www.elsevier.com/locate/pce

Physics and Chemistry of the Earth 31 (2006) 659–668

Access to water and related resources in Ngamiland, Botswana:Toward a more critical perspective and sustainable approach

Phemo K. Kgomotso a,*, Larry A. Swatuk b

a IWRM Programme, Faculty of Science, University of Western Cape, Bellville, Private Bag X17, South Africab University of Botswana, Harry Oppenheimer Okavango Research Centre, Private Bag 285, Maun, Botswana

Accepted 10 July 2006

Abstract

Governance structures in Botswana are highly centralized and top-down in orientation. For water and related resources managementin rural areas, this creates particular difficulties – from lack of decision-making capacity to limited human and financial capital at thelevel of the resource base. In Ngamiland, government is currently undertaking the Okavango Delta Management Plan project as partof its commitment to the Ramsar Convention on Wetlands. The project purports to develop an integrated management plan basedon an ecosystems approach. Meaningful participation by local people is a requirement of the process. Data from 43 village meetingsundertaken under the auspices of the ODMP process reveal that local people’s access to their resource base is diminishing. Informationfrom key informant interviews and a variety of government documents, however, suggests that policy makers are either unaware of oruninterested in this growing problem. Although citizens and government are engaged in an on-going dialogue, there is little evidence tosuggest that policy and practice are moving toward sustainable solutions for all. This article highlights these issues in the hope that amore meaningful dialogue among all stakeholders may be undertaken.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Okavango River Basin; Ngamiland; Governance; Sustainable resource use; Integrated water resources management; Access; Livelihoods

1. Introduction

Access constitutes a key indicator of state commitmentto integrated water resources management (IWRM) (Beuk-man, 2002). For Van der Zaag (2005), IWRM involves‘reconciling basic human needs, ensuring access and equity,with economic development and the imperative of ecologi-cal integrity, while respecting transboundary commit-ments’. The focus of this article is the complexity inproviding access to water and related resources to multiple,highly differentiated stakeholders in a physically remotebut increasingly politically important area of rural Bots-wana. Ngamiland is home to approximately 7% of Botswa-

1474-7065/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.pce.2006.07.004

* Corresponding author. Address: Private Bag 285, Maun, Botswana.Tel.: +267 72192123.

E-mail addresses: phemok@yahoo.com (P.K. Kgomotso), swatukla@mopipi.ub.bw (L.A. Swatuk).

na’s entire population. The vast majority of these peopleare either wholly or partially dependent on direct use ofthe natural resource base for their livelihoods. Settlementpatterns and human activities reflect the dramatic varia-tions of an arid environment with a significant sourceof perennial, seasonal and ephemeral water – the LowerOkavango River Basin.

The Okavango River rises as the Cuito and CubangoRivers in the mountains of Angola, passes through Nami-bia and empties inland as the Okavango Delta in Botswana(Mendelsohn and el Obeid, 2004). The size of the delta var-ies dramatically from year to year, depending primarily onrainfall in Angola. The delta is the focal point of Botswa-na’s tourism industry, a fast growing sector that is nowthe second largest contributor to gross domestic product(GDP) in the country. Due to a variety of internal (rapidpopulation growth; land use conflicts; resource degrada-tion) and external (real and prospective development plans

660 P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668

in the upstream riparian states, Namibia and Angola) fac-tors, the government of Botswana declared the entireLower Okavango Basin (68,640 km2) a wetland of interna-tional importance and became a party to the Ramsar Con-vention in April 1997 (Swatuk, 2003). As a party to theConvention, the government of Botswana, with interna-tional support, initiated the Okavango Panhandle andOkavango Delta Management Plan processes. Two keyconditions of the process impact current and will influencefuture resource access, use and management practices: (1)an integrated and holistic ecosystem approach is to beadopted regarding development and management of theresource base; and (2) ‘primary stakeholders’, defined asthose people dependent for all or a significant part of theirlivelihood on the resources of the area, must be involved ina significant way in all planned activities. In addition, theirkey livelihood interests should be held above the interestsof others less dependent on the resource base (ODMP,2005).

Perhaps unexpectedly, the ODMP process has revealeda deep contradiction in Botswana’s water resources man-agement plans and performance. On the one hand, avail-able data shows that countrywide the proportion of thepopulation without access to safe drinking water – a keyMDG indicator – is 12.1% (CSO, 2001), an excellent recordfor a developing country. On the other hand, the ODMPstakeholder consultation process reveals a disturbing pic-ture of heightened levels of resource use conflict and declin-ing access to water and related resources among primarystakeholders in Ngamiland.

As shown below, the present centralized, top-down insti-tutional matrix impedes realization of IWRM principlesand practices. A close analysis of primary stakeholderissues reveals the need for planning at basin level or, inthe case of the transboundary Okavango River, at sub-basin level (the Lower Okavango Basin (LOB)) with a clearview of management practices and developmental inten-tions upstream in Namibia and Angola. The concernsarticulated by stakeholders living within the LOB demon-strate the importance of devising policy within the uniqueparameters set by the river and its ecosystems. However,standard government responses to the issues raised by peo-ple living within the LOB – as revealed in village meetings(known as kgotla) and key informant interviews – showwhat amount to as missed or ignored opportunities formoving toward IWRM in the study area. This article high-lights these issues with the intent to promote further dia-logue toward more holistic and sustainable resource usein Botswana.

2. Methodology

The analysis employs an interpretative social scientificresearch method (Terre Blanche and Kelly, 2001). ‘Inter-pretive research . . . relies on first hand accounts, tries todescribe what it sees in rich detail and presents its ‘‘find-ings’’ in engaging and sometimes evocative language’

(Kelly, 2001) The purpose of our interpretive analysis isto provide ‘thick description’ so as to better understandproblems with sustainable water resources managementin the study area.

In order to achieve ‘exhaustion/saturation’ in data gath-ering and analysis, a triangulation strategy was employed(Kelly, 2001). Triangulation was employed in five differentways: data triangulation; investigator triangulation; theorytriangulation; methodological triangulation; and interdisci-plinary triangulation.

Data was collected from different sources over time.Residence in the study area provides a participant observerperspective. Primary data sources include formal semi-structured interviews with key informants in each of theinstitutions involved in water management in the LowerOkavango Basin. These were recorded, transcribed, trans-lated (where necessary) and written in English and codedaccording to relevant resource use themes. Several informalopportunistic interviews and discussions were also used.This is differentiated in the text below by the terms ‘inter-view’ (where it was formal and semi-structured) and ‘per-sonal communication’ (where it was opportunistic andperhaps only designed to gain clarification on a particularissue).

Analysis was undertaken of raw data on naturalresource use issues in the study area. This data was col-lected from 43 village meetings (kgotla) held between 31January 2005 and 12 May 2005 as part of the OkavangoDelta Management Plan (ODMP) project process.Approximately 18% of Ngamiland residents attended thesemeetings. The percentage of population attending kgotla

varied inversely with village size with less than 0.3% of peo-ple in major villages (i.e. Maun and Gumare) attending.While this seems to suggest that our research findings areexploratory rather than definitive, the similarity of issuesand concerns across all 43 cases leads us to believe thatthe study achieves theoretical saturation (Kelly, 2001).

The data was compiled in two unpublished documents.Bendsen (2005a) presents the unanalysed transcripts fromthe kgotla meetings. Bendsen (2005b) breaks down thetranscripts according to key component areas (12 in all)of the ODMP project. These transcripts were read andkey-word coded. Recurrent themes, such as decliningaccess to fishing grounds or slow processing of land claims,helped structure subsequent key informant interviews.Despite the high quality of the raw data, we did not attendany of these meetings. However, similar meetings attendedfor other purposes (e.g. Wirbalauer et al., 2003; Swatukand Rahm, 2004) show the transcribed data from Bendsen(2005a) to be consistent with expectations.

3. Structures of governance

In Botswana development planning in general, andwater resources management in particular, are highly cen-tralized (Swatuk and Rahm, 2004). The institutional matrixand mechanisms for delivery are designed to implement

P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668 661

policy made and decisions taken at the centre, in the capitalcity of Gaborone. District Development Plans (DDPs)therefore reflect the overall national goals as articulatedin various National Development Plans (NDPs), in thiscase Ngamiland DDP 6 and NDP 9. The belief is that sucha matrix will facilitate smooth delivery while maintaining afirm hierarchical chain of command, thereby avoiding con-fusion regarding or deviation from Central State-articu-lated developmental goals (Edge and Lekorwe, 1998).This process is reinforced by the near absolute dependenceof Districts on the Central Government for human andfinancial resources.

3.1. Water

In the case of water resources, policy is informed by theWater Act (1968), and set by the Ministry of Minerals,Energy and Water Affairs (MMEWA). The guiding docu-ment is the National Water Master Plan (NWMP, 1992),currently in revision. As an arid country, policymakersare centrally concerned with securing supply through avariety of supply-side mechanisms: dams, pipelines, bore-hole and wellfield development, all of which are dependenton sophisticated technology and large amounts of financialand highly-skilled human resources (Swatuk and Rahm,2004). Commitment to demand management is limited to

Map 1. Settlements in the Lower Okavango River Basin. Source: Harr

typical punitive measures (e.g. cut-offs for non-paymentin urban households) (Goldblatt et al., 2000). Implementa-tion is parcelled out to a number of actors. In the studyarea, this includes centrally the Department of WaterAffairs (DWA) and District Council. It tangentiallyinvolves numerous other departments as their activitiesall require secure water supplies. For example, the Depart-ment of Animal Health and Production is responsible forsmall dams development.

In Ngamiland, DWA is responsible for supplying Maunand Gumare (designated ‘major villages’ of more than 5000people) with water and monitoring water levels in theOkavango River Systems. For rural (small/medium village)water supplies, DWA designs and constructs all supply sys-tems and hands them over to North West District Councilfor operation and maintenance. The NWDC is thereforeresponsible for supplying water to all gazetted villages(i.e. according to the National Settlement Policy, thosewith 500 people or more living in government-approvedareas) (see Map 1). Gazettement is a tool used by govern-ment to entice rural people to cluster in accessible and ser-viceable areas. This means that tens of thousands of peopleliving in small, scattered, ungazetted settlements areresponsible for their own water supply. These are usuallysourced from privately owned boreholes, hand dug openwells, rivers or hauling water from major villages or towns.

y Oppenheimer Okavango Research Centre, October 2005, Maun.

662 P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668

Far from constituting a ‘crisis’, however, many are satisfiedwith this arrangement – as long as they retain access tothose resources necessary for the maintenance of their live-lihoods and there is visible and tangible support from gov-ernment toward socio-economic improvements (Bendsen,2005a).

Map 2. The Okavan

3.2. Land

Issues of land and water are not easily separated, espe-cially where household subsistence is so dependent on thenatural resource base of seasonal wetlands. Moreover, ina country heavily dependent on sub-surface water, ‘every-

go River Basin.

P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668 663

thing begins with access to land – even water. So if youwant water, you must first get land’ (Magole, pers. comm.).Land Use Planning is the responsibility of the Ministry ofLands and Housing which formulates policy and legisla-tion with ‘optimal physical and economic use of land’ theguiding principle (NWDC, n.d.). At the District level, theMinistry operates through the District Land Use PlanningUnit and the Land Board. Land Board is responsible forland administration and management while the DistrictLand Use Planning Unit (DLUPU) offers technical adviceto the Land Boards on all land use related activities. TheLand Board is the statutory land allocation authority thatholds land in trust for the citizens. In large districts likeNgamiland the main Land Board works through Sub-LandBoards for allocation of customary grants for grazing, ara-ble, and residential purposes. The main Land Board alsogrants leases under common law for commercial, indus-trial, residential and agricultural purposes. It thus also set-tles disputes. The Land Board is required to consult theDistrict Council and other local authorities in the formula-tion of any policy pertaining to any land use. Given theimportance of land in an arid environment, the land boardsare dominated by the most powerful political and eco-nomic actors in the country.

4. The study area

Ngamiland district occupies the northwestern part ofBotswana. Its northern and western boundaries form partof the international boundary between Botswana andNamibia (Caprivi Strip). It is bordered by Chobe districtin the northeast, central district in the east and Ghanziin the south. The whole Ngamiland district covers anarea of 109,130 km2 (NWDC et al., n.d.). The district liesbetween latitudes 18 030 0S–20 000 0S and longitudes23 015 0E–24 030 0E (see Map 2). The whole of Botswana’spart of the Okavango River Basin (divided into the pan-handle and the delta) is within this district.

Maun is Ngamiland’s district headquarters, its largestsettlement with a population of about 44,000 (CSO,2001), and covers a total planning area of 446 km2. Thereare 124,712 people in the Ngamiland district, marking a5% increase from the 1991 census (CSO, 2001). Seventyper cent of the Ngamiland population is settled within10 km of the Okavango River main channel (CSO, 2001).People depend on the nearby water bodies for resourcessuch as water, fish, building and handicraft materials, jobs,farming and transportation. The surface waters of theOkavango provide a ready source of water for livestockwatering. From the early days, farmers have sustained theirlivestock from the waters of the Delta, and grazed them onthe grasses that grow in the seasonal swamps when thewater recedes. Flood recession (molapo) farming practiceis a popular subsistence farming tradition practiced byinhabitants of Ngamiland (27% in 1997), alongside drylandfarming (73% in 1997). Research suggests that only half of

the potential area that can be ploughed under molapo isbeing used (Bendsen, 2003).

The paved road network runs along the western andsouthern portions of the river basin, facilitates settlementalong these routes, and informs official government settle-ment policy. Approximately 1270 km of disease controlfencing has been erected since 1964 (Van der Post, pers.comm.) to inhibit wildlife–livestock interactions. Many set-tlements are located within fenced areas, so complicatinggovernment development and household livelihoodstrategies.

Settlements vary dramatically in size, with recognisedlocalities varying from 0 to 44,000. The National Popula-tion Census identified 562 settlements in the study area ofwhich 478 had fewer than 250 people (CSO, 2001).Forty-eight are gazetted villages so qualifying for govern-ment services. However, several of these are communitiesof considerably fewer than 500 people gazetted under theRemote Area Development Programme, six being locatedin the Delta. There are eight settlements – Sexaxa (534),Boro (842), Sekondomboro (655), Eretsha (616), Gudigwa(732), Roye (615), Samochima (847) and Etsha 1 (614) –large enough for village status, that remain ungazetted.Clearly, this variation in size and location poses severelogistical and technical difficulties for government servicedelivery; hence, the current policy of encouraging peopleto relocate to gazetted areas.

5. Resource use issues raised by primary stakeholders

People living in Ngamiland use resources found both onland and in the water for their livelihoods. They fish in theriver, use it as a transport mode, and harvest reeds from theriverbank. On land, people harvest grass for thatching,trees for building material for houses and to also makeboats for water based tourist activities; they also brew beerand palm wine, and harvest a variety of medicinal and edi-ble roots and plants. The rivers and wetlands are homes tothe vast wildlife species that communities use for non-consumptive tourism activities or hunt for subsistenceuse. Some of this activity is seasonal, so complicatingstrict use/non-use, access/no access-oriented rules andregulations.

The communities raised concerns on all the issues thataffect the utilisation and sustainability of natural resourcesin the Okavango Delta. Out of 12 components of theODMP project, eight are involved in the direct utilisationof the Delta resources. Bendsen (2005b) clustered issuesraised in the kgotla meetings in terms of these components.The percentage frequency of the issues raised by communi-ties is as follows: Hydrology and Water Resources (15.8%);Wildlife Management (32.8%); Sustainable tourism andCBNRM (7.5%); Sustainable fisheries utilisation and man-agement (6.9%); Vegetation Resources Management (13%);Land Use Planning and Land Management (N/A); WasteManagement (4.3%); and Sustainable Livestock Manage-ment (6.7%).

664 P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668

With regard to ‘access’, our key-word coding of thesetranscripts showed that community concerns could be clus-tered in four areas: access to sufficient good quality water;access to quality land; access to a variety of naturalresources necessary to peoples’ livelihoods; and good gov-ernance. We now turn to a discussion of these issues.

5.1. Water issues

Primary stakeholders are overwhelmingly concernedwith channel blockages and the flow regime of the Delta’swaters. In 22 villages, communities complained that over-growth of vegetation such as hippo grass, Salvinia molesta

(the aquatic alien species commonly called water hyacinth)and reeds blocks their access to the river because they can-not use their boats. As a result of lack of access through theriver, they struggle to reach their fishing grounds, veldproducts and grass harvesting grounds in forests, andother villages in other parts of the river. The communitiesrequested that these blockages be cleared by DWA, butDWA’s response is that it has no mandate to clear channelblockages. In some cases, communities also complainedthat channel blockages prevent water from flowing to theirvillages and as a result they do not have water. Some com-munities want answers to why channels are blocked whileothers blame DWA for disturbing the flow of the Deltaand elephants for pushing trees into the river therebyblocking flow. Members of several different communitiesrequested that they be engaged in clearing channels of deb-ris and Salvinia molesta.

According to the head of Aquatic Vegetation Control,K. Naidu, DWA only clears channels to get to its gaugingstations to collect data on flow levels of the Delta but can-not clear the rest of the river due to lack of funds (Inter-view, 19/09/05, Maun).

With regards to the LOB’s flow regime, there is generalconfusion as well as scepticism as to why the Delta waterflow shifts, fluctuates from year to year or dries up com-pletely in some parts. Some communities suspect largeabstractions and use or storage upstream of the river, eitherin Angola or Namibia. Even though DWA understands thebasic regime of the Delta, no regular information is givento the public on the causes of channel blockages, fluctuat-ing water flows, or any other factors that affect the flow ofthe Delta.

Water pollution and water supply concerns figuredprominently in kgotla meetings. Pollution concerns wereraised in the four villages of Makalamabedi, Seronga, Dit-shiping and Matlapaneng. In the village of Makalamabedi,members of the community complained about a watertreatment plant in the village that disposes grey water intoa lagoon used for watering livestock. Pollution of drinkingwater elsewhere was blamed on elephants since in manycases drinking water sources are shared between communi-ties and wildlife. To reduce this problem, communitiesrequested that boreholes be drilled elsewhere for elephants,far away from the villages. DWA’s response to this is that

the movement of elephants cannot be controlled (Goja-mang, 19/09/05, interview, Maun). In Matlapaneng andSeronga, communities raised concerns about the possibilityof pollution by fertilisers used in vegetable gardening onthe riverbanks. DWA is aware of this possibility but thereis no mechanism in place to monitor these water use activ-ities (Ramoshibidu, 19/09/05, interview, Maun).

In eight other villages, lack of water was regarded as oneof the main water problems people face. A lack of reticu-lated water supply was blamed on the danger to people’slives by wild animals. ‘‘In all the smaller settlements andcattleposts there are no standpipes and kids are exposedto crocodiles when they fetch water from the river. Willwater be provided to smaller settlements?’’ (K. Katwana,Sekondomboro kgotla). Residents in remote non-gazettedareas also expressed interest in reticulated water supply,suggesting ignorance of official government settlementpolicy.

In some cases where water is provided through publicstandpipes, people complained about the long-pendingapplications for connections to individual households. Inmany cases, borehole pumps that provide water to the vil-lages were broken. In the ungazetted village of Ditshiping,people reported having been denied access to a DWA bore-hole even though they had no other source of potable watersupply. Residents did not understand that as an ungazettedsettlement they were not entitled to this water even thoughit was located nearby. Rather than seek a sustainable solu-tion, DWA’s practice is to deter use through the confisca-tion of buckets (Ditshiping kgotla).

Many communities raised concerns about access towater in areas away from the river. Subsequent to beingasked to move their livestock from the river to the drylandsin order to minimise conflict with wildlife at the riverfront,communities raised concerns of lack of water in drylandsand trading one set of predators (crocodiles) for another(lions) (Samochima kgotla).

Clearly, the scarcity of appropriate information – anissue raised in all meetings – complicates sustainable waterresources management. Among many alarming examples isthat of the village of Sankuyo where people complained ofthe condition of ‘our boreholes’, as they had been inun-dated with floodwaters. Villagers wished to know if thewater was still ‘safe to use . . . for drinking’ (Sankuyo kgo-

tla). In response, a Tawana Land Board representative sta-ted that the boreholes in question were in fact test sites forthe Maun groundwater supply (Sankuyo kgotla). Residentswere unaware that the water they were using had not beenintended for them, but for the growing settlement of Maunlocated some 75 km south; neither did they know if it wasfit for human consumption.

5.2. Land issues

In the ODMP project, land issues are the responsibilityof the Sustainable Land Use and Land Management com-ponent. The institution responsible for this is the Tawana

P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668 665

Land Board. Key concerns raised in the 43 meetings wereon allocation regulations, allocation procedures, land userights and conflicts among others. In most cases there isgeneral discontent among communities that their tradi-tional land use rights are and will continue to be erodedby new regulations. For instance, a new regulation wasintroduced recently to stop the use of floodplains for arableagriculture, a traditional agricultural practice dating backhundreds of years. This regulation, which currently appliesto the panhandle area, states that no allocations for arableuse should be made within a 500 m strip from the flood-plains. Traditional farmers pointed out that prior to theenforcement of the Panhandle Management Plan theyhad been advised by Land Board to take their livestockaway from the riverfront and use this fertile area for culti-vation instead. Farmers who grow vegetables that need tobe irrigated or watered by hand are particularly dependenton having their small gardens near open water sources. Indefence of floodplain farming, speakers in several villagesinvoked history, stating that molapo had been practicedfor generations in their area. In response to the 500 m rule,a representative of the Tawana Land Board explained thatploughing in floodplains ‘might have negative environmen-tal impacts like erosion and water pollution’ (Mongati,Eretsha kgotla).

Elements of traditional versus modern values and prac-tices are woven through these issues. For example, resi-dents in Etsha 1 and Seronga complained thatapplications from foreigners (i.e. tourism company ownersand operators) are privileged ahead of local residents. Cen-tral government’s desire to profit from wildlife-based tour-ism have led to the establishment of these enterprises inareas of traditional seasonal molapo farming, seasonaland perennial fishing grounds, in the midst of communaltransit routes, and areas where veld products have been his-torically collected. In response, the Fisheries Division rep-resentative from the DWNP asked the communities to referthe matter to Land Board. He stated, ‘‘traditional subsis-tence user rights should not be affected by the commercialleases’’ (Manyamane, Ditshiping kgotla). The fact is, how-ever, that they are.

Land Board is also accused of discrimination, favourit-ism, corruption, lack of communication with the landusers, conflicting regulations and practices, and longprocesses and procedures for land allocation. Villagersthroughout the study area provided evidence in supportof each of these claims.

5.3. Livelihood issues

Issues of access to key resources found on land and inwater that sustain livelihoods stem from a number of fac-tors including new laws on resource use that often invali-date or challenge traditional harvesting techniques anduses, lack of regulations to control the use of resources,depletion and degradation due to unsustainable use,

resource allocation inequities, pollution, conservation, spe-cies loss and others.

Fishing is one of the most prevalent and important live-lihood activities practiced by people living in the Delta andanywhere else in Ngamiland where there is water. Thus it isunderstandable why concerns related to fishing arose in allof the kgotla meetings. Concerns centred on lack of regula-tions to control fishing, fear of fishing regulations, lack ofaccess to traditional fishing grounds, and legislationregarding appropriate fishing techniques. Declining fishstocks and increased competition and conflict among bothcommercial and artisanal fishers were common prob-lems (Sankuyo, Mababe, Khwai, Maun and Boro kgotlas).People were also of the opinion that since its transfer toDWNP the Fisheries Department was more concernedwith conservation qua conservation, than in achieving sus-tainable fisheries practices (Seronga kgotla).

With regards to wildlife species, concerns centred on acombination of declining numbers of animals and birdstraditionally hunted by people and regulatory frameworksthat impeded their ability to hunt, varying from conserva-tion measures to long distances to travel to secure huntinglicenses. Limited enforcement combined with heavilyrestrictive regulations led many people to ‘hunt for thepot’, i.e. to poach wildlife. People’s abilities to farm, raiselivestock, fish and collect veld products were beingrestricted by government policies that favour large mam-mal preservation. Elephants, and large predators such ashyena, lion, cheetah and wild dog are problem animalsthroughout the study area. Other villages complained ofhaving to compete with hippo, crocodile and rhino foraccess to resources. As local residents become squeezedby changes in land use policy and practice on the one hand,and their own rising populations, on the other hand, theyfind themselves increasingly at risk from wildlife whichshare the resource base.

In response, government has attempted, where possible,to separate people from wildlife through the erection offences. In general, residents of Ngamiland support thefencing policy, but feel that the fences are poorly main-tained. However, a speaker from Gudigwa stated that thefence had not divided wildlife from people and livestock;it had merely divided various forms of wildlife. Not onlyhad it failed in its intended mission, the speaker stated, ithad also reduced villagers’ access to arable lands (Gudigwakgotla).

DWNP imparted contradictory information to resi-dents. In Habu it was stated that ‘The suggestion to elec-trify the buffalo fence was submitted to the departmentbut it was rejected’ (Habu kgotla). One day later, in Tsau,the same person, O. Sapula, stated that electrification was‘subject to the availability of funds’ (Tsau kgotla). And,in Shorobe, the Agricultural Demonstrator, Mr. Kwerepe,claimed, ‘The Minister of Agriculture mentioned in hisspeech that the electrification is being considered; however,it cannot be implemented without an EIA’ (Shorobekgotla).

666 P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668

Compensation paid by government for stock loss due topredation was also a central concern, as was compensationpaid by government for cattle it killed for straying onto thewildlife side of the buffalo fence. Government policy statesthat buffalo which stray into livestock areas will be drivenback; but cattle which stray must be killed. The reason forthis is to protect Botswana access to European Union beefmarkets. EU policy immediately freezes all exports from acountry where it can be shown that a cow has been in pos-sible contact with disease carrying wildlife. Thus, smallholders living alongside wildlife and who rarely if ever selltheir cattle to an exporter, are therefore paying a dear priceto maintain markets accessed only by the large cattleranchers of the country whose stock is located far frompossible contact points.

With regards to vegetation resources, most people com-plained about lack of grazing lands and depletion of graz-ing resources due to concentration of livestock in smallgrazing areas since most have been fenced in to keep wild-life and livestock apart. Some have also lost access to landto large commercial cattle and/or game farmers, putting afurther squeeze on an already narrow resource base. Lossof vegetation resources was also attributed to uncontrolledveld fires. Some complained about the banning of use offire as a vegetation management tool. People in Eretshacalled for regulations for veld products utilisation, whilesome thought it would be better for communities to settheir own regulations because the government only comesup with limits and control measures. In Mababe, peoplelamented on the loss of power by traditional leaders whoin the past could control the use of natural resources.

5.4. Governance issues

For many primary stakeholders, loss of access to theirresource base reflects poor governance. In all 43 kgotla meet-ings, communities expressed dissatisfaction about the waythings are done: from basic communication and informationdissemination to decision-making processes and the imple-mentation of rules and regulations. In some cases the differ-ent institutions agreed that they are not efficient (e.g. DWNPrespondent in Shakawe and Gudigwa kgotla). Issues of par-ticipation, consultation, communication, limited knowl-edge, unclear responsibilities and lack of accountability bythe government and implementing institutions in particularwere some of the key governance issues mentioned.

5.4.1. Participation

There is evidence that communities are not beingincluded in implementation of projects and would like toparticipate more. ‘Participation’ includes not only consul-tation but implementation as well. However, many peoplefelt that government devises and implements policies andplans ‘without listening to communities’ (Sankuyo kgotla).The centralized style of decision-making in Botswana,however, leaves little room for meaningful participationby people at the level of the resource. Passive participation

through consultation is common, as it draws on the kgotla

meeting as a traditional form of governance – where deci-sions continue to be taken by elites (Good, 1992).

5.4.2. Institutional capacity

In many of the villages, there was a serious lack ofknowledge about procedures and general informationabout how things get done. This is compounded by slowor no response from government. Lack of transparencyand corruption on the part of the Tawana Land Boardwas a key sticking-point. A speaker from Bothatogoobserved, ‘Just last year, raffles for commercial livestockfarms were won only by Land Board officials.’ One speakerfrom Matlapaneng requested, ‘Land Board should not onlyallocate land to their relatives, but the community at large.’In neither case did the institutional representative chooseto respond.

Overall, at least one primary stakeholder in every one ofthe 43 kgotla meetings expressed dissatisfaction with atleast one of the following issues:

• Slow and/or no responses to applications (re. Land,licenses), complaints (re. Predator-related losses of live-stock), and inquiries (re. Borehole sitings).

• Corruption and lack of accountability at all levels ofgovernment.

• Imposition of regulations masquerading as ‘consulta-tion’ and ‘participation’.

• Ignoring valuable traditional methods of conservation(re. Fish, fauna, flora).

• Lack of respect for traditional leaders and traditionalknowledge systems.

• Lack of regulations and enforcement of regulations.• Delivery of poor or wrong advice, lack of feedback, lack

of reflexivity.• Generally poor communication skills and practices.• Biases toward modern/foreign applications/enterprises.

5.4.3. De facto governance practices

In the face of so many (real and perceived) policy, legaland institutional barriers to sustainable and equitableresource access and use, local people and governmentresort to a patrimonial form of governance typical of devel-oping countries. This translates as a combination of specialpleading on the part of the relatively unempowered andpromises and chastisements on the part of the empowered.For example, primary stakeholders often prefaced theirremarks, as did one speaker from Shakawe, with ‘We thepoor people’.

Government departments in turn made a variety ofpromises, but, at the same time, they often aggressivelyconfronted villagers. Rather than truly listen to what peo-ple had to say, government departments fell back on recit-ing rules and regulations – regarding, for example, landboard allocation procedures, government policy on theright of affected persons to kill problem animals and the

P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668 667

ambit of DWA activities in the Ramsar site. At several kgo-

tla meetings, government departments acknowledged thatlimited human and financial resource capacity contributedto their poor performance (in maintaining fences, in pro-cessing land applications, in compensating for animallosses) – an explanation to which rural people with limitedresources and capacities themselves could relate.

6. Conclusion

This paper has presented a comprehensive analysis ofthe issues that residents of Ngamiland deal with on aneveryday basis. It reveals what issues are important to peo-ple, the power dynamics between the government as theruler and the people as the ruled. In some cases the govern-ment is perceived as the answer to many resource use prob-lems while in some cases it is perceived as a direct threat tosustainable rural livelihoods.

For people of Ngamiland who have historicallydepended on the Okavango River ecosystem, access tothe waters are more than just about domestic water supply.In fact, water for drinking and cooking purposes is one ofthe least of their problems because it contributes minimallyto their livelihoods. Not that residents of Ngamiland wouldnot want or benefit from a reliable supply of potable water.To the contrary, many residents have expressed their desirefor such a service. But, as shown above, most of Ngami-land people’s livelihoods are dependent on fishing, flood-plain agriculture, transport, harvesting of reeds that growin rivers and lagoons, watering livestock, harvesting of veg-etables that grow in the river and tourism activities thatoften take place in the river such as mokoro (canoe) trips.Often these activities are done on a subsistence level butsometimes profit is made to put children through schools,build better housing and take care of the sick. Access tothe river or open water sources and other natural resourcessuch as veld products is of paramount importance. Gainingaccess to potable water while losing access to the river andits varied resources is not a desirable trade-off.

For the people of Ngamiland, therefore, water is part ofa system that requires a holistic management frameworksensitive to the above mentioned and related needs. Institu-tions that control the use and management of theseresources have however often failed to ensure access tothese resources. As shown above, this is in good part aresult of the policy and institutional framework of the insti-tutions responsible: each has its own limited mandate tocarry out in sectoral fashion; each department faces serioushuman and financial resource constraints; and all depart-ments respond to policy fashioned at central state level.Thus, even if there was the will to respond effectively tolocal issues, centralized structures of governance prohibitsuch initiative. This approach to policy making and imple-mentation is leading, it seems to us, to a gradual marginal-isation of the residents of Ngamiland from their resourcebase. The ODMP process has helped reveal the complexityof arriving at sustainable resource use policy where compe-

tition for key resources is both fierce and contested by fun-damentally different sorts of stakeholders. While there maybe no clear way forward at this time, the ODMP processprovides a framework for thinking more critically aboutthe consequences of current governance practices.

References

Bendsen, H., 2003. The Dynamics of Land Use Systems in Ngamiland:Changing Livelihood Options and Strategies. Harry OppenheimerOkavango Research Centre, Maun.

Bendsen, H., 2005a. Results and Analysis of the Feedback CommunityConsultation Process on the ODMP. Harry Oppenheimer OkavangoResearch Centre, Maun.

Bendsen, H., 2005b. Observations, Suggestions and Concerns fromCommunities Regarding ODMP Activities Grouped According toComponents. Harry Oppenheimer Okavango Research Centre, Maun.

Beukman, R., 2002. Access to water: some for all or all for some? Physicsand Chemistry of the Earth 27, 11–22, 721–722.

Central Statistics Office, 2001. National Statistical Tables Report: 2001Population and Housing Census, Gaborone, Government Printer.

Edge, W.A., Lekorwe, M.H., (Eds.), 1998. Botswana: Politics and Society,Pretoria van Schaik.

Goldblatt, M., Ndamba, J., van der Merwe, B., Gomes, F., Haasbroek,B., Arntzen, J., 2000. Water Demand Management: Towards devel-oping effective strategies for Southern Africa. IUCN, Gland.

Good, K., 1992. Interpreting the Exceptionality of Botswana. Journal ofModern African Studies 30, 1.

Kelly, K., 2001. From encounter to text: data-gathering in interpretiveresearch and Calling it a day: reaching conclusions in interpretiveresearch. In: Terre Blanche, M., Durrheim, K. (Eds.), Research InPractice. Applied Methods for the Social Sciences. UCT Press, CapeTown.

Mendelsohn, J., el Obeid, S., 2004. Okavango River: The Flow of aLifeline. Struik, Cape Town.

National Water Master Plan (NWMP), 1992. Government Printers,Gaborone.

North West District Council (NWDC), n.d. Ngamiland DevelopmentCommittee and Ministry of Local Government, Ngamiland DistrictDevelopment Plan 6: 2003–2009, Government Printer, Gaborone.

Okavango Delta Management Plan (ODMP), 2005. Okavango DeltaManagement Plan Final Inception Report 1 (Executive summary),ODMP Secretariat, Maun.

Swatuk, L.A., 2003. State interests and multilateral cooperation: Thinkingstrategically about achieving ‘Wise Use’ of the Okavango DeltaSystem. Physics and Chemistry of the Earth 28, 20–27, 897–906.

Swatuk, L.A., Rahm, D., 2004. Integrating policy, disintegrating practice:water resources management in Botswana. Physics and Chemistry ofthe Earth 29, 1357–1364.

Terre Blanche, M., Kelly, K., 2001. Interpretive methods. In: TerreBlanche, M., Durrheim, K. (Eds.), Research In Practice. AppliedMethods for the Social Sciences. UCT Press, Cape Town.

Van der Zaag, P., 2005. Integrated Water Resources Management:irrelevant buzzword or key concept? Physics and Chemistry of theEarth 30, 867–871.

Wirbalauer, C., Kgomotso, P.K., Magole, L.I., 2003. Proceedings ofStakeholder Consultation workshops and a National TBNRM Stake-holder meeting Botswana, October 2002–August 2003. IUCN,Gaborone.

Interviews

Mpateng Gojamang, Water Supply Section, Distribution Manager,Department of Water Affairs, Maun 19/09/05.

Koketso Baeti, Hydrology and Water Resources, Head of Section,Department of Water Affairs, Maun, 19/09/05.

668 P.K. Kgomotso, L.A. Swatuk / Physics and Chemistry of the Earth 31 (2006) 659–668

Johnson Reakae, Design and Construction, Head of Section, Departmentof Water Affairs, Maun, 19/09/2005.

Kurugundla Naidu, Aquatic Vegetation Control, Head of Section,Department of Water Affairs, Maun, 19/09/05.

Gabalaiwe Ramoshibidu, Groundwater Section, Groundwater Officer,Department of Water Affairs, Maun, 19/09/05.

Alec Makgasane, District Council–Council Water and Wastewater Unit,Principal Technical Officer, 20/09/05.

Sibangani Mosojane, Department of Wildlife and National Parks, DistrictWildlife Coordinator, Maun, 21/09/05.

Eliot Taylor, Okavango Development Management Plan, Chief TechnicalAdvisor, Maun, 20/05/05.

W.M. Muchina, Department of Animal Health and Production, Veter-inary Officer, Maun, 05/07/05.

Wellington Masamba, Water Chemist, Harry Oppenheimer OkavangoResearch Centre, Maun, 24/10/05 .