a)
C (Atlantis in urban area with D4a)
D (Far West in urban area with D4a)
E (Far West in urban area with D3b)
F (Far West in urban area with D4b)
In scenarios A and C, suitability thresholds are the same (see Table 3): priorities are related to providing firewood and charcoal, so that the remaining woodland can be preserved, and controlling household density, so that it is still possible for people to own the land essential for survival. To guarantee charcoal for urban demand, it is recommended that wood plantations be implemented in the surroundings.
In no-policies scenarios, such as B and D, people will tend to live close to rivers, roads and forests; in these places, population density will increase rapidly. It would be important that policies at least control and protect the forest reserve and avoid encroachments.
A quite risky strategy would be to let the household density increase: as shown in Figure 7, this would be a good outcome for land-use conversion, because less land is required for the same population, but concentrating the inhabitants in a small area can lead to several social problems. Moreover, people would still need to cultivate for subsistence, which would force them to cover a long distance every day.
This paper presented a method to support spatial planning and decision making when they aim at improving the sustainability of the relation between human wellbeing and ecosystem services. The method was consistent with the principles expressed in the Millennium Ecosystem Assessment and addressed the growing need to include ecosystem-based approaches in decision making and land-use planning (Daily and Matson 2008, Carpenter et al.2009, Fisher et al.2009, Hauck et al.2013).
The proposed method allows decision makers to see the consequences of environmental and sectorial policies and to understand how human actions are connected to the state of ecosystem services. Furthermore, it enables identification of the critical issues of the study area, to focus on the aspects that provide the biggest impacts on ecosystems and to identify additional studies needed to improve the analysis of the context.
Many studies were carried out to assess the interrelations between ecosystem services and human wellbeing, but less attention has been paid to the provision of information that can be directly used for land-use planning. This method provided a comprehensive approach,- from the analysis of the ecosystem services to the comparison of scenarios, to fulfil this aim and was applied to a case study in order to show how the analysis can provide guidance for land-use planning. Moreover, the analysis was designed to be flexible, given the study context: lack of data (especially statistical data and adequate satellite images) and tools was overcome through flexibility of the method and the comparison of qualitative and quantitative assessment to ensure that the robustness of the results compensated for the incomplete set of information available. The general approach followed can be replicated in contexts similar to Koboko district: rural areas of Uganda, and sub-Saharan Africa in general, that are characterized by poverty and strong dependence of the population on ecosystem services are good candidates for testing the method, adapting assumptions to specific environmental and social conditions.
The identification of the critical uncertainties was one of the most important steps in connecting the analysis of the linkages between human wellbeing and ecosystem services to the development of scenarios dedicated to land-use planning purposes. The choice of these uncertainties was based on literature, but also on the results of the framework assessing the interrelations between the human and the ecosystem spheres in the study area. To guarantee clear messages for decision makers, the number of key uncertainties was low: this allowed to show straightforward cause-and-effect chains and to point out the environmental effect of the different policies applied (Geneletti, 2012).
Moreover, the comparison between qualitative scenarios, at the district scale, and quantitative scenarios, at the urban scale, fully supported the belief that selected aspects of qualitative scenarios may be underpinned by numerical estimates and that the numerical estimates of a quantitative scenario may be bound together and explained using a consistent storyline. Notwithstanding the different scales between qualitative and quantitative analysis, the results have shown deep consistency, ensuring the quality of the scenarios developed.
The spatially explicit analysis demonstrated how different policies will affect urban development: it was clearly shown by the results of this method how even with very different demographic outcomes, the impact of sectorial policies guaranteed a good level of suitability for the new residential areas.
aIn D2 and D4, the annual growth rate foreseen is 2.6%. This value is consistent with the rate in Kenya and Tanzania, the other two countries in the East African Community that have the same trends in Human Development Index as Uganda (United Nations Development Programme, 2012).
The authors would like to thank ACAV - Associazione Centro Aiuti Volontari for the hospitality and the support provided during field work.
- Alcamo J, Henrichs T: Towards guidelines for environmental scenario analysis. In Environmental futures: the practice of environmental scenario analysis. Edited by: Alcamo J. Amsterdam: Elsevier; 2008.Google Scholar
- Arua DLG: District state of environment report. Arua, Uganda; 2004.Google Scholar
- Ash N, Blanco H, Brown C, Garcia K, Henrichs T, Lucas N, Raudsepp-Hearne C, Simpson RD, Scholes R, Tomich TP, Vira B, Zurek M: Ecosystems and human well-being: a manual for assessment practitioners. Washington: Island Press; 2010.Google Scholar
- Bahiigwa G, Muramira TE: Poverty relevant environmental indicators for Uganda, Economic Policy Research Centre, Makerere, Kampala, Uganda and National Environment Management Authority, Kampala, Uganda. 2001.Google Scholar
- Byakola T, Mukheibr P: Energy systems: vulnerability, adaptation, resilience. Focus, Uganda: Helio International; 2009.Google Scholar
- Carpenter SR, Mooney HA, Agard J, Capistrano D, DeFries RS, Díaz S, Dietz T, Duraiappah AK, Oteng-Yeboah A, Pereira HM, Perrings C, Reid WV, Sarukhan J, Scholes RJ, Whyte A: Science for managing ecosystem services: beyond the Millennium Ecosystem Assessment. Proc Natl Acad Sci U S A 2009, 106(5):1305–1312. 10.1073/pnas.0808772106View ArticleGoogle Scholar
- Daily GC, Matson PA: Ecosystem services: from theory to implementation. Proc Natl Acad Sci U S A 2008, 105(28):9455–9456. 10.1073/pnas.0804960105View ArticleGoogle Scholar
- Daily GC, Polasky S, Goldstein J, Kareiva PM, Mooney HA, Pejchar L, Rickets TH, Salzman J, Shallenberger R: Ecosystem services in decision making: time to deliver. Front Ecol Environ 2009, 7(1):21–28. 10.1890/080025View ArticleGoogle Scholar
- Emwanu T, Okwi PO, Hoogeween JG, Kristjanson P: Where are the poor? Mapping patterns of well-being in Uganda. Kampala: Uganda Bureau of Statistics; 2003.Google Scholar
- Fisher B, Kerry TR, Morling P: Defining and classifying ecosystem services for decision making. Ecol Econ 2009, 68: 643–653. 10.1016/j.ecolecon.2008.09.014View ArticleGoogle Scholar
- Geneletti D: Environmental assessment of spatial plan policies through land use scenarios. Environ Impact Assess Rev 2012, 32(1):1–10. 10.1016/j.eiar.2011.01.015View ArticleGoogle Scholar
- Hauck J, Schweppe-Kraft B, Albert C, Görg C, Jax K, Jensen R, Fürst C, Maes J, Ring I, Hönigovà I, Burkhard B, Mehring M, Tiefenbach M, Grunewald K, Schwarzer M, Meurer J, Sommerhäuser M, Priess JA, Schmidt J, Grêt-Regamey A: The promise of the ecosystem services concept for planning and decision-making. GAiA - Ecol Perspect Sci Soc 2013, 22(4):217–288.Google Scholar
- International Institute for Sustainable Development: Connecting poverty and ecosystem services: focus on Uganda. Nairobi, Kenya: United Nations Environment Programme; 2005.Google Scholar
- Jaeger JD, Rothman C, Anastasi S, Katha Van Notten P: Training Module 6: Scenario Development and analysis. In GEO resource book: A training manual on integrated environmental assessment and reporting. Nairobi: UN Environment Programme and International Institute for Sustainable Development; 2007.Google Scholar
- Kanyamibwa S: Great Lake Region. In Impact of refugees and internally displaced persons on ecosystem integrity, Sudan post-conflict environmental assessment. Kampala: Albertine Rift Conservation Society; 2007.Google Scholar
- Kibreab G: Environmental causes and impacts of refugee movements: a critique of the current debate. Disasters 1997, 21(1):20–38. 10.1111/1467-7717.00042View ArticleGoogle Scholar
- Knöpfle M: A study on charcoal supply in Kampala: final report. Kampala: Ministry of Energy and Mineral Development (MEMD), Energy Advisory Project (EAP); 2004.Google Scholar
- Mendoza M, López E, Morales-Manilla LM, Alcántara C, Geneletti D, Bocco G: Land suitability analysis to support environmental planning: a multicriteria study in Morelia, Mexico. In Spatial decision support for urban and environmental planning, A collection of case studies. Edited by: Geneletti D, Alias A. Kuala Lumpur: Arah Publications; 2009.Google Scholar
- Ministry of Local Government: Koboko town council structure plan, final report. Koboko, Uganda: Ministry of Local Government; 2007.Google Scholar
- National Environmental Management Authority: Ecosystems, ecosystem services and their linkage to poverty reduction in Uganda. Ltd, Kampala: Centre for Resource Analysis; 2006.Google Scholar
- National Environmental Management Authority: Pilot integrated environment assessment of the Lake Kyoga catchment area. Kampala: National Environment Management Authority (NEMA); 2008.Google Scholar
- Nonhebel S: Renewable energy and food supply: will there be enough land? Renew Sust Energ Rev 2005, 9(2):191–201. 10.1016/j.rser.2004.02.003View ArticleGoogle Scholar
- Robinson T, Pozzi F: Wealth index mapping in the horn of Africa: Animal Production and Health Working Paper. Rome: Food and Agriculture Organization; 2011.Google Scholar
- Rogers D, Emwanu T, Robinson T: Poverty mapping in Uganda: an analysis using remotely sensed and other environmental data, pro-poor livestock policy initiative. Rome: Food and Agriculture Organization; 2006.Google Scholar
- Speziale L: Applying the ecosystem approach to land use planning in Koboko district. Master’s thesis. Italy: University of Trento; 2012.Google Scholar
- Uganda Bureau of Statistics: Uganda population and housing census, main report. Kampala: Government of Uganda; 2002.Google Scholar
- Uganda Bureau of Statistics: Uganda population and housing census, main report. Kampala: Government of Uganda; 2010.Google Scholar
- United Nations Development Programme: 2012 Human Development Report, Main Report. New York: United Nations Development Programme; 2012.Google Scholar
- United Nations Environment Programme: Population displacement and the environment. In Sudan post-conflict environmental assessment. Nairobi, Kenya: United Nations Environment Programme; 2007.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.