CASE STUDIES IN WETLAND t VALUATION #2: May 2003 t

n

n e

BAROTSE FLOODPLAIN, e

m

m e

ZAMBIA: e

g

g a

local economic a n

dependence on wetland n

a

a M

resources M

n

n

i

i s

Human use of the Zambezi s a

The various components of the Zambezi a B

River Basin system, and wetlands in B

particular, have a high economic value to r

r e

large numbers of people in Southern Africa. e v

Yet, in spite of their obvious importance, the v i

Zambezi’s wetlands have been vulnerable to i

R R

increasing pressures of economic and

population growth (Seyam et al 2001). o

o t

Resource over-exploitation, land drainage and t n

encroachment for agriculture, and n

i

i

interference with river hydrology for large- s

scale hydropower and irrigation schemes are s e

all resulting in wetland degradation. e u

u l

This case study describes an attempt to l a

articulate the economic value of one of the a

V

V

Zambezi’s largest wetland complexes, the c

Barotse Floodplain in western Zambia. It c i

formed a component of a project concerned i m

with wetland conservation, with field sites in m o

Malawi’s Lower Shire Wetlands, the Zambezi o n

Delta in Mozambique and the Eastern n o

Caprivi Wetlands in Namibia, as well as the o c

Barotse Floodplain itself. A particular focus c E

of the study, and of the project more E

generally, was to assess the value of local- d

level wetland resource use by wetland d n

communities. A major motivation for this n

a

a l

was that in the Zambezi Basin, the ecological l t

and economic value of wetlands to rural t e

communities is not fully appreciated when e W

river basin planning is undertaken or when W

land and water management decisions are g

made. As a result, such decisions often g

n

n i

interfere with wetlands of local economic i t

importance, thereby impacting heavily on the t

a

a r

communities who live beside them. r

g

g

e

e

t

t

n

n

I I

The Barotse Floodplain km2, increases steeply around the floodplain. The floodplain area is occupied mainly by the After rising in north-western Zambia and Lozi people, and falls under a dual passing southward through Angola, the administration − that of the Barotse Royal Zambezi re-enters Zambia in Western Province Establishment under the rule of the King, or and becomes larger and more consolidated, Litunga, and the Government of Zambia giving rise to a series of floodplains through Provincial and District line ministries (Timberlake 1997). These include the Barotse and administrative authorities. The use of Floodplain and other interconnected areas. The floodplain resources was in the past managed exact extent of the Barotse Floodplain is not according to traditional systems, under the easy to determine, because annually flooded customary authority of the Litunga. Today, areas grade into occasionally inundated parts, although formal control over natural resources and it is also difficult to separate the wetlands has been passed over to central and provincial influenced by the Zambezi from those fed by government, the Royal Establishment other catchments. Broad estimates put the maintains a great influence on natural resource Barotse Floodplain area at approximately use patterns and regulations in the region. 550,000 hectares, and the total wetland cover in the region at some 1.2 million hectares. The The Lozi people are also known as the “plains floodplain is mainly comprised of grasslands. or water people”, and local livelihoods and Although trees are largely absent from cultural traditions are linked closely with seasonally flooded areas, there are a number of seasonal flooding. During the dry season, the small wooded areas on higher ground (van Gils bulk of local production, economic activities 1998), and swamp forests are scattered over the and settlement are focused in the floodplain area. The Barotse Floodplain is flanked by area. As the plain becomes inundated, most of plateaux of Kalahari sand covered in semi- the population move to the uplands and plain evergreen woodland, interspersed with low- fringes. This annual relocation of people and lying dambos which are characterised by cattle includes the movement of the Litunga in grassland vegetation (Timberlake 1997). The a highly-celebrated traditional ceremony, called Liuwa Plain National Park and associated areas the Kuomboka (Nkhata and Kalumiana 1997). to the north west of the floodplain are relatively flat, and are waterlogged during the rainy season while remaining extremely dry Wetland resources in local during the rest of the year (Simwinji 1997). livelihoods Most of the population in the Barotse The ecological characteristics and conditions of Floodplain depend on a mixed livelihood the Barotse Floodplain, as well as the human strategy, combining crop farming, livestock production systems it supports, depend largely keeping, fishing and natural resource on the timing and duration of the annual floods exploitation. This diversity of livelihood (Timberlake 1997). The main wet season runs components, many of which depend on from November until March, although wetlands, is an effective strategy for spreading inundation depends mainly on rainfall in the risk, and income and subsistence sources vary upper catchment and on seepage from the at different times, especially according to uplands (Simwinji 1997). The onset of annual season. The rural economy is for the most part flooding varies greatly and may occur anywhere subsistence-based, and is subject to high levels between December and March, although of uncertainty and variability. About 76% of northern parts of the floodplain are generally the rural population in Western Province live in inundated earliest. The maximum flood level is poverty, and lean months are November to attained in April, after which floodwaters January/February when incomes are lowest gradually recede over May, June and July. and expenditures highest, and little food is available (Simwinji 1997). In total, the four Districts of the Barotse Floodplain are estimated to contain just under Almost all of the floodplain population are 225,000 people or 27,500 households. involved in crop farming. Of the total area Population density, which is generally low in under arable agriculture of 280,000 hectares in Western Province with fewer than 5 people per Western Province, about 10% is comprised of

floodplain farming systems. The main growing are concentrated. This activity intensifies from season in the floodplain is between November May until December, when fishermen stop and April, and produces maize, rice, sweet fishing in anticipation of the rains. When the potatoes, sugar cane, fruit and vegetables. floodplain becomes fully inundated, fish are Floodplain farming systems are diverse, and mainly caught using traditional maalelo traps, as include raised gardens (Lizulu), rain-fed village well as with traps and spears. gardens (Litongo), seepage gardens (wet Litongo), drained seepage gardens (Sishango), lagoon The floodplain population also makes use of a gardens (Sitapa) and riverbank gardens wide range of wetland plants, animals and (Litunda). natural resources for their daily subsistence and income. Almost all households harvest grass, Most of the cattle in Western Province are reeds and papyrus for use in house found along the Zambezi floodplain and construction, thatching, mat and basket adjoining plains (Jeanes and Baars 1991), and production, broom making and fishing the Barotse Floodplain is known to be one of apparatus construction. Clay is also important, the most productive cattle areas in the country used for house construction and pottery (Simwinji 1997). Over three quarters of cattle making. Although the loss of many of the in Western Province are pastured in the floodplain’s wild mammal populations has floodplain, including 265,000 head that belong meant that hunting has decreased over time, to floodplain residents. The bulk of herds are turtles, birds and birds’ eggs form an important managed under a system of transhumance and supplement to local diets. move between the floodplain and adjacent uplands, usually spending January to July in the Valuing household wetland use and floodplain and the remainder of the year in the uplands. Primarily driven by the seasonal future management scenarios availability of pasture, annual transhumance is The primary aim of the valuation exercise was also important for the distribution of manure, to assess and articulate the value of wetland and in the floodplain there is a strong goods for local communities. It was carried out interaction between herding, cropping and in two phases. During the first phase, scoping fishing activities (Simwinji1 997). visits were made to the study area in order to assess the status of existing data, meet with The fisheries sector is one of the most government and traditional authorities, and important sectors in Western Province, and is consult with local villagers. This yielded mainly concentrated on the floodplains of the information about which resource were used, upper Zambezi (Timberlake 1997), especially their relative importance and value, and helped the Barotse floodplain (Simwinji 1997). Just in the design of survey instruments and over half of the floodplain population are valuation methods for the subsequent main involved in fishing activities. Fish are an study. During the second phase of the study, important source of protein, and local fish household surveys were used to obtain consumption is five times the national average quantitative data about the use of wetland (van Gils 1998). Bream make up 80% of the resources. Focus group discussions with catch (Maimbo et al 1996), and a number of community leaders, resource users, wetland smaller fish are also caught such as minnows, specialists, and different socio-economic tilapia, bottlenose and silver barbels. Fishing is categories provided a more detailed, and a highly seasonal activity. Between December participatory, means of assessing the economic and April, fish move from the main river importance of wetland resources to the local channels into the wetlands, where they spawn population. before the height of the flood (Nkhata and Kalumiana 1997). As the floodwaters rise, a The data gathered were analysed using a static phenomenon called “red waters” occurs, where economic model to determine the value of each low oxygen water is pushed forward by the wetland resource. The model modified and floods. Only barbel can survive in this water. extended an existing approach, which had The main fishing season takes place as the originally been developed in order to assess floodwaters recede, and gill nets are used in the local and national level returns to wildlife lagoons which have formed and in which fish resources in Namibia (Ashley and Barnes 1996,

Barnes and de Jager 1995). It indicated the applied to the model. These included various financial and economic returns to different combinations of a “do nothing” scenario of wetland utilisation and value-added activities, continuing resource use and human population measured as private net cash income and growth; a “wise use” scenario based on economic net value added to national income. sustainable levels of wetland resource Values were expressed at the level of the whole utilisation; a “protected area” scenario where floodplain, and as gross and net returns per parts of the floodplain were put under strict household. protection which required resource utilisation activities to be reduced or curtailed completely; A dynamic model was then developed to and an “agricultural development” scenario calculate the present net value of wetland assumed the gradual transformation of resources under different future management floodplain wetlands to large-scale irrigated rice. scenarios. This scenario analysis used a An additional scenario, “upstream hydrological dynamic ecological-economic model which developments” was identified and described simulated a simple wetland system and the but not modelled quantitatively, as it depended effects of human activity on that system. A on actions being taken outside the direct study generic Zambezi wetland model was area and because there were at the time no developed, and then adapted to model the such plans for developments upstream of the Barotse area. The model was run from 30 years Barotse Floodplain. before the time of the study to 20 years hence, in order to simulate past resource trends Integrating livelihood values into recorded in the study and investigate how these trends would affect future wetland values. Four wetland planning future wetland management scenarios were The study confirmed the extremely high value identified, each based on likely or planned of wetland resource use in local livelihoods, actions in the Barotse Floodplain region, and and as a way of spreading seasonal risk and

Table 1: Financial and economic returns to wetland resource utilisation Reeds & Cattle Crops Fish Wildlife Palms Grass Clay Total papyrus PER HOUSEHOLD ($/year) Gross financial 120.4 90.8 179.6 5.83 15.12 0.43 8.25 2.39 417 value Net financial 120.4 88.7 174.1 0.41 10.72 0.27 8.07 2.33 405 value Gross cash 11.5 6.1 52.6 0.01 1.61 0.04 0.30 0.02 72 income Gross subsistence 109.0 84.8 127.0 0.42 13.51 0.29 7.95 2.37 345 value TOTAL WETLAND ($’000/year) Gross economic 3,988 1,447 5,947 12 501 12 272 66 12,244 value Net economic 3,908 -75 4,258 10 271 3 221 52 8,647 value Gross financial 3,323 2,507 4,956 12 417 12 228 66 11,520 value Net financial 3,323 2,447 4,803 11 296 7 223 64 11,174 value Gross cash 316 167 1,452 0.3 44 1 8 0.5 1,989 income Gross subsistence 3,007 2,340 3,504 12 373 11 219 65 9,531 value

uncertainty. In total, local use of wetland cost of lost access to wetland resources of resources in the Barotse Floodplain was found between $1.2 -$3.0 million a year. to have a net economic value of some $8.64 million a year (Table 1). At the household level, Developments outside the Barotse Floodplain wetlands were calculated to generate an average area itself were also found to have a profound net financial return of $405 a year. The major effect both on wetland status and on the local proportion, 83%, of this value was comprised economy and livelihoods. A large number of of subsistence values and home consumption, new hydropower schemes, dams and reservoirs and by far the most valuable products were have been identified for development along the found to be fish (43% of total, and 73% of Zambezi River although none are currently household cash income), floodplain grazing planned in Angola or north western Zambia, (29% of total) and crop production (22% of upstream of the Barotse Floodplain. The total). findings of the study emphasised that any upstream development, if it influenced Dynamic modelling indicated that the most downstream river flow and flooding, would be economically valuable future management likely to incur devastating economic losses to option was wise use of the wetland area, the local communities who depend on the possibly combined with small areas under strict Barotse wetlands. protection − in comparison to a “do nothing” scenario, to putting larger areas under strict An important conclusion of the study was the protection, or to converting the wetland for the emphasis it accorded to the economic role of implementation of large-scale agricultural the Barotse Floodplain in local livelihoods. Not development projects (Table 2). The economic only is the value of the wetland great in and financial values yielded by wetland wise use absolute terms, but it also forms a key and conservation were found to be most component of local livelihoods − without pronounced at the local level. As the area access to wetland resources, households would under strict protection is increased, large both lose the values accruing from natural financial and economic losses are incurred by resource use and also be separated from an local communities who must reduce or curtail essential source of support and inputs to other their utilisation of wetland resources. The forms of production such as crop farming and financial benefits of agricultural schemes to livestock keeping. For a long time such values local communities, primarily felt through have not been a factor in decision-making in increased employment opportunities, are also the Zambezi Basin: because they are not likely to be far outweighed by the opportunity reflected in official prices and formal markets, they are treated as being negligible. The study Table 2: Model of economic values associated with made the point that this is clearly not the case. future wetland management options Ecological-economic modelling of future Financial Economic wetland options showed that any management Scenario NPV ($ mill) NPV($ mill) scenario which omitted consideration of these Do nothing 84.2 62.2 values, and did not allow for the local-level Wise use 86.7 64.4 use of wetland resources, would run the risk Wise use and 10% of being both economically and financially 85.5 64.5 protection sub-optimal, as well as jeopardising the Wise use and 25% economic livelihoods of almost a quarter of a 82.7 63.7 protection million people. Wise use and 50% 74.0 58.4 protection This case study is adapted from Turpie, J., Smith, B., Agricultural development Emerton, L. and J. Barnes, 1999, Economic Valuation of 5,000 ha and no 85.4 63.2 the Zambezi Basin Wetlands, IUCN − The World protection Conservation Union Regional Office for Southern Africa, Agricultural development Harare 5,000 ha and 25% 81.4 62.5 protection Agricultural development 20,000 ha and 25% 79.7 60.9 protection

Ashley, C. and J. Barnes, 1996, Wildlife Use for Nkhata, D. and O. Kalumiana,1997, Energy Economic Gain: the Potential fo Wildlife Needs and Shortfall Assessment of the to Contribute to Development in Barotse Flood Plain of Western Nambibia, DEA Research Discussion Province, IUCN − The World Paper No 12, Directorate of Conservation Union Regional Office for Environmental Affairs, Windhoek Southern Africa, Harare Jeanes, K. and R. Baars, 1991, Carrying Seyam, I., Hoekstra, A., Ngabirano, G. and H. Capacity of Western Province. A Popular Savenije, 2001, ‘ The value of freshwater Version of “The Vegetation, Ecology and wetlands in the Zambezi Basin’, paper Rangeland Resources of Western presented at AWRA/ILWRI-University Province, Zambia”. RDP Livestock of Dundee International Specialty Services Report to Department of Conference on Globalisation and Water Agriculture, Lusaka Resources Management: the Changing Maimbo, F., Huijsman, A., Mulwanda, D. and Value of Water. B. Lof, 1996, Opportunities for Western Simwinji, N., 1997, Summary of Existing Province: an Agro-Economic Relevant Socio-Economic and Ecological Reconnaissance Study. Royal Tropical Information on Zambia’s Western Institute, Amsterdam. Province and Barotseland, IUCN − The World Conservation Union Regional Office for Southern Africa, Harare Timberlake, L., 1997, Biodiversity of the Zambezi Basin Wetlands: a Review of Available Information, Zambezi Society and Biodiversity Foundation for Africa This document was produced under the project Report to IUCN − The World Conservation "Integrating Wetland Economic Values into River Basin Union Regional Office for Southern Africa, Management", carried out with financial support from Harare DFID, the UK Department for International Development, as part of the Water and Nature Initiative of IUCN - The van Gils, H., 1998, Environmental Profile of World Conservation Union. Western Province, Zambia, ITC Report to This project aims to develop, apply and demonstrate Provincial Planning Unit, Mongu environmental economics techniques and measures for wetland, water resources and river basin management which will contribute to a more equitable, efficient and sustainable distribution of their economic benefits at the global level and in Africa, Asia and Latin America, especially for poorer and more vulnerable groups.

The views and opinions in this document are those of the authors alone, and do not necessarily reflect those of IUCN, DFID or other institutions participating in the project. For more information, please contact: Lucy Emerton. [email protected] Tel: ++94 1 694 094

For information about project activities in Africa, Asia and Latin America please contact:

Eastern Africa: Francis Karanja. [email protected] Tel: ++254 2 890 605-12 Latin America: Rocío Córdoba. [email protected] Tel: ++506 241 0101 Lower Mekong: Sarah Porter. [email protected] Tel: ++855 23 222 311/2 South Asia: Shamen Vidanage. [email protected] Tel: ++94 1 694 094