Baseline Survey of Conflict on Access to Natural Resources in the Greater Virunga Landscape. REPORT (Draft 1)

Transcription

1 Baseline Survey of Conflict on Access to Natural Resources in the Greater Virunga Landscape REPORT (Draft 1) Submitted by Associate Prof. Gerald Eilu (PhD) Proposed position Team Leader Contact person Assoc. Prof. Gerald Eilu (PhD) Contact address Department of Forestry, Biodiversity & Tourism, Makerere University Postal address P.O. Box 7062 Kampala, Uganda Telephone Tel:+256 (0) ; +256 (0) JANUARY 2017

2 ACKNOWLEDGEMENTS This work was commissioned by the Greater Virunga Transboundary Collaboration (GVTC). The GVTC was established through a treaty signed by the three partner states, namely the Democratic Republic of Congo (DRC), Rwanda and Uganda in Kinshasa in September It is a collaborative transboundary framework of programs, plans and activities to conserve a network of Protected Areas (PAs) in the Greater Virunga Landscape (GVL). The GVTC is a mechanism for strategic, transboundary, and collaborative management of the GVL. Field supervisors, field assistants, and enumerators helped with data collection. This Report was prepared by: Assoc. Prof. Gerald Eilu.(Overall Team Leader) Dr. LawrenceOrikiriza (Overall Field Coordinator Uganda). Dr. Julius Nziza (Co-Team leader Rwanda) Prof. Jackson-Gilbert MajaliwaMwanjalolo (Co-Team leader-drc). AnthonyTumwesigye (Field Coordinator-Uganda) Joseph Oriekot (Statistician/Data analyst) Mr. Robert Mugabe Data Entrants The GVTC Team guided the study and provided comments to improve the report.

3 TABLE OF CONTENTS Content Page ACKNOWLEDGEMENTS... 2 SUMMARY INTRODUCTION BACKGROUND CONTEXT OF THE BASELINE SURVEY OBJECTIVES OF THE BASELINE SURVEY SCOPE OF WORK STUDY AREA OVERVIEW METHODS APPROACH TO THE BASELINE SURVEY TARGET POPULATION FROM WHICH TO SELECT PARTICIPANTS RESEARCH DESIGN, STUDY SETTING AND SAMPLING SELECTION OF PAAS AND NGOS SELECTION OF HOUSEHOLDS DATA COLLECTION METHODS DATA PROCESSING RESULTS SOCIO-ECONOMIC AND DEMOGRAPHIC CHARACTERISTICS OF HOUSEHOLDS PERCEPTION ON EXPLOITATION OF NATURAL RESOURCES AS ORIGIN OF CONFLICT EXTENT AND SEVERITY OF CROP RAIDS BY WILD ANIMALS FROM THE PAS CROSS BOARDER EFFORTS IN MANAGING HUMAN WILDLIFE CONFLICT DISCUSSION SOCIO-ECONOMIC CONDITION OF COMMUNITIES ADJACENT TO THE PAS COMMUNITY PERCEPTION ON EXPLOITATION OF NATURAL RESOURCES AS ORIGIN OF CONFLICT EXTENT AND SEVERITY OF CROP RAIDS BY WILD ANIMALS CROSS BORDER COLLABORATION ON HWC IN THE GREATER VIRUNGA LANDSCAPE CONCLUSION SOCIO-ECONOMIC CONDITION OF COMMUNITIES ADJACENT TO THE PAS COMMUNITY PERCEPTION ON EXPLOITATION OF NATURAL RESOURCES AS ORIGIN OF CONFLICT EXTENT AND SEVERITY OF CROP RAIDS BY WILD ANIMALS CROSS BOARDER COLLABORATION ON HWC IN THE GREATER VIRUNGA LANDSCAPE RECOMMENDATIONS AND PROPOSED M & E INDICATORS SOCIO-ECONOMIC CONDITION OF COMMUNITIES ADJACENT TO THE PAS AND COMMUNITY PERCEPTION ON EXPLOITATION OF NATURAL RESOURCES AS ORIGIN OF CONFLICT EXTENT AND SEVERITY OF CROP RAIDS BY WILD ANIMALS CROSS BORDER COLLABORATION ON HWC IN THE GREATER VIRUNGA LANDSCAPE PROPOSED GVTC M&E INDICATORS REFERENCES...87 APPENDIX 1: DATA TABLES...89 I

4 SUMMARY This work was commissioned by The Greater Virunga Transboundary Collaboration (GVTC). The GVTC is a collaborative transboundary framework of programs, plans and activities to conserve a network of Protected Areas (PAs) in the Greater Virunga Landscape (GVL). The GVL is part of the Central Albertine Rift which stretches from the eastern part of DRC, the northern part of Rwanda to the western part of Uganda. The aim of the work is to establish the origin and magnitude of human-wildlife conflict emanating from management and access to natural resources. The specific project objectives are: i) To establish the socio-economic condition of communities adjacent to the PAs; ii) To assess community perception on exploitation of natural resources as origin of conflict; iii) To determine the extent and severity of crop raids by wild animals from the PAs in the Greater Virunga Landscape; and iv) To explore cross boarder collaboration undertaken by Protected Area Authorities and Non-Governmental organizations in management of wildlife in the Greater Virunga Landscape. Fieldwork was done around seven PAs that include Virunga National Park (NP), Semuliki NP, Queen Elizabeth NP, Volcanoes NP, Rwenzori Mountains NP, Mgahinga NP, and Bwindi Impenetrable NP. In total, over 1043 questionnaires were administered to communities living in the vicinity of the Protected Areas. In addition, several Focus Group Discussions and Key Informant Interviews were conducted. The results of the survey will be used as baseline to monitor future trends in these target parameters. The survey fits within the GVTC Mandate i.e. coordinate and convene stakeholder partnerships and meetings; Information and information management clearing house; brokering/mediation; monitoring and evaluation; serving as warning bell/early warning system; and implementation of some interventions. The key findings show that the population within the areas around the PAs is relatively high, most of it youthful and within the reproductive age thus rapidly growing. The communities are generally poor with an annual average annual income of USD 478, which is generally low. Moreover, the households are generally food insecure and lack basic health as well as sanitary facilities. Rwanda appears to fare relatively better than Uganda and DRC. Overall most households (68%) perceived restriction of exploitation of natural resources in PAs as the main origin of conflicts around Protected Areas. This is generally the scenario around the PAs in the GVL. The quantity of crops raided by wild animals in the areas adjacent to the PAs, on average, amounted to 609 Kg for Maize, 248 Kg for Cassava, 267 Kg for Sorghum, 121 kg for millet, and 476 kg for Irish Potatoes. This, on average, amounted to losses of USD per household. In terms of acreage, it was between half and a whole garden. There are several cross-border collaborative efforts within the GVL. There is relative success of cross boarder efforts meant to create awareness and promote conservation within the GVL. The fact that at least a member of each household belonged to a CBO, shows that communities were fully aware of the value of collaboration and are willing to contribute. However, the involvement of women in the groups is generally less than the males. This calls for deliberate efforts to encourage the women. The key indicators to be monitored include the following: i) Socio-economic condition of communities adjacent to the PAs. This include livelihoods, level of poverty, food security among others; ii) Percentage of the population in areas adjacent to PAs that perceive the exploitation of NR 2

5 as the major origin of conflict; iii) Amount of crops in PA adjacent areas raided by wild animals in 2015 reported in kilograms of crops lost, hectares destroyed and the estimated financial loss in USD $.; and iv) the number of collaborative cross border actions taken by GVTC, NGOs operating in the GVL in partnership with PAAs and focussed on natural resources management and the PAAs in In terms of cross border collaborative efforts, the strategy should be to emphasise the area of intervention rather than number of collaborative initiatives. It is necessary to strengthen the transboundary collaboration between the PA Management within the three countries as an appropriate response to the human wildlife threats in particular the issue of compensation needs to be harmonised across the region. Moreover, where compensation is practised, it should be streamlined and handled efficiently to minimise local population complaints across the GVL. 3

6 1.0. INTRODUCTION 1.1. Background Human Wildlife Conflict (HWC) refers to an interaction resulting in negative effects on humans or wild animals and their environment (adapted from the IUCN/SSC Africa Elephant Specialist Group). HWC is just one of the natural-resource related conflicts. There are several conflicts within and between communities over resources. These frequently involve negative interactions among people who have different goals, perceptions and levels of empowerment. The history of conflicts between humans and wildlife, or humans and other humans over resources, dates back to the dawn of humanity. It occurs on all continents within the developed and developing countries. It varies according to the particular environment and the way of life of the local people. World over, natural resources including land, water, timber, fauna, flora, minerals and oil constitute the most important sources of livelihoods and power (UN DPA and UNEP, 2015). These often provide major sources of conflict if not managed carefully. In Africa, the conflicts are common and pronounced. Rural and peri-urban communities are affected. The consequences of human wildlife conflicts are direct and indirect. The direct effects include injury or death from encounters with dangerous large animals such as hippopotamuses, elephants, lions and baboons. Direct conflict, can easily be measured by counting or measuring the damage done. The indirect effects are also common and are manifested as loss of crops, livestock or damaged infrastructure. The small fauna (e.g. birds, rodents or insects) may also cause enormous crop losses when the attacks are made in large numbers. Indirect conflict also includes wider issues such as competition for wild fruits and water, which are needed by both people and wildlife, or the fear of being attacked while walking within the Protected Area or periphery during day or night. Indirect conflict is much more complicated to measure because it deals with issues like fear, competition and attitude rather than actual damage. Direct contact with wildlife is generally more common inside and around protected areas, where wildlife population density is higher and animals often stray into adjacent gardens. Human-wildlife conflicts have become more frequent and severe over the recent decades. The reasons include: human population growth; extension of transport routes; expansion of agricultural activities; and expansion of industrial activities. These have led to increased encroachment on previously wild uninhabited areas (WWF, 2006; Conover 2002, 2003). Competition for natural resources and habitats has increased. Moreover, the impacts of climate change have exacerbated the conflicts. One of the areas where this is the case, is the Greater Virunga Landscape. The GVL is categorized by IUCN as a high biodiversity hotspot and ranked among the highest priority areas for conservation in Africa (Plumptre et al., 2007). However, the landscape is prone to conflicts in various forms. There is a high human population within the GVL comprised of 95% subsistence farmers. The area faces pressure from the surrounding communities from poaching, encroachment, fuel-wood harvesting, timber harvesting, illegal grazing and harvesting non-timber forest products. However, there is permitted access to some natural resources from PAs mainly under the Multiple Use arrangements but conflicts still persist. Due to the high and increasing human population, the rural communities around PAs have limited development opportunities and are heavily dependent on natural resources. Although some alternative means of livelihoods have been provided by PAAs and their partners, the scale is too small to compensate for the loss of PA resources. This aggravates the situation and necessitates the establishment of the socio-economic condition of communities 4

7 adjacent to the PAs in the GVL. It is inevitable that perceptions of communities adjacent to the PAs are considered if conservation efforts and harmonious relationships between the management of the respective PAs and neighboring communities are to be achieved. Equally important, is the human-wildlife conflict manifested in terms of crop raids by wild animals from the PAs. PA frontline communities are the most prone. Conflicts due to crop raiding constitute a major conservation and management challenge in the GVL. Conserving the wildlife in PAs is one objective while ensuring local community benefits from the conservation programs is another important but controversial objective. The impact of crop raiding on attitudes of local communities towards PAs can undermine conservation efforts (Fungo 2011) if not addressed. Wild animals are forced to raid crops in adjacent gardens when human populations invade wild animal territories through human activities that reduce the availability of natural prey/food sources. Instances of crop raiding within the GVL are reported in the Virunga-Bwindi region, where habitat destruction and human population growth has caused the mountain gorilla and other forest animals, such as bush pigs, elephants and buffaloes, to increasingly come into contact with people resulting into conflicts. In the Volcanoes National Park, Rwanda, Buffaloes are the main cause of cropraiding. Some communities around the Mgahinga Gorilla National Park continue to experience severe problems with buffaloes and elephants outside the park. Around Bwindi Impenetrable National Park, crop raiding by wildlife is an issue that contributes to hostility between the park and local communities (Kalpers et al., 2011). With such human-wildlife conflicts, community access and exploitation of natural resources is a necessity. Strategies for addressing conflicts and mechanisms for collaborative are required urgently. Analyses of HWC cases demonstrate that they are a major concern to local livelihoods and to national economies (WWF, 2008). Impact of HWC such as damage to crops or livestock by wildlife has the potential to threaten key livelihood options, for example by creating food insecurity and loss of investments. Crop damage is a particular risk to poverty stricken households that do not have access to alternative sources of income. Encounters with small animals, exposure to zoonotic diseases, physical injury or death caused by large mammal attacks, have high financial costs for individuals and society. Negative social impacts also include missed opportunity for school and other domestic work as time is spent guarding property. Other impacts include loss of sleep, fear, as well as restriction of travel (Hoare, 1992). The wide range of costs is an indication that stakeholders such as government, wildlife managers, scientists and local communities need to adopt long-term measures to resolve the conflicts. Policy and legislation on human-wildlife conflict in GVL Uganda The Uganda Wildlife Statute (1996) and Uganda Wildlife Act cap 200 of 2000 provide the legal framework for the management of human-wildlife conflicts in Uganda. Section 3(1 (f), of the statute provides for the promotion of ecologically acceptable control of problem animals. The Local Government Act 1997 assigns the responsibility of managing crop-raiding animals declared vermin by UWA to the local government authority at district and sub county levels. Currently the animal species declared as vermin in Uganda are Cercopithecus aehtiops (Vervet Monkey), Potamochoerus larvatus (Bush pig) and Papio anubis (Olive baboons). Within their natural habitats, the animals are not considered to be vermin. Many animals not listed as vermin are legally protected outside the PAs and it remains the responsibility of UWA to control and protect them. These include animals of 5

8 international significance such as the Buffaloes, Hippopotamuses, Mountain Gorillas, Elephants, Chimpanzees and Leopards, among others. The Uganda Wildlife Statute (1996) does not provide for compensation to over crop losses to crop raiding wild animal. There are indirect mechanisms to benefit the affected communities. Examples include sharing the revenue generated from the park for community development projects or allowing frontline communities access to some of the park resources. Whereas, the responsibility of management of vermin animals is vested in the district local governments, there is inadequate technical and financial capacity to deal with the problem. The challenge with the compensation schemes is that it is difficult to cover all threats, property damage, and losses to human life caused by all problem animal species. This leaves the local community in dilemma on how to address the situation properly. This provides the basis for arguments in favour of non-compensation schemes. However, experiences from Rwanda are generating lessons that will guide future options. Rwanda The wildlife laws have provision for compensation over damage caused by wildlife. The law in force (Law No. 26/2011 of 27/07/2011 covers compensation for damage caused by animals; Law No. 52/2011 of 14/12/2011 established the Special Guarantee Fund for automobile and damage caused by animals; Prime Ministers Order No. 26/03 of 23/05/2012 determines the rates, calculation method and criteria for determining compensation to the victim of damage; the Ministerial Order No. 14/MINICOM/2012 of 18/04/2012 stipulates the list of wild animal species covered by the law on compensation. Claims are now being processed for compensation over wildlife damage around Volcanoes National Park. In case of loss of life, the laws cater for compensation to family members of the deceased, to the victim of corporal injury caused by an animal, and for loss or damage to property. Rwanda Development Board (BRD) provides money for compensation by allocating 5% of the annual revenues from tourism through the Special Guarantee Fund (SGF) in line with the government s wildlife damage refund scheme. One of the problems with the compensation law is question of costs of valuation of damage (and complication), the valuers, costs, and accessibility to sites of damage. Total cost of damage may be higher than park revenues unless compensation for damage is limited. Also red tape, long legal processes and corruption may derail the process unless clear guidelines are established. Democratic Republic of Congo In the DRC, The wildlife resources law (No of 1982) regulates hunting of animal species under total protection status. The ordinance-law No (1969) on the conservation of nature, gives a framework for improved conservation of wildlife in general, especially the great apes, covered under the nature reserves. Law No (1975), establishes sanctuary areas. Arrangements for species protection are derived from provisions of the hunting regulations (Law No ). Under this law, animals are categorized into three: i) Fully protected; ii) Partially protected, and iii) Non-protected. The law also forbids killing, capturing, hunting, pursuing, deliberately disturbing, or illegally and with prejudicial intent causing any of the listed animals to flee, unless furnished with a scientific permit issued by the ministry responsible for hunting issues. There are no specific sectoral policies. Tropical forests and biodiversity are managed through legislation only. The legislation does not provide for strategies of dealing with problem animals. Moreover, the communities are not allowed to kill crop-raiding animals under the hunting law. 6

9 Reforms to law No (1969) provide for mechanisms of addressing Human Wildlife Conflict in an acceptable manner for conservation and local community livelihood. However, there are no specific provisions in the current laws dealing explicitly with Human-Wildlife Conflict. In 2012, a ministerial directive was made to ICCN to examine the modalities for compensation due to wildlife damage to property and injuries or deaths of persons Context of the Baseline Survey The Albertine Rift and its network of PAs and stakeholders cover a long stretch with rugged terrain. The poor road network combined with remoteness and insecurity curtails communication. Access to natural resources in the PAs is a source of conflicts between communities living adjacent to the PAs and PA Authorities (PAA), as well as between the communities. While, for example, there are efforts to minimize poaching and illegal trade in animal species and derivatives, extraction of these resources persists in Albertine rift in general and the GVL. Poaching still exists as one of various threats to Wild animals. This is a persistent and sophisticated form of conflict over wild life resources. Equally important, is the human-wildlife conflict that occurs in various forms including crop raiding by wild animals from the PAs. This mostly affects, but is not limited to PA frontline communities. This survey of community perception of conflict over natural resources in the GVL was undertaken to contribute towards a solution to this problem. Hence, this task is set up within the context of the various mandates of the GVTC. Cross border actions taken by GVTC and NGOs operating in the GVL (in partnership with PAAs and focused on natural resources management), particularly in the context of conflict, however, remain largely un-quantified. This study seeks to provide baseline information on conflicts on natural resources access under the current transboundary conservation efforts in the GVL. The objectives of GVTC as enshrined in the treaty include: collaboration for improved transboundary conservation of natural resources; ensuring improved and coordinated landscape level planning and management of natural resources; sharing skills and expertise; and ensuring the sharing of benefits and costs across borders. In order to achieve these objectives, GVTC exercises its functions based on the following mandates: Coordinating and convening stakeholder partnerships and meetings Serving as clearing house for information and information management Brokering/Mediating Monitoring and Evaluation Providing early warning systems Implementing some interventions The results of the survey will be used as baseline to guide future efforts of managing the conflicts. Specifically, the findings will be used in the following ways: Help GVTC and other Protected Area Authorities (PAAs) start accounting for the improvement (or otherwise) in conservation arising out of efforts invested Contribute to the annual conservation status reports of PAAs and implementing NGOs in the GVL. Gauge the effectiveness of interventions undertaken by GVTC and stakeholders It is vital that the GVTC and partners measure and monitor these aspects annually. 7

10 1.3. Objectives of the baseline survey The aim is to establish the origin and magnitude of human-wildlife conflict emanating from management and access to natural resources. The specific objectives are: i. To establish the socio-economic condition of communities adjacent to the PAs ii. To determine the extent and severity of crop raids by wild animals from the PAs in the Greater Virunga Landscape iii. To explore cross boarder collaboration undertaken by Protected Area Authorities and Non- Governmental organizations in management of wildlife in the Greater Virunga Landscape iv. To assess communities perception on exploitation of natural resources as origin of conflict 1.4. Scope of work Seven PAs within the Greater Virunga Landscape were covered. In addition the surrounding frontline communities were covered by the study. Conflict is understood to mean any situation that results into displeasure, harm or loss to the parties involved. Concerning the conflict on natural resources access, we targeted access to natural resources interpreted to mean the resources found within or originating from the boundaries of the protected area. The study did not quantitatively assess conflicts over resources that do not occur in or originate from the PA. We also catered for the fact that conflict goes beyond human-wildlife conflict, embracing other forms of conflict over resources. The baseline survey determined the current status of the conflict situations and provides units that will be used to measure aspects stipulated in the objectives. Ultimately, the study provides workable strategies and coordinated approaches between countries and PAs The rationale of the study is that access to natural resources in the PAs is a source of conflicts between frontline communities and park management. Equally important, is the human-wildlife conflict emanating from crop raiding by wild animals from the PAs. However, data are still required to inform policy making and implementation meant to address threats to wildlife such as poaching and other challenges to conservation. This work generates an understanding of the community perception of conflict over access to PA resources as well as conflicts resulting from crop raiding and destruction by problem. The results of the survey will serve as baseline to guide future efforts of managing these conflicts in the Greater Virunga landscape.. 8

11 2.0. STUDY AREA 2.1. Overview General Description The GVL is part of the Central Albertine Rift, which stretches from the eastern part of the Democratic Republic of Congo (DRC), the northern part of Rwanda to the western part of Uganda. It covers seven PAs that include Virunga National Park (NP), Semuliki NP, Queen Elizabeth NP, Volcanoes NP, Rwenzori Mountains NP, Mgahinga NP, and Bwindi Impenetrable NP. The coverage includes the human communities around the PAs. The baseline survey was conducted mainly among frontline communities living around PAs in the GVL. The administrative units are referred to by different terminologies in the target countries: i.e. parishes (Uganda), sectors (Rwanda) and Groupment (DRC). These parishes are adjacent to the PAs in the GVL. The GVL (Figure 1) is an interconnected chain of seven PAs that straddles the borders of Uganda, Rwanda and DRC covering a total area of about 13,200 km 2 of varied habitats including glaciers and rock, montane and lowland forest, savannas, wetlands, bamboo, hot springs and active volcanoes with lava flows (Nyiragongo and Nyamulagira) ( 1 indicates a summary how each objective of the baseline survey will be addressed. 9

12 Figure 1: The map shows the study area from which we will select the sampling sites 10

13 2.2. Description of the PAs (National Parks) of the GVC landscape Over view of flora and fauna within target PAs of the GVL The Greater Virunga landscape covers about 13,190 km2 and includes a wide variety of habitats and altitudes, ranging from 600 to 5100m above sea level. It is also incredibly rich in total species as well as species of conservation concern (Box 1) and is one of the most biodiverse sites in the world. Virunga National Park ranks high for various taxa because it contains a very diverse suite of habitats ranging from glaciers to lowland forest and savannas. Some of the key flora and fauna are described here: BINP QENP RMNP SEMILIKI VIRUNGA VOLCANOES MGAHINGA Mammals Spp No ?? AR Endemics ?? Threatened ? Birds Spp No ? 79 AR Endemics 24(1) (5) 27 (2) Threatened ?? Reptiles Spp No AR Endemics ?? Threatened ?? Amphibians Spp No ?? AR Endemics ?? Threatened ?? Plants (Trees) Spp No. Plants ?? ?? AR Endemics ?? Threatened ?? Mammals Two endemic genera occur of mammals occur within the GVL (Rwenzorisorex, and Delanymys). Virunga National Park in eastern DRC has the highest number of endemic mammal species (21). Bwindi Impenetrable National Park (20) and Rwenzori Mountains (18) are next highest. The mammals include the mountain gorilla (Gorilla beringei beringei), golden monkey (Cercopithecus kandti), and Rwenzori otter shrew (Micropotamogale ruwenzorii), and the Rwenzori duiker, (Cephalophus rubidus). The number of globally threatened mammals (13) in Virunga National Park is one of the highest in the region. The fauna of the Rwenzori also includes endangered species such as the Chimpanzee (Pan troglodytes), elephants (Loxodonta africana), L Hoest s monkey (Cercophithecus l hoestii), Rwenzori blackfronted duiker (Cephalophus nigrifronsrubidus). Other fauna includes the Rwenzori Colobus Monkey (Colobus angolensis ruwenzorii), Rwenzori Hyrax (Deadrohyrax arboreus ruwenzorii) among others. 11

14 Large predators such as leopards Panthera pardus, lions Panthera leo, some of the larger primates (chimpanzees Pan troglodytes, gorillas Gorilla beringei), and large ungulates (elephants Loxodonta africana, hippopotamuses Hippopotamus amphibious) are species that need large areas to maintain viable populations. The largest and most critical of the landscapes includes the Virunga National Park in DRC, with the Parc National des Volcans in Rwanda, and Semliki, Rwenzori, Bwindi Impenetrable, Queen Elizabeth National Parks in Uganda. In the Mgahinga NP, the Nyakagezi group of Mountain Gorillas and buffalo (Syncerus caffer) occur. There is also the golden monkey; the blue monkey (Cercopithecus mitis sp.). Other mammals are the golden cat (Profelis) aurata), serval cat (Felis (Leptculurus) serval), leopard (Panthera pardus), spotted hyena (Crucuta crocuta), side-striped jackal (Canisadustus), black-fronted duiker (Caphalop husnigrifrons), bushbuck (Tragelaphus scriptus) and giant forest hog (Hylocheorus meinertzhageni). The Semiliki National Park has species including the pygmy antelope and two flying squirrel species. It is also home to the peculiar water chevrotain, known as the fanged deer. Other mammals include leopards, elephant, forest buffalo, hippopotamus, civets, potto, bush babies, mona monkeys, water chevrotains, and nine species of Duikers, including the Bay Duiker (Cephalophus dorsalis) and the Pygmy Flying Squirrel (Idiuus zenkeri) that occur nowhere else in East Africa. The mammals of Volcanoes NP include Golden monkeys, spotted hyena, bushbucks, black fronted duikers, buffaloes and forest elephants, among others. Birds Virunga National Park has one of the highest numbers of species of endemic birds (27). Virunga National Park has a large number of endemic species (11). In the Rwenzori, some of the species of conservation concern include Rockefeller s Sunbird and Shelley s Crimson-wing. Others include the Rwenzori Turaco, Bamboo Warbler, Golden-Winged Sunbird, Scarlet-Tufted Malachite Sunbird and Stuhlmann s Double-Collared Sunbird and the White-Collared Olive-Back (Nesocharis ansorgei). Semliki National Park has some unique forest birds such as the rare Sassi s Olive Greenbul (Phyllastrephus lorenzi) and Forest Ground Thrush (Turdus oberlaenderi). Nine species of hornbills have been recorded. as well. Reptiles Sixteen endemic reptile species occur in the Rift, of which Virunga National Park contains the highest number (11), followed by Rwenzori Mountains National Park (9). The reptiles include the Three Horned Chameleon (Chamele ojohnstonii) and other species of conservation concern. Amphibians Virunga National Park had the highest number of endemic species (16). Sixteen Albertine Rift amphibians are globally threatened, of which 14 are endemic. Virunga National Park with 10 and Bwindi Impenetrable National Park with six have the highest numbers of threatened species. Fiftysix fish species are endemic to lakes George and Edward. Plants Virunga National Park in eastern DRC and Bwindi Impenetrable National Park in Uganda had the highest numbers of plant species recorded. The more conspicuous of the endemics, for example in the Rwenzori, include Hypericum revolutum ssp. revolutum; Senecio adnivalis and Senecioerici-rosenii two of the iant-groundsels; Alchemilla argyrophylla and related species which carpet much of the moorlands. Other species only recorded on Rwenzori in Uganda are Adenocarpus mannii, Agelanthus dijurensis, Canthium lactescens, Desmodium repandum, Jasminum abyssinicum, Rubus rosifolius and Schefflera polysciadia (Howard 1995). 12

15 Rwenzori Mountains National Park (RMNP) Location and overview The RMNP is located in western Uganda bordering the Democratic Republic of Congo (DRC) in the West. It is part of the Rwenzori ranges, which rise from about 1670m to 5,109m at Margherita, Africa s third highest peak. The mountain ranges of the Rwenzori are shared between Uganda and DRC. Over 75% of the range is located in Uganda. It covers 995km 2 and was gazetted National Park in Before this, it was a Forest Reserve where the communities had access to resources including timber, poles, bush meat, and firewood among others. When it was gazzeted National Park, management was transferred from the Forest Department to Uganda National Parks, now Uganda Wildlife Authority (UWA). The communities were, thereafter, not allowed to access the resources thereby creating conflicts with authorities. The communities now access some resources (e.g. Herbal medicine and local construction materials) under collaborative resource management arrangements in selected parishes but coverage is still low and few resources are permitted. Conservation Values It supports several species of global conservation concern. The Park is globally important for its biodiversity and cultural values. In 1994, the United Nations Education, Scientific and Cultural Organization (UNESCO) designated the Rwenzori Mountains a World Heritage site. The history, culture and beliefs of the Bakonzo, the main ethnic group in the area, are closely woven around the Rwenzori Mountains. The Bakonjo have a number of traditional practices, some of which are intended to preserve the mountain resources. They have strong respect for the ranges and the life it supports. Isolated places such as thickets, huge trees, deep valleys or top of hills are believed to be residences of forest gods. Such places are used for worship and offerings. The glaciers on Rwenzori are melting away, but it is clear that they still play an important role in shaping the landscape and ecology of the range and are very important attractions for visitors. The high rainfall (over 2000 mm annually) in most parts of the range on the Uganda side and the melting of the snow, results into numerous rivers flowing from the mountains and presence of extensive bogs (wetlands). These are associated with numerous plants and animals. The Rivers flowing down the Mountain range feed the economically important lakes, Edward and George, and constitute a major source of the White Nile. A combination of beautiful peaks, glaciers, valleys, and lakes, occurs on the mountain range making Rwenzori Mountains beautiful. The stratified vegetation of Rwenzori is one of the main attractions for visitors. Five major vegetation zones can be observed. These are grassland ( m), montane forest ( m), Bamboo/Mimulopsis zone ( m), Heather/Rapanea zone ( m) and the afro-alpine moorland zone ( m). The formation of Rwenzori, its landscapes, flora and fauna are fascinating areas for research. The interactions of these with the climate, hydrology, neighbouring communities management actions are important areas for research. Management The RMNP covers four Districts of Kasese, Kabarole, Ntoroko and Bundibugyo in Uganda. In DRC the mountains are part of the Parc National des Virunga (PNVi) that is contiguous with the RMNP for about 50km (RMNP GMP, 2004). The mountains are a source of various resources for the communities, mainly the Bakonzo, who live on its slopes. 13

16 The management zones of the park include the Wilderness Zone intended to provide a high level of protection to representative park ecosystems; the Tourism Zone, designated along the established or planned tourist trails and camps; and Integrated Resource Use Zone where communities are permitted to harvest resources regulated by agreements with the park Management In western Uganda specifically, the Allied Democratic Forces (ADF) rebel group was active from 1995 to The group was especially active in the Rwenzori Mountains and during this time many Chimpanzees and other primates were poached. The ADF became more dependent on poaching for subsistence as displaced villagers abandoned their crops and rebels ran out of fields to raid. Illegal hunting of wild animals and illegal trade persist. Hunting is sometimes done as one of the traditional practices of the Bakonzo. Poaching mainly targets the primates, bush pigs, and elephants. Others include the black-fronted and red duikers as well as rock hyraxes. Crop raiding is one of the most prominent human-wildlife conflicts in the areas around the protected area. Elephants, buffalo, monkeys (Blue monkeys and Vervet monkey, bush pigs, chimpanzees, rodents and porcupines and baboons Queen Elizabeth National Park The area of QENP was originally the ancestral grazing area of the Basongora pastoralists. In 1952 L. George and L. Edward Game Reserves were combined and gazetted as QENP. The park is Uganda s most popular tourist destination. It has over 600 species of birds. It covers 1,978 km², forming part of a much larger transboundary grassland-forest-wetland ecosystem. Its western border is contiguous with the Parc National des Virunga in the DRC. It has several species of birds and is famous for its tree-climbing lions and large hippo population. Lakes George and Edward found within the park have the most productive fisheries in Africa. A number of conflicts exist. These include competing and conflicting land use practices (cultivation, pastoralism and wildlife protection, conflict over grazing land and water sources, cattle keeping within fishing enclaves). There is encroachment and illegal extraction of park resources. Others include community perceptions of being marginalized and the right to be on the land, and claims by communities that they own part of the park land Bwindi Impenetrable National Park (BINP) Bwindi Impenetrable National Park (BINP) located in South-western Uganda is the home to the mountain gorillas, constituting approximately half the World s population. Before it was gazetted as a National Park, Bwindi Impenetrable Forest provided multiple livelihood and subsistence resources for people living around it. These included domestic timber products such as firewood, building poles, forest products such as medicinal plants, basketry materials and foods that included honey, edible plants and bush meat. In 1991, the forest was gazetted as a national park, which changed its conservation status and meant that forest access by the communities including the Batwa (the forest dwelling people) were prohibited. This created conflict between the local community and the Park Authorities.Despite numerous efforts and programmes these conflicts though reduced still exist Mgahinga Gorilla National Park Mgahinga Gorilla National Park (MGNP) is located in the extreme south-western corner of Uganda in Kisoro district. It covers 33.7 km 2 and is the smallest park found in southwestern Uganda at the border of DRC and Rwanda. It is located on the slopes of three of the six virunga Volcanoes: Mt. Gahinga, Mt. Sabinyo and Mt. Muhavura. It is important as habitat of Mountain Gorillas. It is contiguous with Parc National des Virunga in the DRC, and Parc National des Volcans in Rwanda, forming transboundary Protected Area known as the Virunga Conservation Area of 434 km 2. The 14

17 Mgahinga Gorilla National Park is habitat to golden monkeys, the Black & White colobus monkeys and blue monkeys, and forest elephants among others. It has several species of birds. Surrounding the park are various local communities including the African pygmy Batwa tribe which many years ago survived by gathering fruits and hunting in the Bwindi Impenetrable Forest. They were relocated after the establishment of the park. This together with restricted access to resources is a source of conflicts. Conservation values The area is characterized by two rainfall seasons (March-April and September-December) and two dry seasons (January-February and May - August). MGNP is frequently shrouded in high mist and humidity with temperatures ranging from 4 o C at the highest elevations to 18 o C at the lowest points. The park includes Mt. Muhabura (4,127 m), Mt. Gahinga (3,474 m) from which it derives its name, and Mt. Sabinyo (3,645 m - three of the Virunga volcanoes) (MGNP GMP, 2014).The three volcanoes in MGNP are thought to have arisen in the early to mid-pleistocene era, and formed through a deposition of layers of ash and cinders from successive lava flows. Sabinyo is believed to be the oldest Volcanoes, followed by Gahinga, which is younger, with a swamp crater of about 180m diameter at its summit. Muhabura is believed to be the youngest Volcanoes. It is cone-shaped with a small crater lake approximately 36m in diameter at its summit. There are numerous caves on the slopes of the mountains, caused by lava tubes. The soils are all derived from volcanic activity and reflect the relatively short time since the last eruption (MGNP GMP, 2014). The MGNP is an important water catchment supported with a high protective cover of vegetation, numerous flowing streams, a Crater Lake and swamp crater on Mt. Gahinga. The vegetation in MGNP consists of woodland, and only a small area of pure montane forest still remains at the base of Mt Muhabura following encroachment in the 1950s. Above the montane forest belt is the bamboo (Arundinaria alpina) zone that stretches from the western boundary on Sabinyo to the lower slopes of Muhabura. The Hagenia-Hypericum zone appears above the bamboo zone on Mt. Sabyinyo and below it on Gahinga. The Afro-Alpine Belt, characterised by giant Senecio and Lobelia species, occurs above the Ericacious Belt and reaches its maximum development on Mt. Muhabura. Management The park is located in Bufumbira County, covering Nyarusiza and Muramba sub-counties and neighbours three parishes of Gisozi, Rukongi and Gitenderi. The park was originally established as a Forest reserve. It was made a Sanctuary for protection of Gorillas in 1930, upgraded to a Game reserve in 1965 and a National park in Following the establishment of the park in 1992, communities were relocated and access to park resources restricted. However, communities surrounding the park depend on it for bamboos, bean stakes, building poles, firewood, thatching grass, footpaths and water. The MGNP is thus an important source of livelihoods. The Batwa community were known to have survived in the park by gathering fruits and hunting wild animals. The park area was previously heavily encroached on with settlements until 1992 when MGNP was created. Its creation led to the eviction of over 2,400 people. For the last twenty one (21) years, the Uganda Wildlife Authority and other stakeholders have implemented programs and activities that have seen the park recover to its original vegetation, though not yet fully recovered at the moment (MGNP GMP, 2014). 15

18 Due to the high population pressure (UBOS, 2011), there is acute land shortage in Kisoro district. In addition, communities surrounding the park cultivate the slopes of Mt. Mgahinga and Mt. Muhavura up to the boundary of the park and continue to experience severe problems with buffaloes and elephants outside the park thereby increasing the risk of crop raiding and human-wildlife conflicts due to crop damage by wildlife- a big challenge around MGNP Semliki National Park Semuliki National Park is located in Bundibugyo District, western Uganda. It covers 219 km 2 and is part of the Central African Congo Basin forest system of the DRC, being separated from the Ituri forest of the DRC only by the Semliki River. It is separated from the rest of East Africa by the Rwenzori Mountain range and with it being located within the Albertine Rift, the western arm of the Great Rift Valley it is included within the Eastern Afromontane biodiversity hotspot. Four distinct ethnic groups live near the park Bamba farmers live along the base of the Rwenzori while the Bakonjo cultivate the mountain slopes. Batuku cattle keepers inhabit on the open plains and Batwa pygmies, traditionally hunter gathers, live on the edge of the forest. Conservation values It has a high diversity of plant and animal species and many microhabitats. Most of the plant and animal species in the park are also found in the Congo Basin forests, with many of the species reaching the eastern limit of their range here. The flora is dominated by a single tree species, Cynometra alexandri, mixed with tree species of a more evergreen nature. Swamp forest communities are also found. The park has several of the Charaxes butterfly genera. Management The Conservation of the park is faced with challenges including illegal access of resources from the park by neighboring communities and human-wildlife conflict mainly due to crop raiding and vermin which is the focus of this baseline study Virunga National Park Located in eastern DRC, Virunga is Africa s oldest national park. It is a UNESCO World Heritage Site and is known for its wildlife-rich vegetation mosaic, active and dormant volcanoes and permanent glaciers. It is home to the critically endangered mountain gorillas. It contains more species of mammals, reptiles and birds than any other PA in Africa, and possibly in the world. Several plant species occur, many of which are endemic to the Albertine Rift. Virunga is Ramsar site (Wetlands of International Importance). Despite its status as protected wilderness, the park has been under threat for more than two decades by armed groups that engage in poaching, deforestation and other unsustainable and illegal resource exploitation. As a result, Virunga has been included on the List of World Heritage Sites in Danger. Currently, all five of DRC s World Heritage Sites are listed as in danger. The park is under attack from oil exploration with concessions covering 85 per cent of its territory having been designated, and exploration companies are on its doorstep. With oil come risks from pollution and destabilization, which could destroy this extraordinary place forever Volcanoes National Park Rwanda Volcanoes National Park lies in northwestern Rwanda and borders Virunga National Park in the DRC and Mgahinga Gorilla National Park in Uganda. It covers an area of 160 km 2. It has mountain gorillas. The park was first gazetted in 1925, as a small area bounded by Karisimbi, Visoke and Mikeno, intended to protect the gorillas from poachers. It was the very first National Park to be 16

19 created in Africa. Subsequently, in 1929, the borders of the park were extended further into Rwanda and into the Belgian Congo, to form the Albert National Park, a huge area of 8090 km2, run by the Belgian colonial authorities who were in charge of both colonies. In 1960, the park was split into two, and upon Rwandan independence in 1962 the new government agreed to maintain the park as a conservation and tourist area. The park was halved in area in 1969 and between 1969 and 1973, 1050 ha of the park were cleared to grow pyrethrum. The potential sources of conflict with local communities range from the displacement of local communities from their land, to crop raiding by wildlife and restricted access to natural resources. Conservation values The park is home to the iconic Mountain Gorillas and also a home to five of eight volcanoes of the Virunga massif (Karisimbi, Muhabura, Sabyinyo, Bisoke, and MGahinga) that are covered by rainforest and bamboo. The park is rich in both flora and fauna. The vegetation belts include the lower montane forest (now mainly lost to agricultural activities). Between 2400 and 2500m, there is Neoboutonia forest, 2500 to 3200m, has Arundinaria alpina forest (bamboo) that covers 30% of the forest. From 2600 to 3600m, Hagenia-Hypericum forest occurs, covering 30% of the park. The vegetation from 3500 to 4500 is dominated by Lobelia wollastonii, Lobelia lanurensis and Senecio erici-roseni covering 25% of the park. Secondary thickets, meadows, marshes, swamps and small lakes also occur, but their total area is relatively small. Management Volcanoes National Park is has many threats affecting the protection of biodiversity and causes conflicts. Some of these conflicts arise from forest dependence practices of the surrounding communities that illegally enter the park to extract non-timber forest resources such as honey, bush meat, water, bamboo, and others. On the other hand, frontier communities have suffered from crop raiders (wild animals from the park) that damage and destroy their crops resulting into open conflicts with the park management. Conflicts in VNP mainly result from crop raiding (Irish potatoes, pyrethrum, maize), destruction to property (houses, beehives), and loss or injury to livestock or humans. The animals that cause damage include buffaloes, duikers, elephants, golden monkeys, gorillas, and jackals. Efforts to manage conflicts over natural resources in the GVL In order to address the conflicts over natural resources, stakeholders have attempted to find various strategies. In Rwanda, for example,, the Rwanda Development Board and its conservation partners (local and international conservation NGOs) sought interventions that could minimize or eradicate the conflicts around National Parks such as Volcanoes. Physical barriers Some strategies re used around VNP included the setting up of physical barriers ( ) such as buffalo wall or stone wall, trench digging and related regular maintenance mechanisms ( ). Though these physical barriers have played a significant role in reducing crop raiding cases at more than 80%, they have not fully controlled the increasing frequencies of animals ranging outside VNP (RDB, 2016). This failure is attributed to the gaps on the stone wall, insufficient funds as well as the critical animals (Mountain Gorillas, Golden Monkeys). 17

20 Revenue-sharing The revenue sharing scheme was launched in 2005 where by 5% of the annual tourism revenues from all the protected areas are put into a fund for community projects in administrative sectors that neighbour national parks. The RDB issues calls for proposals, and a project selection process is made at sector and district levels. Sectors are the second major administrative entities, which will be autonomous when the decentralization process is complete, and are being coordinated by the districts. Selection criteria include positive impacts on conservation of biodiversity in protected areas, and to local community. Areas that register more cases of conflict between protected areas and the community, according to results of Ranger- Based Monitoring (a system used by RDB to monitor biodiversity) have preferential access to funds, as do those which are located closer to protected areas. For example, in revenue sharing contribution ratios of infrastructure to small income generating projects was 66% and 34% respectively. However, park management is advocating for more support to small income generating projects for communities neighbouring protected areas especially cooperatives involved in community conservation and park management like the crop ranger associations (Uwingeli, 2012). In Uganda, the revenue sharing program is implemented by UWA and local governmental systems whereby 20% of park entry fees are given to local governments around the National Parks. Money from revenue sharing is allocated to frontline parishes. The funds are used to implement a variety of projects to benefit the community such as building clinics or schools or income-generating projects (Mukumi, 2013). The revenue sharing policy focuses on ensuring equitable benefit to the communities in the front line parishes who suffer the effects of Human Wildlife Conflict. During the year , the local communities around RMNP received amounting to UGX.135m = (US $53,465). Patrols and Law Enforcement A variety of strategies and methods are used around the protected areas to minimise HWC including law enforcement to stop poaching, physical deterrents to prevent wild animals from crossing over to cultivated fields. Conservation methods aimed at raising awareness about the importance of Conservation and appreciation of the value of the parks and improving park community relations. These initiatives vary from park to park depending on animal species, farming systems and strategies, measures that have been tested and accepted by local communities, and biophysical features of the area/site and institutional and legal framework of each Country. In Uganda, the UWA is responsible for managing the National Parks and wild animals including the vermin and crop raiding animals. It works to monitor and eradicate poaching. The UWA currently works with the tourism police, the Uganda Peoples Defence Forces (UPDF), to enforce the law against poaching activities Regular patrols are done within the park boundaries while intelligence networks consisting of informants from the UWA as well as community members. However this approach has a number of shortcomings including, lack of resources which limits the coverage of the PA. Community Conservation The Community Conservation (CC) department of the UWA, as well as several NGOs, have programs which encourage communities around PAs to indulge in conservation. The programs aim 18

21 to instil a belief that conservation is beneficial, while overuse and exploitation of natural resources is detrimental to the community. Apart from Revenue sharing (indicated earlier), other initiatives include sensitization and mobilization of surrounding communities, employment opportunities, community-based tourism, and controlled access to PA resources. Other initiatives include the planting of unpalatable buffer crops such as green pepper, onions and garlic by frontline communities. These double as cash crops to generate household income (Kizza Fredric 2014). The Maurithius thorn has also been planted to establish a barrier to incursions of wild animals into community crop fields. Rwenzori mountains National park is also an important source of resources for communities. The resources include Smilax spp. and Acalypha spp. (for basket making), medicinal plants, mushrooms water, honey, fibres from tree bark, bamboo stems and sheath. A number of resource use agreements have been signed to enhance sustainable access to PA resources. Also the park authorities provide protection to communities bordering PAs by deploying rangers at outposts. These help to scare away animals that stray into the gardens. In order to strengthen levels of protection for community members, districts with parishes bordering protected areas are also supposed to recruit vermin control officers and rangers. However, support from UWA or the government in protecting peoples lives and property is still inadequate (Madeleine 2013). Despite of the efforts to mitigate these conflicts, a recent study conducted by Babaasa (2016) to assess the effectiveness and perceptions towards physical deterrents to crop raiding animals around Volcanoes National Park, shows that the strategies have been ineffective and have failed to eradicate the problem though they have played a role in minimizing crop raiding cases. Other reports from the community conservation department in the Volcanoes NP indicate that the physical barriers have not fully controlled the increasing frequencies of animals that move out from Park. Collaborative Efforts One of the key collaborative efforts that it covers is the ranger collaboration to protect mountain gorillas in Mgahinga, Bwindi, Virunga and Volcanoes NPs. The collaboration includes aspects of tourism, community conservation and research/ monitoring. RMNP fully participates in the Trans - boundary collaboration initiative supported by Wildlife Conservation Society (WCS). RMNP jointly planned and conducted six coordinated patrols (one quarterly) with Parc National des Virunga (PNVi) Nord, along the contiguous boundary of 50 Km though confined to the southern sector of RMNP. The main challenge encountered in the current transboundary collaboration initiative is the communication barrier, movement restrictions at the borders, and failure to interpret the respective wildlife laws. RMNP participates Wardens Transboundary committee meetings to share field reports, experiences and plan for quarterly coordinated patrols. This needs to be strengthened with formalization of the coordinated patrols and to initiate cross border Tourisms. (Kiza Fredrick 2014). Monitoring programs on conflicts The management of Volcanoes National Park and other NGOs that help in the conservation of VNP has systematic procedures of monitoring conflicts. Through ranger based monitoring, conflict cases are reported and documented. Previously, Volcanoes National Park management relied on informal data gathered by the community based conservation education officers (ANICOs). However, since 2012, in order to facilitate the implementation of the 10 year- management plan, and wildlife compensation policy, the community partnership programme in Volcanoes National Park 19

22 initiated the wildlife data collection system in 12 sectors bordering the park, with more focus on the critical zone (mainly sectors of Musanze district). The database partially serves as a monitoring/ planning tool for the above highlighted policies and mechanisms against wildlife conflicts around Volcanoes National Park. Efforts in Uganda are mainly part of the Ranger Based Monitoring Programme and stored within the MIST database. 20

23 3.0.METHODS 3.1. Approach to the Baseline survey The baseline survey intended to examine the current situation and generate indicator values against which future GVTC project performance will be measured. Therefore, a mixture of methods involving quantitative and highly participatory qualitative methods were used as elucidated in this chapter. The survey process was structured into the four phases, as shown in Figure 2. Figure 2: Phases of the baseline study process 1. Planning and Inception 2. Fieldwork/Data collection 3. Data analysis and synthesis of information 4. Report writing and Validation 3.2.Target population from which to select participants The baseline study targeted the following stakeholders; a) Households living adjacent to the protected areas (PAs) b) Local communities living adjacent to the protected areas (PAs) within the GVL c) Protected Area Authorities (PAAs) and NGOs Research design, study setting and sampling A cross-sectional study design was used. A cross-section of the population living within the parishes (Uganda), sectors (Rwanda) and Groupments (DRC) adjacent to the PAs in the GVL was selected to participate in the survey. Sampling units were households adjacent to the PAs. Information was collected at a point in time i.e. in reference to the GVTC baseline year. Each household was represented by the household head (i.e. main income earners and the decision-makers of the household) or any adult responsible member of the household, found at home at the time of the survey. Multi-stage sampling design, that is, a two-stage stratified cluster sampling design was used. Based on information obtained from the Local authorities (e.g. PAAs about the human-wildlife conflict levels, parishes (Uganda), sectors (Rwanda) and Groupment (DRC) adjacent to the PAs the study area were stratified into 4 strata (i.e. low, moderate, high, and very high). Villages in each stratum were listed with their approximate sizes from respective local authorities or PAAs. UBOS (2014) National Population and Housing Census 2014 statistics (Household population) were also used to inform the sampling process. In the first stage, a probability proportional-to-number of households in a village sampling scheme was used to draw the villages. In the second stage i.e. within the selected villages a random sample of households was drawn. Proportional sample size allocation was used for strata allocation. The sample size (n) is a function of size of the population of interest (i.e. Households adjacent to PAs), the desired confidence level, and level of precision or permissible error. The formula below was used to calculate the representative sample size (n) of households. 21

24 z 2 ( * p(1 p)* Deff ) 2 1 n= * () i 2 r Whereby; n = the required sample size of households. Z = Value of the standard normal distribution given the chosen confidence level of 95% such that Z = 1.96 at 95% level. l = Level of precision or permissible error which is i.e. 0.1(Kish, 1965) p = Probability of success in choosing a household =0.5 (i.e. 50%). q = (1-p) = probability of failure= 0.5 (i.e. 50%). r = Response rate (80% minimally used in most household/socioeconomic surveys) by UBOS. Deff 1 = Design Effect 2 is as used by UBOS in sample surveys. Sample size characteristics and Response rate A total representative probability sample (n) of 1,043 households participated in the study (See Table 1 below) thus generating a Response Rate (RR) of 98.68% higher than the 80% RR anticipated at the inception phase of the study. Each household was represented by the household head or any adult responsible members of the household present at the time of the survey. Table 1: Sample size characteristics by gender of respondent Protected Areas (PA) No. of males in the No. of females in Total Number of sample the sample Households Rwenzori Mountains National Park (RMNP) Queen Elizabeth National Park (QENP) Bwindi Impenetrable National Park (BINP) Mgahinga Gorilla National Park (MGNP) Semuliki National Park (SNP) Virunga National Park (PNVi) Volcanoes National Park (VNP)-Rwanda Total sample size(n) Selection of PAAs and NGOs Purposive sampling was used to select all the PAAs and natural resources management focussed NGOs operating within the GVL. The PAs were also purposively selected because they areas of conservation concern which the GVTC has targeted. Purposive sampling is an informant selection tool widely used in ethnobotany. The purposive sampling technique, also called judgemental sampling, is the deliberate choice of an informant due to the qualities the informant possesses. Simply put, the researcher decides what needs to be known and sets out to find people who can and are willing to provide the information by virtue of knowledge or experience (Bernard 2002, Lewis & Sheppard 2006). 1 Design effect (DEFF) is a correction factor to account for the heterogeneity between clusters with regard to the measured indicator. 2 DEFF is defined as the ratio between the standard error using the given sample design and the standard error that would result if a simple random sample had been used. A DEFF value of 1.0 indicates that the sample design is as efficient as a simple random sample, while a value greater than 1.0 indicates the increase in the sampling error due to the use of a more complex and less statistically efficient design. 22

25 3.5. Selection of Households Households living adjacent to the PAs were selected from the complete list (sampling frame) obtained from the PAAs located within the sampled parishes (Uganda), sectors (Rwanda) and Groupements (DR Congo) adjacent to the PAs in the GVL and more so frontline areas which mainly experience human-wildlife conflicts. Households were arranged in ascending order, for each gender categories and headship (i.e. male and female household heads). Each household was allocated a unique ID-Number in the listframe. Households were then selected randomly using random number table generated in MS-Excel 2013 under each gender category. Each of the gender categories was proportionally allocated the number of households. A member of the local community accompanied the enumerator mainly to help identify the actual household in the village that matched the allocated unique Identifier number & Household head. Random selection of households is appropriate because it gives each household an equal-likely chance (i.e. probability) of inclusion into the final sample (Cochran, 1977; Wandiembe, 2009) Data collection methods i. Household Questionnaire The semi-structured questionnaire was administered by trained enumerators, to household heads (Figure 3) or other adult responsible member of the household found at home at the time of the survey. The questionnaire was used to collect data from household to address the objectives of the baseline study. The household questionnaire was administered to gather data on: a) socioeconomic and demographic characteristics of households, b) Severity of crop raiding, c) practices used by the different stakeholders to reduce crop raiding, d) Contribution of stakeholders (government, PAAs & NGOs) in reducing HWC especially crop raids. Detailed information was also collected on the following (i) crop species mostly affected, (ii) animal species responsible for the damage to crops, (iii) the perceived extent to which each primate species damages crops, (iv) seasonality of raids, and (iv) the preventive practices by different communities. Households ranked constraints and use a 5- Point Likert Scale to give perceptions about the origin and extent of the human-wildlife conflict and approaches used to overcome. Figure 3 Data collectors (right) administering a household questionnaire to a respondent 23

26 ii. Focus Group Discussions (FGDs) Three FGDs targeting each gender category (Figure 4) were conducted per PA to ascertain the origin of the human-wildlife conflict and practices used by frontline local community to cope with the human-wildlife conflict. An FGD guide was used to facilitate the discussions. Interventions were evaluated and analyzed. Each FGD consisted of a total of 8-12 participants who had not participated in the household questionnaire interview. Participants spanned a wide range of socioeconomic profiles e.g. gender (male & female), age (youths, elderly,), resource use groups e.g. farmers, hunters, herbal medicine practitioners, charcoal burner, firewood collectors, craft makers, timber dealers, etc. The FGDs generated qualitative information which was triangulated with the quantitative information obtained from the household survey questionnaire. Information was use to augment, explain and strengthen the quantitative findings on the objectives of the baseline survey. Figure 4. Focus Group Discussion with Women in Musanze district, Rwanda iii. Key Informant Interviews (KIIs) Key informant interviews were conducted with representatives identified from PAAs, local governments and NGOs and CBOs involved in conservation of wildlife in GVL. The interviews were conducted face-to-face, or through phone where it was applicable. The prevailing situation in relation to HWC and cross-border collaborations was examined. Resource mobilization, joint monitoring, collaborative research, experience sharing/knowledge transfer, among others were also discussed to generate the baseline information. The KII guide was developed to have questions that trigger off a discussion. iv. Desk-Review of documents Desk-based reviews of documents were conducted to generate background information, benchmarks or best practices. This process involved a critical examination and analysis of documents containing information related to the GVTC project to give background information about baseline indicators, planned activities and how to measure performance against targets. Literature on human-wildlife conflicts in the Albertine rift and Virunga ecosystem, strategies and cross border collaborations was also reviewed. v. Visits to project sites Site visits were conducted to affected sites adjacent to the PAs, for example, on-farm in order to ascertain existence of HWC and determine the magnitude of crop damage (if there was evidence at 24

27 the time of the survey), some gardens were inspected with respondent to ascertain and obtain estimates of loss or damage, size of garden, type of crop damaged, part of crop mainly damaged, distance moved, and prevention methods used by households. Digital photography was also used to document extent of crop raiding and enriched the visual evidence Data Processing a) Verification review of data collected During data collection, the research supervisors conducted routine reviews and field editing of questionnaires to identify errors (i.e. response errors and recording errors) and practical solutions to on a daily basis to ensure data quality. In addition, a team of experts from GVTC (Project staff) also came to the field and visited the consultants, research supervisors and enumerators to ascertain the progress of fieldwork activities, application of methods, coverage, quality of data collected, challenges, mitigations measures and also shared their knowledge and experiences and advised on the best practices that ensured data quality. b) Office editing and Data coding, cleaning and entry i. Office editing of questionnaires: Prior to data entry, the questionnaires returned from the field were further reviewed, edited and new responses from open ended questions in the Household survey Questionnaire were also coded (i.e. numerical codes attached). A data entry screen was also developed using EpiData 3.02 in which to enter data from the household questionnaire. EpiData was appropriate for data entry because it has the in-built auto skips and checks to minimize entry errors. ii. Data entry process and team composition: Data entry clerks were trained by the statistician on how to enter data into EpiData 3.02 software. Training helped to minimize some non-sampling error e.g. entry errors. Data was entered by a team of six (6) data entry clerks. The statistician supervised and verified the data entry process and reported progress to the overall Team leader and Co-team Leader. c) Data analysis and synthesis of information Final data screening: Data were exported to STATA 12 to support the cleaning process. Prior to final analysis, data were further cleaned using both visual and computed aided checks in STATA 12 in order to identify errors, missing cases, outliers and extremes. Data screening was done by generating summary statistics and using Box and whisker plots. Hard copies were retrieved based on Unique ID codes and further cross-checked to verify the errors identified. Final data analysis of quantitative and qualitative data: Both quantitative and qualitative data analysis techniques were used. a) Quantitative Data analysis Quantitative Data analysis- Quantitative data were analysed using descriptive statistics (i.e. means, frequencies and percentages) generated in STATA 12 and the Statistical Package for Social Scientists (SPSS) version 17. The summary statistics were also disaggregated by PA and key socio-economic characteristics of the households e.g. gender and other variables where it was applicable. 25

28 Qualitative data analysis- Qualitative data collected especially from FGDs and KIIs were analysed using thematic, content and discourse analysis. The transcripts were read to extract the emerging themes. Coding of themes was done in MS-Word 2013 software. The themes were collapsed into super themes. The findings were used to strengthen the interpretation of the quantitative findings and provide additional information. Final data analysis was done where salient issues were unearthed through triangulations. In some instances direct quotes from participants in FGDs were used. 26

29 4.0. RESULTS 4.1. Socio-economic and Demographic Characteristics of Households Gender of the HH respondents: Overall, out of the 1043 HHs that participated in the study, almost equal percentages of males and females (50%) responded to the HH questionnaire (Table 2). This is sufficient for capturing the differences in perception between the females and males Age of the HH respondents Overall, the average age of respondents was years (SD=0.4884). There is very little variability around the average age of respondents living adjacent to all the sample protected areas within the GVL. In relation to gender, the overall average age (42.5 years) fell within the female reproductive age group (15-49 years) and the working age group (14-64 years) for all people within the GVL. This has implications for population growth and the workforce in general Household size Overall, the average household size was 6 persons. Household sizes were highest in RMNP (average 7.2 persons) and lowest in MGNP and Volcanoes NP (5.3 persons). On average a household had 3.5 children and up to 58% of the household members were children. The households living adjacent to Virunga NP had the highest proportion of children (64.5%), while QENP was lowest (56.2%) Household income Overall the annual average household income earned by a respondent was USD 478. On average, female respondents earned relatively less income (USD477) compared to the males (USD520). This was the case around all sample PAs except Virunga NP in DRC. 27

30 Disaggreg ation RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall Baseline Survey on Conflict on Natural Resources Access in the Greater Virunga Landscape Land owned by households living adjacent to protected arears Overall, the average size of land owned by the respondent was 2.7 acres. Overall female respondents owned smaller average land size (2.34 acres) as compared to the males (3.04 acres). Overall female headed households owned small average land sizes as compared to the males headed households across all the seven protected areas within the GVL. Table 2: General socioeconomic and demographic characteristics of households Variable Gender Average age (years) Average Household size by gender. Average Number of Children per Household. Proportion of children in the household. Annual Average Household income (USD) Average size of land owned by HHs Males Females Total No. of HHs Males Females Overall Males Females Overall Males Females Overall Percentage Males , Females , Overall , Males Females Overall

31 Disaggregation RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall Baseline Survey on Conflict on Natural Resources Access in the Greater Virunga Landscape Education level of the HH respondents: Overall, more female respondents than males had no formal education, Table 3. The respondents mainly had primary education. The number of respondents decreased with higher levels of education Marital status of respondents Overall, most of the respondents were married. The scenario was similar around the individual PAs with percentages ranging between 56% (around QENP and 87% around Volcanoes NP. Table 3: General socioeconomic & demographic characteristics of households Variable Education level Marital status Religion Relationshi p to the head of the household No formal education Pre-Primary Primary O Level A Level Vocational education University degree Postgraduate Others Overall Single/Never married Married Cohabiting/Staying together Separated Divorced Widowed Total Catholic Protestant Moslem Pentecostal SDA Other Total Head of household Spouse Son/Daughter Brother/Sister Mother/Father In-law Grandchild Other Relatives Total

32 Religion of the respondents Overall, the respondents were mainly Protestants, Catholics or Seventh Day Adventists (SDA). The Protestant population was highest around BINP, while the catholic population was highest around Virunga NP, SDA was highest around RMNP, and Moslem was highest around Semuliki NP Relationship to the household head Overall, about 15% of the households were female headed. The highest percentages of female headed households were recorded around QENP and BINP while the lowest was recorded around RMNP Economic Activity/Livelihood strategy Overall, the main economic activity for most of the households living adjacent to the protected areas was agriculture (Appendix 8) and especially crop farming (82%). This was the case for all sample PAs ranging between 97% in Volcanoes NP. to 58% around Semuliki NP. Activities that could impact directly on resources in the park include Charcoal burning/selling, Firewood selling, Firewood selling, Handcraft and artefacts, Herbal medicine extraction and trade, bee keeping and Ecotourism activities : Availability of adequate food staples for household consumption in the 12 months On average, most households (71%) do not have adequate supplies of staple food for home consumption (Table 4). The Households are, therefore, generally food insecure across the whole of the GVL. The worst hit areas are around QENP and MGNP. Areas around Semuliki NP and RMNP are relatively the most food secure. Table 4: Availability of adequate food staples for household consumption in the 12 months. Protected Areas (PA) HHs had adequate/produced food staples in year 2015 HHs did not have adequate/produced food staples in year 2015 Total No. of HH RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

33 Out of the 742 households that did not have adequate food staples, the majority (53%) were headed by females (Table 5). Table 5: Inadequate food staples for household consumption by gender of respondent Household had inadequate food staples in year 2015 Protected Areas (PA) Male respondents in HHs Female respondents in HHs Overall No. of HH RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall The households did not have adequate food staples for an average period about five months per year (Table 6). Male headed households experienced slightly longer periods of inadequate food staples (4.67 months) as compared to the females (4.44 months). Overall, the longest period of inadequate food supply was reported around Semuliki NP covering 7.39 months with the females most affected (8.22 months). Table 6: Average number of months of inadequate food staples with HHs Average number of months of inadequate food staples within the Protected Areas (PA) households per Gender of respondent. Male respondents Females respondents Overall RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

34 Reasons why households experienced food insecurity The inadequacy of food was attributed to crop raiding (56.87%), unfavourable weather conditions such as drought/ flooding, limited land for cultivation, poverty, and crop diseases (Table 7). Table 7: Causes of inadequate availability of food staple within households adjacent to PAs. Cause of Food insecurity RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoe s NP Overall No. of HHs. Sold all farm produce Unfavorable weather conditions (drought, flooding) Crop raiding by wild animals Crops attacked by other pests e.g insects & birds Crop diseases Poor storage facilities Post-harvest losses Limited land for cultivation Inadequate labor force Household size is big Low income levels/poverty/low purchasing power. Others e.g poor farming practices

35 Head Spouse Children Head & spouse Head, spouse & children Total No. of HHs. Baseline Survey on Conflict on Natural Resources Access in the Greater Virunga Landscape Decision making on crop production, marketing/sales and income Overall, most of the household heads were the main decision makers on production (51%), marketing/sale (52%) and utilization of income generated from crops (54%), (Table 8). The situation is similar regarding decision making on livestock production, marketing/sales and income at household level (Appendix 8, Table 327b). With regard to Decision making on non-farm activities production, marketing/sales and income at household level, the trend is similar, but the percentages are on average generally higher (over 64%). Table 8: Main decision maker on Crop production, marketing and use of income generated Component Protected Areas (PA) Crop production Marketing & sale of crops Income generated from crops RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

36 Head Spouse Children Head & spouse Head, spouse & children Total No. of HHs. Baseline Survey on Conflict on Natural Resources Access in the Greater Virunga Landscape Decision making on resources from protected areas Overall most of the household heads made decisions on extraction of resources (63%), marketing/sale (64.4%) and utilization of income generated from extraction of resources (64.4%), Table 9. However, in a few cases, the spouses also made decisions on extraction of resources, marketing and sales as well as utilisation of income generated from non-farm activities. Table 9: Main decision maker on resource extraction from the PA, marketing and income generated Component Protected Areas (PA) Resource extraction from the PA Marketing & sale of resources extracted from the PA Income generated from resources extracted from the PA RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

37 4.1.19: Welfare and WASH Indicators Most household heads owned the main house in which they lived (Table 10). Table 10: Mode of ownership of the main house Protected Areas (PA) Owned Rented Owned by relative House provided as part of work Other Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall The main houses were predominantly roofed iron sheets (71%) or grass (21%). In general, grass thatched houses constitute less than 50% with the highest percentage recorded around Semiliki NP. The highest proportion of houses roofed with Iron sheets was recorded around BINP and MGNP (Table 11). Table 11: Main Roofing material of the main house Protected Areas (PA) Grass Iron sheet Tiles Tarpaulin Other Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall The walls of the main houses were mainly of mud and wattle (61.5%), followed by bricks/stone (27%), Table 12. Table 12: Main wall material of the main house Protected Areas (PA) Mud & wattle Bricks/ Stones Iron Sheet Wood Plastere d Others Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

38 The main floor materials of the main house were earth (80%) and cement (16%), Table 13. Table 13: Main Floor material of the main house Protected Areas (PA) Earth Cement Wood Tiles Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall The main toilet facility used by the household was the pit latrine (96%), Table 14. There are still households that use the bush as the main toilet facility all through the GVL with the percentages ranging between 0.7 (around Volcanoes NP) and 4.1 (around Virunga NP). Table 14: Main type of Toilet facility used by the household Protected Areas (PA) Pit latrine Flush toilet Bush Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall Overall, the main type of energy used for cooking by the household was firewood (93%), Table 15 while the lighting used in the min house by the household was the Kerosene tin lamp (Candle) (76.4%), Table 16. Table 15: Main type of energy used for cooking by the household Protected Areas (PA) Firewood Charcoal Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

39 Radio Television Print media e.g Newspapers & mag Social media e.gfacebooks, Whatsup Telephone/mo bile phone Religious institution e.g church Word of mouth e.g at drinking joints Others Total No. of HHs. Protected Areas (PA) Tin lamp uses paraffin Baseline Survey on Conflict on Natural Resources Access in the Greater Virunga Landscape Table 16: Main type of Lighting used by the household Lantern Pressure lamp Electricity Solar power Fuel wood Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall The main mode of access to information used by the household was the radio (70%), followed by mobile telephone (12.5%), and word of mouth (13%), Table 17 Table 17: Main mode of access to information used by the household Protected Areas (PA) RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall Overall the main means of transport used by the household was walking (78.6%), followed by motorcycles (11.4%) and bicycles (6.9%), Table 18. Walking was relatively east important around RMNP and QENP but most important around BINP. Table 18: Main mean of transport used by the household Protected Areas (PA) Walking Bicycle Motorcycl e Bus/Mini-bus Trucks/Lorries Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

40 Piped water into house/ Compound Public tap/stand pipe Borehole Protected dug well/spring Rainwater Unprotected dug well/spring Surface water (Lake, river, stream Total No. of HHs. Baseline Survey on Conflict on Natural Resources Access in the Greater Virunga Landscape Overall, the households did not have proper garbage disposal methods (53%), followed by dumping in garbage pit (44%) and burning/incinerating (2%) among others, Table 19 Respondents around QENP and Virunga had the smallest percentages of respondents without proper disposal methods. Table 19: Main method of garbage disposal used by the household Protected Areas (PA) No proper disposal mechanism Dump in Garbage pit Burn/Incine rate Others Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall Overall the main source of water for domestic used by the household was the Public tap/stand pipe (47%), followed by Surface water e.g. Lake, river, stream (15%) and piped water into house/compound (15%). Nearly 5% used a borehole (Table 20). Table 20: Main source of water for domestic use by the household Protected Areas (PA) RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

41 Distance to the nearest facility or service centre around PAs in the GVL The average distance from households to the nearest social service facility ranged between , Table 21. In general, the respondents moved over distances of at least half a kilometre to reach the nearest social facility. Table 21: Estimated average distance (km) from home to the nearest facility or service centre Estimated average distance (km) from home to the nearest facility or service centre Facility or service centre RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoe s NP Overall No. of HHs Source of water for drinking Health facility School: Market Motorable road Agro-processing & value addition facility Police post PAA station/office

42 Percentage Baseline Survey on Conflict on Natural Resources Access in the Greater Virunga Landscape 4.2. Perception on Exploitation of Natural Resources as Origin of Conflict Perception on origin of conflict Overall, most households (about 68%) perceived restriction of exploitation of naturals resources in PAs as the main origin of conflicts around Protected Areas. This is generally the scenario around the PAs in the GVL (Figure 5). For example, a total of 92% of households mentioned than they did not access water from the park because it was prohibited which creates conflict between PAAs and frontline communities RMNP QENP BINP MGNP Semuliki NP 20 Virunga NP Volcanoe NP 0 RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoe NP Ptotected Area (PA) Figure 5: Household that perceive access to PA resources to be the origin of conflict (n=709) Patterns of Extraction of the resources from Protected Areas Overall up to 68%of the respondents reported that they did not harvest resources from the park (Table 22). The percentage that acknowledged harvesting resources from the park varied from one park to another. The highest percentage of people who harvested resources from the park was recorded in Semuliki NP (68.89%) followed by RMNP (41.16%) and Virunga NP (30.14%). The lowest percentage of resource harvesters was Volcanoes and BINP. Table 22: Extraction of the resources from the protected area by the household Protected Areas (PA) Household extracted resources from the park Household did not extract resources from the park Total No. of HHs. RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall

43 Extraction of firewood from the park Overall, nearly one in every four households obtained firewood from the protected area (Table 23). They mostly harvest the tree stems and only a tenth of the household harvest the whole tree. Household firewood extraction also varied with national park. Semliki NP has the highest percentage of households extracting firewood. No firewood extraction was reported in Volcanoes NP. Table 23: Extraction of Firewood from the park Protected Areas (PA) Household which extracted Firewood from the park Quantity (Bundles) of firewood extracted from the park Main part of resource extracted Households Average No. of Bundles Stem Whole plant RMNP QENP BINP MGNP Semuliki NP Virunga NP Volcanoes NP Overall Figure 1 Carrying firewood-collected from the park as one of the resources permitted through MOUs with Semiliki NP authorities Extraction of firewood from the PAs is on a monthly (37%), weekly (29%), quarterly (13%), or daily (8%) basis (Table 24).In RMNP the majority of the respondents reported that firewood is extracted weekly, in Semliki and MGNP firewood collection is done monthly. The firewood is mainly collected in the morning (74%) and early afternoon (13%), Table 25. In Virunga, BINP and Semliki NP firewood collection was reported to take place anytime, though they seemed not to indicate night collection. 41