The Lower Usutu Smallholder Irrigation Project EIA

LUSIP Background | The LUSIP EIA | The Project Brief |

Background

According to Environmental Audit, Assessment and Review Regulations of 1996, an Environmental Impact Assessment (EIA) is required for the Lower Usuthu Smallholder Irrigation Project (LUSIP) as it is classified as a Category 3 project. The EIA covers the proposed project as designed in the Pre-Feasibility Study (Booker Tate, 1996) and the Feasibility Study (GFA-Agrar, 1998). The financing of the EIA has been provided by the European Union under the 8th European Development Fund. The contract for the EIA was awarded by the NAO to the consulting engineering Company Vakakis International S.A. in association with WS Atkins International on 4th August, 1999. The contractor mobilised on 15th August, and started work in Swaziland on 16th August, 1999.

A scoping report was delivered on 20th September, 1999, and represented the output of the first stage of the study. It contained a review of the key information and data collected during the scoping period, either from reports and documents or from contacts with a large number of identified stakeholders. It also proposed a screening of environmental issues and impacts to be included for detailed analysis, and some updating of approach, organisation and inputs for the EIA (Environmental Impact Assessment) / CMP (Comprehensive Mitigation Plan) work. The First Workshop was held in Big Bend on 21st September, 1999, and the first Steering Committee meeting was held on 23rd September, 1999. On 28th September, 1999, SEA issued a letter addressed to the Ministry of Agriculture and Cooperatives (MOAC) with ref MTEC/PRJ/5 giving clearance to start the EIA/CMP preparation subject to amendments, recommendations and comments as detailed in the letter. A meeting was held at the SEA to discuss issues raised in this letter. The meeting was attended by three key members of the Consultant’s team.

The EIA field work started on 4th October, 1999 and ended on 29th November, 1999. The Second Workshop was held in Big Bend on 1st and 2nd December, 1999. The first day was held in Siswati and was aimed particularly at local people residing in, or near to, the proposed Project Area. The second day was held in English and aimed particularly at Ministry staff, policy makers, potential investors, and representatives of funding agencies. From field work and the Second Workshop, an Interim Report was prepared and submitted to the members of the Steering Committee, the last meeting of which was held on 9th December, 1999. According to the contractual documents, the contractor then continued with the 5th stage of the EIA – the Economic Valuation - and the 6th stage – the preparation of the Draft First Final Report, which was completed during the last week of January 2000. A third workshop was organised in Mbabane on 27^th March 2000 and the Final Report was completed end of April 2000.

Executive Summary

The Proposed Project

The main objective of the project is poverty alleviation to allow smallholder farmers to shift from risky rainfed agriculture and low-productivity extensive cattle grazing, to high-value irrigated crops, notably sugar cane and cotton. This follows the National Indicative Programme, which allocates 70% of development funds to agriculture and rural development. The NIP aims to bridge the gap between small holder farming and commercial enterprises, particularly to diversify and improve productivity, rural incomes and nutrition, as well as to increase agricultural exports.

The project will embrace a 10-year, 2-phase programme, finally providing smallholder irrigated cropping on 11,600ha. The main infrastructure includes a low weir on the Usuthu at Bulungapoort, 23km of feeder canal leading to Bovane reservoir which will have 155m m³ storage capacity. Two main canals serve the irrigation blocks. The northern canal commands 2000 ha (including 1300ha by 30m lift pumping) ; the southern canal commands a further 4500ha in the Bovane Block and 5100ha in the Matata Block, some of which would involve pumping. The southern canal would run for a total of 75km.

Crops to be grown would include sugarcane, cotton, vegetables, beans, corn, bananas, and other fruit. Half of the irrigation planned is sprinkler (particularly on the more marginal soils), the other half surface irrigation. (Surface irrigation is considerably cheaper in both capital and running costs but is less efficient in water use, with only 50-60% efficiency being common, as compared to 70-80% efficiencies with sprinklers.) Drainage is planned on 10% of the surface-irrigated area. Sugarcane would be transported to the Ubombo Mill (a maximum distance of 35km – average distance 15km); cotton would be processed at Matata. Both facilities would be expanded by the private entities involved.

Each smallholder family would be allocated an irrigated land holding of 3.5ha, the size optimised for family labour to be fully used. Smallholders would be organised into Water User Groups (WUGs) and Farmers Associations (FAs), each covering 55-100 ha of irrigated land. WUGs and FAs would be responsible for organising irrigation at farmer level, collection of revenues, and organisation of some maintenance activities.

The Existing Environment

The area lies between elevations of 100m in the east and 340m in the south-west, sandwiched between the Lebombo Mountains and the Sinceni Mountains, respectively rising to 750 and 940m. Eastern areas are thus in a considerable rain-shadow.

Mean annual rainfall varies across the project area from 520mm in the east to about 750mm in the west, with a long but very unreliable rainy season extending from mid-October to end-April. Rainfall variability is very great from year to year. Winters are cool and generally dry – slight frosts occur very occasionally. Wind damage, and hail damage, associated with thunderstorm activity occasionally occurs. Cyclones may occur every 25-30 years, with damage from flooding being much more severe than from wind. However, in general, the climate of the area can be considered as being highly suitable for irrigated sugarcane, leading to excellent yields on the best soils, high sucrose contents, and facilitating a long sugar milling season (8 months).

Soils, however, are more problematic. Of the total mapped project area of 26,300ha, soils with little restrictions for irrigated agriculture make up only 21.5% of this area; soils which may have water table and sodicity build up after irrigation occupy 13.9%; excessively drained sandy soils near to rivers 6.1%; very marginal soils with depth limitations make up 36.6%; and totally unsuitable soils 21.9%. However, of the gross irrigation area of 13,740ha, the respective percentages are 40.4%, 25%, 10.2%, 24.3% and 0%. (The dominant soil groups –or ‘Sets’ – represented by these divisions are: R, K/C, B/W, S and U/H/Z respectively.) Of greatest concern are the shallow soils – the so-called S-sets- whose depth is very variable over short distances. They are not suitable for furrow irrigation under any crop: with sprinkler irrigation, they are suitable only for well-managed sugarcane. Where they are below 30cm in depth, they are considered not suitable for any crop.

Most of the area overlies hard geology, from basalt sheets in the east to dolerite dykes in the west. Hard rock material occurs at shallow depth (generally 50-150cm) which means that deep excavations would entail a great deal of expensive blasting. Sandstone areas occur to the southwest of the project area, and soils occuring on this material are extremely susceptible to erosion.

Extensive studies have been undertaken on the hydrology. The Usuthu River at the proposed weir site has a catchment area of 12,600 km², of which 60% is in South Africa, although this only contributes 40% of the flows, due to diversions by four dams and large areas under forest plantations. Current annual flows average 1450 million cubic metres (46m³/sec), and the gross annual irrigation requirement is 197m m³, representing only 12% of the total flow. The tributary Mhlatuzane river, draining a catchment area of 540km², has an annual flow of 39m m³, equivalent to a mean flow of 1.2m³/sec. Dry season water allocations for irrigation are excessive, and in some years the Usuthu dries up completely in the section just above Matata.

Current sediment load in the rivers reflects major soil erosion problems in the catchment areas. For the critical Mhlatuzane catchment, net sediment yields are estimated at some 15 tonnes/ha/year, sufficient to reduce the storage capacity of the proposed reservoir by 10-15% within 25 years. Erosion control, and community participatory approaches to land use planning and soil conservation are thus of critical importance.

Other water quality parameters of concern include phosphate and nitrate, which lead to problems of eutrophication, particularly during low flow periods. Although concentrations of anions and cations increase by a factor of 4 across the project area, irrigation water quality remains generally very good. However, close monitoring of all key water quality parameters are necessary.

Vegetation in the project area is dry acacia bushland of various types which are closely related to soils and geomorpholgy. About half of the area would appear to have been cleared in the past for cultivation, with about one third of this used in any one year. For the remaining areas of relatively undisturbed vegetation (about 8,600ha), five main vegetation associations were recognised associated with the following soils : (1) shallow soils (U-set); (2) erodible duplex soils over sandstone (H and Z sets); (3) red upland soils of various depths, derived from basalt or dolerite (S and R-sets); (4) black clay soils in footslope positions (K and C-sets); (5) vegetation associated with drainage network, on alluvium and other soil types. Type (5) is of by far the greatest value in terms of species-richness and total dry matter productivity. Type (4) has particular value for associated rangeland species, together with the deeper soil areas of Type (3). Total range productivity (graze and browse value) was estimated to be in the order of 1.3 tonnes dry matter / ha / year, higher than average for the Lowveldt. Total standing timber / bushwood is estimated to be in the order of 16-24 m³/ha in the Bovane Block (wetter western area) to 5-8 m³/ha in the drier Matata Block, with annual wood increments of 0.5-0.73 and 0.16-0.23 m³/ha respectively, although all of these figures vary enormously depending on rainfall and soil and water table conditions.

There are some 28 rare and endangered species of plants in the project area and immediately adjacent areas, and a further 9 plants which, because of high economic value, are likely to attain that status in the near future.

For the Fauna, a total of 23 orders of aquatic macro-invertebrates and 34 species of fish were recorded from the Lower Usuthu and Mhlatuzane rivers within the Lower Usuthu Basin. Amongst the vertebrates, 2 species of fish, 3 species of amphibians, 14 species of reptile, 10 species of mammals and 13 species of birds are of special conservation status both within Swaziland and in the Lowveldt. Species diversity and population sizes of birds and mammals in most areas of Swazi Nation Land are generally very low. Areas which were relatively undisturbed included the Sinceni Mountains, Maphilingo (proposed feeder canal area), Bovane Ridge, the southern and south-eastern areas of the proposed Bovane Dam, and TDL areas to the south-east and south-west of St Phillips. Fragmentation of these areas by existing unplanned development is an increasing threat: these areas need to be linked by ‘green corridors’ so that wildlife can migrate across the area.

Socio-Economic studies covered administrative framework, demography, homestead and household characteristics, economic activities, land, livestock, infrastructure and social services, public health and cultural heritage. The area is under eight chieftaincies, but only five would benefit significantly from the irrigation project: Ngcamphalala, Mngometulu, Gameze, Matsenjwa, and Shongwe, with total irrigated smallholder settlement capacity, based on a 3.5ha model, of 1532, 570, 441, 410, and 97 respectively. Some 132 homesteads (4% of the total) are in areas apparently disputed between adjacent chieftaincies. The domain of Maja Mamba to the south controls much of the key catchment area of the Bovane reservoir, so this chieftaincy is important to be included in the project. Land of the late Shwapha Dlamini and the late Sonjane Dlamini occurs in the Weir area .

Within the tentatively proposed gross areas for the irrigation blocks, the number of existing homesteads is 1160, with an estimated population of 9860 based on a de-facto household size of 8.5. 1998 census data shows 13,300 people, the difference being explained by settlements on adjoining land, not included in the irrigation block areas. The population is very young (overall average recorded age of 21.3 years), women slightly outnumber men, and the population is increasing rapidly. Education attainment is low, 30% having had no education, and a further 37% with only primary education. Homesteads are in both dispersed and clustered patterns, with a large majority concentrated on the less-favourable soils.

Household and per capita income shows an enormous variation, due to some households having had appreciable savings or family members with relatively well-paid employment elsewhere. Employment breakdown, for working age people, were 32% wage employed, 39% self employed (mainly farmers) and 29% not employed / unemployed.

A large majority of households (91%) have access to land for cultivation, with an average holding size of 8.8 ha. Rainfed cotton for cash sales, and maize for home consumption, were the main crops. Of the households who owned cattle, the average number of cattle owned is 17.6, and cattle held, 19.5. A further 17.3 goats, 11.4 sheep and 5.3 pigs were held.

The Major Impacts

The project is expected to have very major impacts on the surrounding area and on the country as a whole. It would represent a key focus of Swazi activities for the Lubombo Spatial development Initiative (LSDI) which also embraces Mozambique and South Africa, both of which are located a few km to the East of the project area. Major positive impacts are expected to be on income for some 3200 farm families (2200 of them from surrounding areas), and probably an equal number of families brought in to work in secondary and tertiary activities: some 50,000 people in total would thus derive their income directly or indirectly from the project. Land productivity would be expected to increase by a factor of up to 25: from a calculated total annual production from bush rangeland of E310/ha/year to a net revenue for a family-run sugarcane holding of E8000/ha/year. This rangeland is already in a fairly degraded state. If the project were not to go ahead, range productivity would certainly decline due to increasing soil erosion and loss of fertility. An average loss of some E80/ha/year would be expected over a 25year period.

Other positive impacts are :

• a much greater security of income: even in the worst drought years a high proportion of average income would be secured;
• concentration of population into planned settlements, where infrastructure and social facilities can more easily be provided;
• generation of a considerable amount of work during the 10-year construction period: Swazi contractors would gain valuable on-the-job training and experience working to a high standard of specifications;
• a multiplier affect on the local economy, with development of entrepreneurial activities;
• further development of a cash, rather than subsistence economy;
• downstream benefits in the form of increased return flows, which may be as high as 2.9m³/sec – this would prevent certain parts of the lower Usuthu from drying out, which currently happens periodically in the dry season due to over-allocation of low-season water.

The project, however, is likely to have certain major negative impacts:

• Very unequal allocation of resources within the Lowvedlt;
• Unequal benefits and negative impacts between adjacent chieftaincies: one chieftaincy will receive half of the benefits;
• Initial displacement of some 12,900 cattle from project lands, these lands representing some of the best grazing in the Swazi Lowveldt;
• Likely influx of cattle on to adjacent and highly erodible lands;
• Disruption of seasonal cattle migration patterns – loss of access to perennial rivers in the dry season, together with grazing on rainfed crop residues;
• Influx of relatively youthful, single males for construction transforms the population profile/equilibrium and behaviour patterns;
• Required resettlement of some 88 households (the ‘involuntary’ resettlers – see definitions in Glossary), and possible resettlement for a further 232 ‘recommended resettlers’;
• Major disruption and dislocation of life for host resettlement communities, although all should greatly benefit in the long-term;
• Loss of traditional food products, herbal medicines, firewood;
• Loss of natural vegetation - over approximately 8,600 ha :850ha of valuable riverine vegetation;
• Increased threat to 28 potentially threatened species of flora protected by legislation;
• Further fragmentation of Swazi lowveldt habitats;
• Possible rapid and uncontrolled growth of informal settlements.

Major Areas for Concern

Mitigation and compensation measures have been designed to counteract the above negative impacts. However, there are some features of the project as designed in the feasibility study, or are inherent of the project or study area, that represent a significant risk factor to successful performance of the project. Although some of these features are not strictly environmental in the narrowest sense, they could have major negative environmental impacts in the long term if the project were less successful than currently projected. They include the following :

• Sugar prices on international markets. These are likely to fall. A moderate fall has been built into the project economics;
• Management problems of shallow S-set soils (also K and C-set soils). These require good management for land preparation, irrigation techniques and agricultural practises. The range of cropping is very restricted. For S-sets only sprinkler-irrigated sugarcane is suitable;
• Sedimentation, particularly in the Mhlatuzane Reservoir;
• Further water allocations up-stream, both in Swaziland and South Africa;
• Initial displacement of cattle from much of the project area, and subsequent integration of some cattle, into an intensively managed mixed farm system;
• Resettlement issues, particularly the relationship of chieftaincies to the project and the issue of equity;
• Requirement for a rapid and radical change in the way of life on the part of the small-holders.

Another major area of concern lies in engineering infrastructure, and project safely and performance, the main problems being:

• Flood peaks on both Usuthu and Mhlatuzane Rivers are likely to be greater than those predicted in the feasibility study: redesign of dam height and spillway characteristics, and redesign of the weir are both necessary to accommodate this;
• Canals as designed are narrow, deep, with steep gradient, and hence fast-flowing. Command area (some 1500 ha) is hence lost; canals are expensive to construct (because more blasting of the hard bedrock would be required), and also dangerous from drowning of children and animals.

Thus the first component of the mitigation measures includes a redesign of these key components. It is likely that increase in costs for the dam would be more than compensated by savings on the canal designs.

Mitigation and Compensation Measures

The mitigation measures aim to address the identified negative environmental impacts of project activities and have been grouped under the following components:

• CMP Stage 1: Final Design Phase: Components 1 to 4
• CMP Stage 2: Preparatory Phase: Components 5 to 7
• CMP Stage 3: Implementation Phase: Components 8 to 12

Component 1. Project design improvements

This component groups together the Engineering Modifications which need to be incorporated into the project during the final design phase. These modifications are considered necessary for a number of reasons: (i) to improve project safety; (ii) to improve operational efficiency and project economics; (iii) to reduce any negative ecological impacts; (iv) to overcome specific and important objections by some of the Interested and Affected Parties (IAPs). The subcomponents would be:

Component 2. information, communication and institutional building

This component seeks to undertake the essential groundwork to continue good working relations between Interested and Affected Parties (IAPs) and the LUSIP Implementation Agency (LUSIP-PIA), some 90% of IAPs having stated that they are keen to join the project as smallholder irrigation farmers. Good groundwork at this stage is considered necessary for a number of reasons: (i) to ensure that IAPs are fully informed of final project designs and exactly how they might be affected; (ii) to fine-tune some aspects of final designs (e.g. canal location cf. villages); (iii) to establish local committees who will integrate new settlers with existing households; (iv) to allocate land efficiently and equitably, and establish the base for Farmers Associations/Water Users Groups: (v) to overcome specific and important objections by some of the Interested and Affected Parties (IAPs). The subcomponents would be:

Component 3. Specifications for the Contractor and Works Supervision Contracts

This component regroups additional contractual clauses for Contractor and Works Supervision Contracts to be introduced in the Tender Documents. Those clauses aim at mitigating the risks for health and safety of the IAPs, some ecological impacts on fauna and vegetation and to ensure that any archaeological resources, if found, are identified and if necessary rescued. The subcomponents would be:

Component 4: Mitigation and Compensation Measures: Update and Final Design

This component aims to update the CMP on the basis of refinements and changes made in the detailed designs over those given in the feasibility study. The subcomponents would be:

Component 5. Institutional Development and Capacity Building

This component seeks to continue to aim for good working relations between IAPs and the LUSIP-PIA.

It aims to strengthen the grass-level institutions already established in Component 2, and rectify any problems that become apparent. The eventual aim is: (i) to ensure that IAPs are fully informed of final project designs and exactly how they might be affected; (ii) to fine-tune some aspects of final designs (e.g. canal location cf. villages); (iii) to establish local committees who will integrate new settlers with existing households; (iv) to allocate land efficiently and equitably, and establish the base for Farmers Associations/Water Users Groups; (v) to overcome specific and important objections by some of the Interested and Affected Parties (IAPs). The subcomponents would be:

Component 6. Negotiation and agreements with IAPs (through CRPCs) for land use and land management

The long-term success of the project will depend heavily on the degree of success of intensified use of natural resources, with the prospect that the rural population density may increase by a factor of about 6 within the immediate project area. Not only with increased numbers of irrigation farmers have to be accommodated (increasing the rural population in the command areas from 1160 to some 3200 families), but for every on-farm job there is likely to be at least one off-farm job, most of these being in the service sector. The eventual aim is: (i) to control the increasing pressure on the range resources within the project and in neighbouring areas; (ii) to preserve the long-term storage capacity of the reservoir by soil conservation measures in the immediate catchment area. The subcomponents would be:

Component 7. Compensation and Resettlement

Compensation and resettlement aim at  inter alia: (i) ensuring that resettlement activities are planned and implemented with the full co-operation and participation of those being resettled; (ii) paying full, fair and prompt compensation, and provide other appropriate assistance to any person whose occupation of or interest in land is ended or interfered with to their detriment, or otherwise suffers loss by reason of any scheme of resettlement; and (iii) encouraging the development of fair tenure arrangements contributing to greater security of tenure on SNL and any other land where people are occupying land under Swazi customary rights. The subcomponents would be:

Component 8. Institutional Development and Capacity Building

This component seeks to continue to aim for good working relations between IAPs and the LUSIP-PIA. It aims to strengthen the grass-level institutions already established in Components 2 and 5, and rectify any problems that become apparent. The particular aims of this component is to: (i) to ensure that IAPs are fully informed of final project designs and developments; (ii) to fine-tune some aspects of final designs (e.g. canal location cf. villages); (iii) to strengthen the CRPCs in their dealings with land allocation, and particularly the technical issues. The subcomponents would be:

Component 9. Investments in Social Development

The mitigation measure M602 "Establishment of community based local development plans for service provision will allow to identify, design and localize the additional or relocated social facilities as a consequence of the Project implementation. The subcomponents would be:

Component 10. Investments in Conservation of Natural Resources

There is a need for a long term positive symbiosis between the project area and adjacent areas under forestry, bush range, and low-intensity rainfed agriculture. Conservation measures have to be funded by profits made by project activities on the long term, and this is the fundamental philosophy behind this component. A small levy will thus be charged on production, and the fund so established would go towards activities in the following four areas (M1001 to 1004 respectively.) It is also proposed that a small levy be made on any commercial tourism activities benefiting from these activities. In return, the management of this Project Environmental Fund should be transparent, with key contributors to the fund, or representatives of these contributors, being represented on the committee responsible for spending these funds. Any private land owners, whose land is involved in these activities, would automatically be included on this committee. The subcomponents would be:

Component 11. Implementation of the Resettlement Plan

Implementation of the compensation and resettlement as elaborated in Component 7 (M1101).

Component 12. Investments in Ubombo Sugar mill

Ubombo Sugar, a private company with a majority shareholding held by Illovo, stands to gain considerably from LUSIP in that they would be the sole miller for the sugarcane produced. The mill itself is a considerable source of pollution, both air pollution from inefficient mill boilers, and water pollution from currently inefficient water treatment plant. Both areas would have to be improved and extended together with mill size.

Organisation and management

The CMP will be implemented in three different stages: (i) Stage 1: Final design: occurring simultaneously with the Project Final Design stage. This would be carried out in Years 1 and 2; (ii) Stage 2: Preparatory phase: corresponding to the initial stages of construction, and the end of Year 2 and beginning of Year 3; (iii) Stage 3: Implementation phase: simultaneously with Project Construction Works and Land Development, starting in Year 3.

Stage 1 : Final Design

Stage 1 of the CMP should be contracted to Consultants in two separate lots as follows: (i) Lot 1: Component 1and 3; (ii) Lot 2: Component 2 and 4. Lot 1 would form part of the ToR for the Consultant Engineering Firm contracted for the Project Final Design. Lot 2 would be contracted to Environmental Consultants, not necessarily within the same group as those responsible for Lot 1.

The management and supervision of the Consultant Contracts is the responsibility of the Project Proponent. Final documents produced by the Consultants shall be endorsed by SEA on the sole criterion of conformity with the present CMP.

Stage 2 : Preparatory phase and Stage 3 : Implementation phase

At the end of this study, no consensus had been reached on the desired institutional framework for the project, and the Consultant assumes that in due course this level of institutional support will be clarified. In the meantime, such an agency is referred to as the "LUSIP-Implementation Agency – LUSIP-IA".

Within this LUSIP-IA, it is proposed to create an Environmental Unit in charge of ensuring a cost-effective implementation of the CMP. This Unit would be divided into two sub-units, one sub-unit in charge of Land Use Planning and Resettlement (LUPR), the other sub-unit in charge of the Institutional Development, Capacity Building and Public Awareness (IDCB) . Both sub-units would work in close co-operation with each other, with some activities requiring particularly close collaboration.

The key to the success of the Lower Usuthu Smallholder Irrigation Project (LUSIP) is the design of an appropriate and acceptable institutional structure for all stakeholders so that confidence in the methods of communication and delivery of all project related activities is established as early as possible in the project cycle. Therefore, it is proposed to create Community Resources Planning Committees (CRPCs) coordinated by the Chiefs and regrouped into the LUSIP Development and Planning Committee.

LUPR and IDCB will be responsible for the implementation of the set of mitigation measures during stage 2 and 3 (components 5 to 11) and Ubombo Sugar will implement Component 12.

CMP Monitoring will be undertaken by a separated Environmental Monitoring Unit , reporting directly to the Project Board of Directors, in which SEA should be represented.

Key indicators for Monitoring

Monitoring would be undertaken on a number of key indicators by the above unit. The main subjects to be covered, and a selection of the most significant components under each subject are:

• Soils: S-set soils; erosion and infiltration rates ; water tables/sodicity in K/C-set soils;
• Resettlement: quantitative baseline information; real income of weakest in community;
• Socio-economic: population; growth of informal settlements;
• Health: malaria; HIV/AIDS/TB;
• Fauna: macro-invertebrates in rivers (as indicator of water quality); snakebites; water birds; fish;
• Vegetation: 28 potentially threatened species; medicinal plants;
• Water: water flows at GS6,16,19; water quality monitoring; baseline pesticide study;
• Rainfall-runoff in the Mhlatuzane catchment;
• Air: around sugar mill; selected areas within rural sugarcane areas.

Publication from monitoring activities will be undertaken on a regular basis. Should anomalous results be obtained (e.g. alarmingly high levels of phosphate in the water samples) determinations would be repeated immediately, and if these are confirmed, information would be passed to IAPs. For water parameters, circulation of results to TPTC members would be undertaken as soon as possible, so transparency of information is maintained. Standardisation of laboratory and sampling procedures needs to be undertaken at the beginning of the monitoring programme, and for water parameters, these need to be recorded across the whole river basin, according to the same methodology.

Cost of the Comprehensive Mitigation Plan

The total cost of the CMP may be summarised as follows:

Comprehensive Mitigation Plan: Capital Cost (E): E64,255,000

Annual Recurrent Cost (E) E2,464,200/Year

Environmental Monitoring System: E1.291.733

Project description and the baseline environment

Project Objective

The objective of the project is poverty alleviation allowing the people on Swazi Nation Land (SNL) to shift from risky rainfed agriculture to high value irrigated crops. This follows the National Indicative Programme (NIP), which allocates 70% of development funds to agriculture and rural development. The NIP plans to integrate small holder farming into the commercial market, particularly to diversify and improve productivity, rural incomes and nutrition standards, as well as to increase agricultural exports. It aims for grass roots development, especially the integration of women into the rural economy. The project area is semi-arid, of mainly over-grazed bushland, with some rainfed cotton. Adjacent to the river, there is an expanding area of irrigated sugar cane using water pumped from the Usuthu. These areas are managed by small holders who already get yields comparable with the commercial plantations and, with the project, many others will have the same opportunity.

Project description

Engineering and infrastructure

The project will provide a two stage development of small holder irrigation on 11,600 ha, served by diverting part of the peak summer flood flows of the Usuthu River to off-river storage. The main hydraulic structures are the following:

Headworks and Bovane Dams

The conveyance system to the Bovane reservoir consists of :

• a diversion weir at Bulungapoort on the Usuthu River, a sand trap and a 23km of feeder canal of 13m³/s capacity to Bovane.
• a reservoir in Bovane on the Mhlathuzane River impounded by three dams, Golome dam, Mhlathuzane dam and South Saddle dam, with a surface area of 1,440ha at Full Supply Level, FSL, 224m msl, with a live storage capacity of 155MCM. Maximum flood level is 228m, lowest operating level 207m and bottom outlet 183m.

At Bulungapoort there is sufficient flow for seven months of the year to supply the project with year round water from seasonal storage in Bovane reservoir. The best site for the Bulungapoort weir is about 300m upstream of Holomi railway station, where a slight curve gives good conditions for sediment exclusion. The weir will span 220m with a crest limited to 252.5m msl, so that the 100yr flood of 7,750m³/s at 258m does not impinge on the railway at 258.5m. A "Tyrol" weir is proposed with a flat trash rack and a scour sluice to remove sand from in front of the offtake. The weir height will be up to five metres above river bed. The right bank abutment is set at 259m, left bank 256m.

The head reach culvert of 150m length consists of two reinforced concrete box sections. The sand trap design is based on four indicative sediment samples taken by GFA in December 1997 and February 1998. The head reach flow of 15m³/s includes 2m³/s for flushing the sand trap. It consists of three parallel bays of 23m length and 5m top width, controlled by a weir into which the scouring sluices are set. It is aimed to extract all sediment >2mm.

Canals and associated infrastructure

Two irrigation offtakes at 205m elevation supply water for the project :

• the North Main Canal runs in a NW direction for 5.7km from the dam to serve 2,000ha (1,300ha by 30m pumping and 700ha by gravity)
• the South Main Canal will extend for 24km in a SE direction to serve 4,500ha in the Bovane block, and will run for a further 51km, including a 5km siphon, to the Matata block to serve a further 5,100ha.

Main canals will generally have a concrete lining 150mm thick with transverse joints at 7.5m centres to control cracking. Sulphate resistant cement will be used where groundwater is high, with weep holes against uplift. Transitions will be in reinforced concrete. Escape structures will be side weirs to streams located above siphons and checks. The canal alignments have been decided by balancing cut and fill and by short cutting valleys by siphon or pipe bridge. The North Main Canal pump station consists of 5x250-l/s pumps plus two standby, powered by an 11Kv line taken 50km from St Phillips.

Pumps will be horizontal centrifugal for lifts of 20m to 50m, vertical turbine for 10 - 30m and submerged axial flow for <10m. Each installation will have at least two pumps to achieve three levels of output with a high load factor. Storage will be provided at the head of each pump line for 2hrs flow per day. Night storage will be required at the small holder units, of 2,600m³ for a tertiary peak of 0.61-l/s.ha for 12hr service on 100ha. Sprinkler pressure of 42m head will be generated by the small holders' own pumps. Many pump stations are on the main canals and a power line along the canal service road will be needed. Pumps will serve 28% of the Bovane block and 20% of Matata, requiring 625kW and 236kW respectively. The main canals will run 24hrs/day with flow adjusted to demand.

Road Infrastructure

Well built two lane gravel roads give access to the weir site and close to the dam sites but additional access roads will be needed to Bovane ridge. The Usuthu will be crossed at Siphofaneni for light traffic and Mapobeni for heavy traffic. Most materials will be obtained from within the reservoir and borrow pits, so that quarries and construction roads eventually will be submerged. Access to farms will be from the local road network and by the Maphobeni bridge from the Bovane block to the sugar mill. The feeder, main and secondary canal service roads will become public roads. The cost estimates include a provision for farm roads but details will await the design phase.

The service road will be 3m gravel with 1m shoulders and interception drainage ditches in cut reaches. Bridges will be 6m two lane for major roads and 4m one lane on minor roads constructed of single span reinforced concrete. Precast concrete pedestrian/cattle bridges will be provided at 0.6km intervals.

Agricultural

The planned irrigated areas can be summarized as follows :

Main canal and secondary system layouts were prepared on 1:5,000 ortho-photo maps with 5m contours. All-weather roads, settlements and cadastral boundaries for title deed land were added. Soil units were transferred from the 1:50,000 soil maps. S soils were planimetered separately. Command area was taken from 205m with a canal slope of 3dm/km and structure head losses. The reduction in area from the pre-feasibility study, particularly in Matata, is said to be due to inaccuracies on the 1:50,000 maps. The secondary canals will follow ridges below the main canals and contours when pumped above. The smallholder units will draw water directly from secondary canals, or tertiaries when topography dictates. The units will be adjacent to natural water courses for drainage. Bovane land slopes often exceed 5% which would require drop structures at 25m intervals in open canals; these which would be difficult to excavate in shallow S-set soils. Open canals are prone to cattle damage and water theft. Pipes are therefore proposed for slopes >2%.

There is a thriving sugar industry and a cotton ginnery, both ready to expand. Input suppliers and transport exist in the project vicinity. Small holders get comparable sugar yields to the estates, which should be replicated on the project with more crops. Crop choice will be free depending on agro-climatic conditions, soil suitability and farmer experience, but the study recommends sugar, cotton, dry beans, seed maize, green beans, bananas plus others.

The direct beneficiaries will be approximately 3,400 farm households or 30,000, people and 5,600 jobs will be created. During the eight years of implementation there will be some 5,300 "man years" of employment in construction.

Project Costs and institutional arrangements

The components of project cost are (in Euro)^:

Foreign exchange will be 25% and local costs, including Rand, 75%. It is proposed to divide the work into several construction lots.

For sustainability the project needs professional management and an organisation to co-operate with traditional authorities. Central Government and the MOAC do not have a strong presence and lack experience of large project implementation. In any case, Government is committed to downsizing. It was therefore proposed (Feasibility Study) to set up a contract Project Management Unit (PMU) which would develop an institutional framework to manage irrigation after implementation. Farmers would own and manage the system, which must be self financing through the Lower Usuthu Irrigation Management Association, LUIMA, with a voting majority by elections from water user groups. There would also be Government and agro-processor representatives. LUIMA would be financed by monthly metered water charges levied on participatory small holders.

Current Status of Project

Although Economic Internal Rates of Return (EIRRs) are somewhat marginal, funding for the project is being sought. Agricultural processing investment will be covered by private sources; finances for headworks will be sought from regional and international sources; farmers themselves would be expected to cover a portion of field preparation and working capital, and certainly cover operation and maintenance charges. With sugarcane and cotton being the major crops, and thus central processing and marketing facilities being involved, cost recovery will be much easier. Key members of the GFA Team are in the country to assist in formulating suitable financial packages for the programme (and other activities).

Key to the project being able to go ahead will be the results of this EIA study. By Swazi Law, a an environmental compliance certificate must be issued by the SEA before the project is allowed to go ahead: for international funding, equally strict criteria also apply.

Proposed schedule/workplan for project implementation

The timing proposed is Yr0 for the finance package; survey and geotechnical work would be undertaken in Yr1, final design and tender documents in Yrs 1/2, tendering and award in Yrs 2/3, upstream construction in Yrs3/5, downstream construction and land development in Yrs4/6, first water in Main Canal North (MCN) in Yr 5, MCS in Yrs6/7, start of Stage II in Yr 7.

Description of the affected area and surrounding location

The Study Area lies on the South side of the Usuthu and is some 40km East-West and an average of 8km North-South. It refers to areas mapped and discussed in the Feasibility report. It is likely that exact location and areas may change somewhat after detailed design stage. It comprises the following components :

Table EIA-1 - Affected areas

The location of the Project area in relation to the total Usuthu watershed and adjacent rivers is shown in Maps Section - Map 5. More details of the Project Area in relation to existing settlements and existing proposed infrastructure is shown in Appendixes-Maps 1 and 7. Areas of the main chieftaincies are shown in Maps Section - Map 6: areas of apparent overlap of allegiances are reflected by diffuse shading.

Physical

Geology

The Study Area is underlain by predominantly igneous rocks, which range from basalts in the East to mainly dolerite in the West, although intrusions of acidic rocks also occur in the extreme West. In the West and particularly the South-west, Carboniferous sandstones of the Karroo Formation occur, including finer-textured lenses with more saline materials. The area is bounded in the East by the very prominant Lebombo Ridge, again of basaltic material. N-S-oriented fault lines to the immediate West of this ridge mark almost the southern extemity of the Great African Rift Valley. Because of this tectonic feature, there is a moderate risk of earthquakes, although during recorded history only minor tremors have occurred. Moving away from the rift fault-lines, earthquake risk declines markedly from East to West : at the Bovane Reservoir site risk is only very slight.

Depth of weathering the basalt and dolerite is rather shallow, with hard bedrock being encountered at 50 to 150cm depth for the vast majority of the area. Any requirement for deeper excavations will thus entail much higher costs, involving considerable blasting. Much of the bedrock shows deep fissures, which are very significant for water movement, tree rooting, and will necessitate grouting for dam foundations.

Superficial formations are almost notable for their absence. Some sandy river terrace materials occur at some 7-20 m above current river bed level, but these are neither continuous nor extensive, extending back only for a maximum of 200m in places. The Usuthu River is down-cutting into the hard bedrock westwards from Maphobeni, whilst further downstream sandy and silty materials fill most of the riverbed, but are very shallow. Any sandy alluvial aquifers are thus very localised. Colluvial deposits in concave footslope positions are rarely much more than 1.5 m in thickness, and are rich in montmorillonitic clays. Any construction in these sites thus entails problems with foundations.

Physiography

The area varies in elevation from 100m at the Usuthu river in the East to some 340m in the undulating plataux in the south west. The Usuthu in the extreme west of the area at the Bulungapoort Gorge is at 250m : the Sinceni Mountains to the southwest rise rapidly to some 930m, with mass-movement scars being very visible on the steep mountain sides. The Usuthu is joined by two prominent rivers on its south bank, the Mhlatuzane, draining a large catchment area of 540km², and the Mhlatuze, draining a larger area. A much smaller river, the Mfulangwenya, effectively drains the major part of the Matata Block, with a catchment area of some 60km². Gradients are more gentle in the Matata Block in the East, with a very gradual rise from the banks of the Usuthu at Matata at 120m to the interfluve, representing the water divide between the Usuthu and Mwgauma Basins at 200m.

Further to the west of the Matata Block the land rises to some 400m elevation, with rolling to hilly land showing very shallow soils, and characterised by dolerite dykes running in a marked N-S direction. Still further west, a very pronounced dolerite dyke is represented by the Bovane Ridge, which rises from the bed of the Mhlatuzane River at 200m to the ridge top at 250m.

Within the proposed irrigation areas, slopes are mostly in the range 0-6%, but steeper land, going up to 10% slope, is also common. Below the line of the main canal, much steeper areas, including land of around 25% slope, also frequenly occurs, but this corresponds to shallow soil areas and is not planned for irrigation.

Climate

Rainfall in the area varies from about 520mm in the East (at Matata or Big Bend), in the rain shadow of the Lebombo Ridge, to some 750mm in the West, with most rain occuring in the period from mid-October to end-April. The rainy season can be considered as rather long, but very unreliable. Dry periods of 3-6 weeks duration can occur at any time, making for very low average yields of rainfed crops, and frequent crop failure. For two years in five, crops give practically no harvestable yield, although livestock can make use of the crop renmants. Appendix 2 graphically illustrates the lack of reliability of the the rainy seasons over 16 consecutive years for Kabuta (just to the West of the Project Area) and St Phillips (East-Central). (Methodology used for this rainy pentade analysis is discussed in Section 6.1 below whilst more detailed discussion on rainfall and other parameters of climate is given in the Technical Annex C "Hydrology").

Temperature in the area averages 22 deg C thorughout the year, vaying from a monthlyt mean of 16 deg in July to 28 deg in January. Absolute maximum temperatures may touch 40 deg and minimum 0 deg, although incidence of frost is rare. Temperatures may change markedly from day to day, at any time of year, due to changes of wind direction: cold air form the southern Ocean inter-changing with hot air from the African continent.

Wind is rarely a major hazard, although downdraughts associated with thunderstorms can cause damage locally, as can hail. Cyclones occur periodically – maybe one event in 25-30 years – but damage from heavy rainfall and accompanying flooding exceeds by far any wind damage.

Potential evapotranspiration varies from some 2.8mm/day in July to some 6.5mm/day in January. Daily irrigation demand can thus be much greater during dry periods in the rainy season than during the dry season

In general, climate can be considered as highly suitable for irrigated sugar cane, with a long dry season, cooler winter temperatures facilitating cane ripening, high sucrose contents, and a long milling season. Winter temperatures and low humidities are also conducive to good-quality irrigated vegetables. Any damage by the very occasional slight frosts can be mitigated by spray-irrigating on nights of frost risk.

Soils

Soils depth is the major limiting physical factor in the project area, with soil texture and profile differentiation being important, but subsidiary, characteristics. By potential land use characteristics, there are essentially six major groupings of soils:

• Moderately deep to deep, reddish, well-structured clay loams to clay soils, of good fertility status, moderate to high CECs and AWHCs (mainly R; also F, L, N sets – 21.5% of the project area, 40.4% of the gross irrigated area: high to mod.high suitability for most crops);
• Deep to moderately deep, black, well-structured vertisolic (montmorillonic) clay soils, of good fertility status, very high CEC; moderately high AWHC (mainly K and C sets – 13.9% of the project area, 25% of the irrigated area: mod.high suitability for cotton ; marginal for sugarcane – high yields are possible but good management is needed; drainage and sodicity problems may also be expected);
• Moderately shallow to shallow, reddish, moderately well-structured clay loams, moderate fertility status, moderate CECs and AWHCs, total AW only limited by soil depth (S set – 36.6% of the project area, 24.3% of the irrigated area: soils are considered only marginally suitable, and only under sprinklers, for sugar cane);
• Moderately deep to deep, greyish brown, very poorly structured duplex soils, comprising very sandy topsoils over compacted sandy clay subsoils, of limited permeability to roots and water. Soils are considered non-irrigable and are highly erodible. (H and Z-sets –3.9% of the project area but not included in the irrigation area; considered not suitable for irrigation, but deserve further research);
• Shallow to very shallow, generally poorly structured soils of variable texture, (Mainly U-set; 18% of project area, but not included in the irrigation area; not suitable for irrigation, may be suitable for social forestry, houselots, construction, and some rough grazing);
• Deep but sandy riverine soils, generally poorly to moderately structured, of moderate to low CECs and AWHCs (B and W-sets, 6.1% of project area, and 10.2% of irrigated area; considered variably suitable for a variety of uses, depending on soil texture).

The mapped project area here totals some 26,300ha with the gross irrigation area 13,740ha, and the net irrigation area 11,600ha.

Hydrology and water quality

The Usuthu River at Bulungapoort has a catchment area of 12,600km², of which 7,800 (60%) is in RSA. This RSA portion, however, contributes only 40% of the flows : four dams controlling some 2,600km², and large areas under forest plantations, explain the lower than expected contribution to flow. Mean annual run-off (MAR) at Bulungapoort, measured at GS6 is 1,450m cu m (equivalent to an even flow of 46cu m/s, and a net run-off annual depth of 115mm). Usuthu flows are highly variable from day to day, or even hour to hour. (The annual irrigation requirement of 197m cu m thus represents 12% of total MAR). The tributary Mhlatuzane river has a catchment area of 540m cu m and a MAR of 39m cu m, (equivalent to an even flow of 1.2 cu m/s, and a net run-off depth of 72mm). Open-water evaporation in the area is some 1300 mm / year.

The Usuthu River has low flows fully allocated for irrigation, and in some dry seasons flow stops entirely. Zero flows more frequently occur in tributary rivers, such as the Mhlatuzane.

The current cross-border flows into Mozambique have not been measured, but were calculated in the Fesaibility Study on the basis of the low (September) flow of 17cu m /s at GS6 and applying a deduction of 80% of the normal and surplus water rights, giving an average September border allowance of 6.6cu m/sec at Bulungapoort. On this basis Feasibility Study derived allowances for November to May as 7.4, 8.6, 8.6, 7.6, 7.3, 6.6, and 6.6 cu m/s.

Current sediment load in all rivers is considerable, with actual measurements almost certainly being a gross under-estimate of actual amounts. Frequent observation by staff of Swazi Tails, undertaking white-water rafting in the Usuthu, attest to very high sediment loads during high river-flows. More importantly, for the critical Mhlatuzane catchment, net sediment yields are estimated at some 15-20 tonnes/ha/year, sufficient to reduce the storage capacity in the proposed reservoir by 10-15% within 25 years. Erosion control, and community participatory approaches to land use planning and soil conservation are thus of critical importance.

Other water quality parameters are of somewhat lower concern. Anions and Cations (see Appendix IV – Glossary) are at low concentrations in all river samples, although there is a four-times concentration of ions in the low-flow periods between Siphophaneni and the Mozambican border. With more irrigation, and increasing efficiency of irrigation use, ion concentrations are likely to rise. Nitrate and particularly phosphate concentrations are also at rather high levels during these low-flow periods.

Hydrology, sedimentation and water quality are considered in detail in Technical Annes C "Hydrology", and the hydrological interface with engineering is discussed in Section 3 of Technical Annex B "Engineering".

Vegetation

A vegetation field survey was undertaken within relatively undisturbed natural vegetation areas within the proposed irrigation areas, Bovane reservoir area, and proposed canal areas. Thirty seven 200 sq metre sample plots were selected, based on lithological and vegetation structure stratification. Close relationships were seen between vegetation and soil type; soil depth, parent material, and rainfall zoning all being important factors.

Some 37 existing and potentially rare and endangered species were identified as occurring in the Study Area. Eight of these are currently not so rare, but, because of high economic demand, they are likely to come under threat in the near future. Vegetation type 5 (alluvial and riverine soils associated with the drainage network) was shown to have by far the highest biodiversity value ( and also high biological productivity).

The most serious negative impacts from the project would be loss of dense bushland over some 8000ha; further fragmentation of the Lowveldt bushland; increased pressure on remaining natural bushland areas by a much increased population; and a much increased threat to the 37 rare and endangered species of plants. The Bovane Reservoir area in particular shows a relatively high concentration of dense bushland, and much of this would be lost to reservoir inundation. The most valuable habitat, in terms of species richness as well as biological productivity, is the riverine floodplain and river bed communities, and this would be under threat from increased pressure from humans and cattle. (see details in Technical Annex E "Vegetation").

Fauna

The largest catchment in Swaziland is undoubtedly that drained by the Great Usuthu and its feeder tributaries. The river has its main source in South Africa from catchments such as the Little Usuthu, Mpuluzi, Great Usuthu, Ngwempisi and Mkondo. Within Swaziland some of the headwaters of the Usuthu's tributaries, such as the Motshane, originate in the Malolotja Nature Reserve on the slopes of Ngwenya Mountain, in the Highveld region. From western Swaziland through to eastern Swaziland the Great Usuthu River system therefore supports one of the largest fish biomasses in the country.

The vertebrates of Swaziland have been relatively well documented. The Swaziland bird atlas (covering a seven year period from 1985 to 1991) provides detailed distributional information on an eighth-degree grid cell resolution (approximately 11 km × 12 km) (Parker, 1994). The distribution and general biology of the mammals of Swaziland have also recently been published (Monadjem, 1997a, 1998). The distributions of Swaziland’s mammals have been plotted on the same grid cell resolution as the bird atlas. Distributional information (at the quarter-degree grid scale) on the reptiles and amphibians of Swaziland have also been published (Boycott, 1992a,b), together with an annotated checklist of the fishes of Swaziland (Hyslop, 1994). To date, 493 species of birds, 127 species of mammals, 111 species of reptiles, 44 species of amphibians and 56 species of fish have been recorded from Swaziland, underscoring the high diversity of these groups in the country (Monadjem, 1997b).

Despite a fairly good national coverage, the vertebrates of the Lower Usuthu Basin have not previously been adequately documented. The core of the proposed development area of the Lower Usuthu Basin falls predominantly within four eighth-degree squares, although the impacts of the proposed development will be felt over a far larger area. The numbers of bird species recorded in these four squares, from north-west to south-east, are: 180, 152, 196 and 181. Neighbouring squares to the north and south (falling within the same vegetation zones) recorded far more species 229, 267 and 300 species. These differences in species diversity can be mostly accounted for by differences in sampling intensity. The neighbouring squares mentioned above were far better surveyed than the squares falling within the Lower Usuthu Basin core study area (Parker, 1994). From a mammal perspective, the study area is even less well covered. Although extensive field surveys have been conducted in adjoining areas to the north and south of the study area, none have yet been conducted within it Monadjem (1998). The subtropical lowveld of Swaziland has a rich and diverse herpetofauna with over thirty (30) species and subspecies of amphibians and more than seventy (70) species and subspecies of reptiles having been recorded (Boycott 1992a, 1992b). A large variety of fish species was recorded by Clay (1976), in the Usuthu River system. However, recent surveys have indicated that many of these may now be absent from the Usuthu (Hyslop 1994).

Very little work has been done on the invertebrates of Swaziland (with the possible exception of the moths and butterflies). Consequently, this group has not been adequately represented in previous Environmental Impact Assessments and ecological surveys. Aquatic macro-invertebrate communities are made up of a heterogeneous assemblage of animal phyla. The structure and function of these communities may reflect the underlying abiotic and biotic conditions of stream ecosystems. These organisms occupy an important role in the food chains of aquatic systems. The various members of this community responds to environmental stresses in different ways. In addition to their relatively long-lives (over a year), macro-invertebrates are sedentary and are thus unable to easily escape pollution, as more mobile organisms like fish can. These characteristics consequently make this group of organisms suitable for use as biological indicators.

Macro-invertebrates inhabit all levels of a river’s ecosystem. These vary from surface dwellers, e.g. water striders (Family Gerridae), those just below the water surface, e.g. backswimmers (Family Notonectidae) as well as benthic and vegetation dwellers, e.g. water scorpions (Family Nepidae), water bugs (Family Belostomatidae), damselflies, mayflies (Ephemeroptera) and dragonflies (Order Odonata). Some occupy shallow, slow flowing waters, e.g. whirligig beetles (Family Gyrinidae). All aquatic macro-invertebrates have various adaptations which enable their survival within their habitats. These may either be the ability to withstand the water current, e.g. mayfly nymphs which have flattened bodies and attach themselves onto vegetation or by sheltering under rocks, stones or mud.

Socio-Economic

Administrative framework

The study area is situated in the Lowveld area of Swaziland, in the Lubombo District with the district capital of Siteki. Two administrative bodies operate alongside each other:

The Traditional Authority system: clusters of dispersed homesteads form part of a local community (sigozi), led by a headman (induna) appointed by a chief (sikhulu). These communities make up a chiefdom (sive), which is an administrative and political unit which recognises the chief as its head. The position of chief is hereditary and passed on patrilineally. Chiefs rule their chiefdom with assistance from a council of advisors (libandla).

The proposed project covers portions of seven chiefdoms. Areas claimed by the chiefdoms were indicated by chiefs or their representatives on 1:50 000 maps, confirmed by information gained from the socio-economic surveys:

• the weir site at Bulungapoort is situated in the late Chief Shwapha Dlamini’s land. It adjoins that of the late Chief Sonjane Dlamini, who is referred to in the Pre-Feasibility and Feasibility reports as the chief responsible for this area; some residents did indicate at the interviews that they are registered with the Phunga community and named Siyendle Lukhele as their "chief";
• Chief Maphilingo Shongwe resides in the Habelo community area. His chiefdom extends as a strip of land south of the Usuthu River to the Mhlathuzane River and for a portion beyond, adjoining the late Chief Gamedze’s domain to the east and TDL to the west. He is apparently a "new" chief and is not recognised by all;
• the late Chief Madlenya Gamedze’s domain spreads south of the Usuthu River near Siphofaneni, with the Mhlathuzane River as the eastern boundary and Chief Shongwe’s chiefdom to the west;
• Chief Lubhodzela Ngcamphalala is presently acting on behalf of the late Chief Mhawu Ngcamphalala, in a chiefdom that extends over a large tract of land south from the Usuthu River to King Mamba’s chiefdom, west from the Mhlathuzane River, and east to the Lubombo mountains. Portions of land claimed by Chiefs Matsenjwa and Mngomezulu are historically claimed by the Ngcamphalala clan;
• King (Makhulane) Maja II Mamba’s chiefdom stretches southwards of the Ngcamphalala community and westwards of the Mngomezulu community, across the Mhlathuzane River, with Sithobela as the main centre. It is divided into two main portions, Jeremiah Mamba being the indvuna responsible for land falling within the study area;
• Chief Ntunja Mngomezulu’s area lies to the east of King Mamba and south of Chief Ngcamphalala; and
• The Matsenjwa’s claim land stretching from the main Big Bend/Lavumisa road towards the Lubombo mountains, bounded by Ndzevane to the south and Ngcamphalala territory to the north. The majority of residents in the area recognise Chief Mbekwane Matsenjwa, who lives across the mountains in South Africa, while others recognise MP Joseph Matsenjwa, resident in the area. The Ngcamphalala and Mngomezulu chiefdoms historically lay claim to portions of this land since the Matsenjwa’s moved into the area within the last few decades, having sought refuge from Ngwevuma in South Africa after unsuccessful attempts by Swaziland to incorporate the area into the country.

The parliamentary Tinkhundla system: constituencies elect a Member of Parliament (MP), and a counselor (bucopho) to the MP. The study area is made up of four Tinkhundla: Siphofaneni, Sithobela, Nkilongo and Lubuli.

Demography

The estimated population figure of 13 303 for the study area was gained from the most recent GOS Central Statistical Office’s census data (1998). This coincides with the broader 1996 population estimates presented in the Pre-Feasibility study, where chieftaincies were used to demarcate boundaries; it was estimated that 1194 homesteads were situated in the Bovane Block, giving a population of 8 358, and 823 homesteads in the Matata (Mfulangwenya), giving a population of 5 761, to total approximately 14 119 people.

Table EIA-2 - Population census figures for the study area by gender and age

Source: CSO, 1999

All evidence points to a youthful population, with 58% of the population below the age of twenty. The tables are supported by the results of the socio-economic survey (Table EIA-3), with 56.8% of the sample below the age of twenty (median of 17 years) and an overall average recorded age of 21.3 years.

Table EIA-3 - Age distribution

Source: Socio-Economic Survey, 1999

This age distribution will have a future impact on the number of people who theoretically will be available to work on the irrigation scheme and take advantage of related secondary employment opportunities.

The sampled survey shows that there are slightly more women (50.2%) than men (49.8%) in the study area. The census data support this; Table EIA-2 indicates that there is a larger percentage of women (51.2%) than men (48.8%), and particularly in the middle age group. This may indicate that men of working age leave the area in order to seek and obtain wage employment.

The survey also indicated that 19.6% of the sampled population were of pre-school age, at crèche or at nursery school. Of those of school-going age or over, educational levels attained were relatively low: 20.9% had never attended school, 11.1% had reached Grades 1 or 2, 23.3% Grades 3 to 5, 17.6% Grades 6 or 7, 9.7% Forms 1 or 2, 12.9% Forms 3 to 5, and 1.4% had some post-school qualification. This is supported by data on educational levels achieved in the 20-49 age bracket calculated from a Demographic and Housing Survey (1991), which showed that for the Lubombo Region 30% had no education, 37% had attained primary school level, 27% secondary and high school, and 4% university or vocational training (Booker Tate, ibid). According to Ministry of Education projections cited in the Physical Development Plan 1996-2006, the region also has the highest population per high school, namely 5 000 (Booker Tate, ibid).

Although there is an indication that formal educational levels may be low, surveyed household members displayed a range of additional skills. The most predominant centered around farming, driving, mat and other handicraft making, and sewing and knitting. Others included semi-skilled activities such as building.

Homestead and household characteristics

Homesteads in the area are generally dispersed, with people living close to the land which they have access to. In some parts of the study area, such as in the Matsenjwa and Ngcamphalala communities near the Lubombo mountains, people are grouped in settlements as a result of government development schemes, where land was allocated specifically for grazing, cropping and residential purposes and people were moved accordingly.

The majority of households surveyed have been living at the same place of residence for more than ten years (65.0%). Reasons given for moving homestead relate to breaking away from a primary household in the same area (usually a son establishing a family, or a new wife creating a separate household), a member returning to the area of his or her family, or relocation through a resettlement scheme. Any population growth can therefore be seen primarily as a natural increase rather than a large-scale movement of people into the area.

According to the survey, the average de facto household size is 8.5, ranging from one to 38 household members (the median is 8.0 and the mode 7.0); this average will be used throughout the study. Other studies give varying figures:

• the Pre-Feasibility study (Booker Tate, ibid) states that, while the 1986 CSO census found an average homestead size of 7 people for the Lubombo District, higher averages have been recorded; for example the 1994/95 CSO Annual Agricultural Survey found an average of 8.8 persons per homestead on SNL in the Lowveld, and the International Fund for Agricultural Development (IFAD) baseline study (1995) an average of 10;
• the Feasibility study (GFA-Agrar, ibid) found an average of 9,5 persons per homestead amongst the sample of rain-dependent farmers surveyed, greater than the national average of 7,7 to 8,8 people as per the IFAD-financed SADP (1995);
• the study conducted by the Economic Planning and Analysis Section (EPAS) of the MOAC (MOAC, ibid) found a de jure household size of 7.2 and de facto of 8.2.

Table EIA-4 - Household Size

Source: Socio-Economic Survey, 1999

Ninety-two percent of the household members sampled lived permanently at the homestead, reflecting a relatively stable resident population. A further 11% lived away as migrant workers, and four percent were at educational institutions.

By far the majority of household heads were male, and mostly were resident. Some had more than one wife, and hence more than one homestead if a wife had elected to stay elsewhere. The composition of households reflected extended families, with 11.0% being spouses of the household head, 61.3% children or their spouses, 1.4% parents, 7.1% siblings or their spouses, 15.1% grandchildren or great-grandchildren, 3.1% nieces or nephews, and the remaining non-family members.

The mean number of structures per household for those interviewed for the socio-economic survey was 4.7, with a median of 3.5 and maximum of 16. The mean number of rooms was 1.4 per structure. The majority of homesteads were made of wood and mud (39.7%), followed by stones and mud/cement (31.3%), cement blocks (17.7%), wooden poles (7.7%), a combination of wood, stones and mud/cement (1.3%), or other materials such as reed, iron sheeting or bricks. Where structures were complete, roofing was of thatch (54.7%) or tin (45.3%). Although some houses had "no floor" (19.0%), those that did were made of cement (71.3%) or dung (28.7%). Eighty-seven percent of the structures were being utilised.

Sixty-five percent of the surveyed households had a kraal, either for cattle or goats, with an average of 1.6 kraals per household; 28.5% of households had a storage tank for maize; 22.3% had a water tank or water jars; 10.5% had a poultry shed; 8.6% had a pigsty.

In terms of movable assets, 85.0% of households had a radio, 22.3% a television set, 21.9% had access to a motor vehicle (generally a motor cycle), 9.5% a tractor, 9.1% a plough, and 8.8% a trailer.

Economy

It has been stated in previous reports (Booker Tate, GFA-Agrar, EPAS, ibid) that the area is one of the poorest, least developed in Swaziland, with income amongst the lowest. This can be supported to some extent in cash income terms by the findings of the socio-economic survey, "poverty". It became evident in the study area is:

• that there is a heterogeneous rather than an homogenous population;
• that levels of poverty vary between different areas, depending on access to natural resources, services, and opportunities for employment;
• that the large majority of households have access to land and attempt to grow a cash crop, even if they are dependent on rainfed agriculture;
• given the high proportion of households, and members, that are involved in farming activities, on an individual household and group (association, co-operative) basis, there is a solid accumulative skills base in this sector;
• that there is a level and mix of skills within the population that relate to activities other than farming;
• that although there are no significant income earning opportunities within the area apart from farming, many household members are involved in some informal self-employment activity, from marketing home-made goods to offering a local service;
• that people, and particularly women, have formed a range of organisations, from money lending bodies to vegetable gardens, as places of interaction and support;
• that people have developed coping mechanisms for critical times; for example social support networks and organisations.

The following trends emerged from the socio-economic survey:

Employment status: Whereas 29.1% of those of employable age were not employed, 39.0 % were self employed (including cash crop farmers) and 31.9% were wage employed. This indicates a predominantly productive society, even though returns on labour are relatively minimal in cash terms.

Main employment activity: Farming (comprising over one half of the employed), either working on the household’s land or as a labourer, driver or supervisor for a farming group such as a local farmers’ association or co-operative, a sugar cane plantation (Simunye, Mhlume, Illovo and Crookes), or other plantation (Tambuti).

Other wage employment: Wage-employed household members In the building/construction sector, with the army or police, domestic work, sales personnel, office work, semi-skilled work in the motor industry, technical labour, teaching or as a health motivator. Most of the work involved living away, in different parts of Swaziland. Only 4.9% of those employed lived and worked in South Africa, indicating little dependence on that country for migrant remittances.

Cash income: Using all income sources, the Mean/average gross household cash income was estimated at E1 179.70 per month (N=259 households), translating into an estimated E14 156.40 per annum (E138.79 monthly per capita, using average household size of 8.5). This Mean may be regarded as skewed because of the inclusion in the sample of a few high-income earners; the Minimum income was nil and the Maximum E25000 per month. If households which rely on irrigation for crop cultivation are excluded from the above equation, the following estimates were found: Mean monthly household income, E1 063.88 (household annual at E12 766.56, per capita monthly at E125.16); range of income from nil to E11 633;

If only regular sources of income are included in the estimates (i.e. remittances, employment, pension, property and agriculture), the average monthly household income was estimated at E904.31 (annual at E1081.72; per capita monthly at E106.39). These figures are higher than those found by Feasibility Study (1998) and the EPAS (1998) studies. This is discussed in detail in Technical Annex D "Socio-Economics".

Primary income sources (monthly average): crop and vegetable sales, E5 444; livestock sales, E1 004; self-employment, E596; remittances from South Africa, E569; locally based wage employment, E460; remittances from Swaziland, E429.

Savings: Only 22.1% of the sample stated that they have cash savings. If short of money, the majority ask family, neighbours or informal lending facilities to assist rather than a bank or other formal institution.

Land

Swazi Nation Land (SNL) is held in trust by the King for the nation. However, in practical terms under Swazi customary law the chieftaincy is the primary organisation base for the regulation, management and control of the use of community resources, including the allocation of land to homesteads, and control of the use of other natural resources such as grazing.

All married Swazi males may acquire the "right of avail", which includes the right to cultivate land, to residence, and to use communal property resources. Married men may allocate fields to women in the household; in some instances unmarried women with children may be allocated land in their own right by a sympathetic chief; land is also acquired by inheritance, by men and women.

The majority of households sampled in the socio-economic survey (that is, 91.4%) indicated that they had access to land for cultivation, land which belonged to them. Most were smallholders, with an average of 8.8 hectares (median of 6 ha; mode of 3 ha). Of the land cropped, maize and cotton were the primary crops, with all cotton being sold, and all maize being consumed by the household. Labour was mostly family. Some farmers were involved in sugar cane production, usually through a farmers’ association or co-operative. Reasons given for land not planted were, in order of importance:

• a shortage of water / drought or late rain,
• no available labour,
• no capital,
• land not having been cleared,
• no tractor for ploughing, or fields lying fallow.

The problem of cattle or goats eating crops was cited by a number of respondents, and in some instances farmers who had no fencing stopped growing crops solely because of the risks of the plants being destroyed. Of those with fencing (74.8%), most had brushwood.

There was an average of 2.9 fruit trees, of various types, and 4.4 shade trees found on the property of households across the sample.

Livestock

According to the socio-economic survey, the average number of animals kept by households were:

• cattle owned, 17.6; sisa’d in, 12.3; sisa’d out, 10.4
• goats, 17.3
• sheep, 11.4
• pigs, 5.3
• poultry, 18.9

Some households owned horses or donkeys, but numbers were limited.

Most cattle graze around homesteads, or in communal areas. Animals are watered at rivers or streams (69.3%), from a homestead’s private supply (13.3%), from a spring or borehole (9.8%), at a dam (4.2%), or from a communal tap (3.4%).

Cattle are sisa’d out for a number of reasons: a household may not have an available person to look after the cattle, or have too many cattle to look after, or there may not be enough grazing land to support the cattle. On the other hand cattle are brought in to provide milk, as draught power, to pull sledges or to carry loads.

Infrastructure and social services

There is a reasonably good infrastructure of roads to and within the study area. A main tarred road passes from Manzini through Siphofaneni, following the north bank of the Usuthu River to Big Bend, and thereafter curving south towards Ndzevane and the South African border post at Lavumisa. An untarred road branches off from this main road at Siphofaneni, across the Usuthu River, dividing into two access roads:

• to Sithobela, passing through Habelo; a further right branch goes through Mphaphati to the weir site, towards Sidvokodvo and eventually Manzini;
• to St Phillips and beyond, across the Mhlatuzane and Mhlatuze Rivers through the major part of the study area, with a branch leading into the Bovane dam site area. This road joins an untarred road from Big Bend to Nsoko, which links with a tarred road to Matata at Mndobandoba.

At present these roads cross the Usuthu River into the study area at two low-level bridges, the first at Siphofaneni and the second near Mndobandoba. A further low-level bridge is nearing completion, at Maphobeni.

A railway line runs parallel with the Usuthu River, on the north bank, past the weir site and Siphofaneni to a station at Phuzumoya. It crosses the river, running close to the south bank of the river through to Matata station, where it curves through the middle of the Manyonyeni Cattle Breeding Station on to Nsoko.

Other infrastructure and services (formal and informal) are largely limited to the main centres serving the study area, namely Siphofaneni, Matata/Mndobandoba (see Appendix II, of Technical Annex D, Status Quo Reports) and Big Bend. Facilities include supermarkets, petrol stations, police stations, banks, clinics and churches. Smaller villages exist within the study area, with basic service provision such as shops, schools and transport systems; at Mphaphati, Habelo, Madlenya, Mahlabatsini, St Phillips, Mahlabaneni, Matsenjwa and Ndzevane.

Schools in the study area:

• Mphaphati Primary School;
• Thandweni Primary School;
• Madlenya Primary School;
• Mahlabatsini Primary School;
• Sibetsaphi Primary and Secondary/High Schools;
• St Phillips Primary and Secondary/High Schools;
• Elulakeni Secondary School;
• Mndobandoba Primary and Secondary/High Schools;
• Mahlabaneni Primary School;
• Ndzevane Primary and Secondary/High Schools;
• Manyonyeni Training Centre, at Manyonyeni Cattle Breeding Station.

Although most children attend local schools at primary school level, many are sent to educational institutions outside the study area, particularly for higher-level schooling; this includes to Siphofaneni and Big Bend, Mbabane and Manzini, and as far as South Africa.

Energy and telecommunications

In general, there is no electricity supply to communities in the study area; the exception is in the St Phillips and Mndobandoba communities where some Swaziland Electricity Board (SEB) infrastructure is in place. Electricity is also available to homesteads in the village of Siphofaneni. Households rely on natural resources, paraffin, liquid gas and candles as fuel and energy sources. Wood is the primary source of fuel for cooking, collected from surrounding areas.

Currently telecommunication facilities in the study area are limited. A telephone service is available to a few, such as at St Phillips or in the Siphofaneni surrounds; of all sampled households only two indicated that they had a telephone on their property. Residents in the study area travel to St Phillips, Siphofaneni or Matata to make use of public telephones. At present cellular phones can only be used in certain areas, near to Matata and the main road between Big Bend and Lavumisa.

Water supply and sanitation

Forty-one percent of the surveyed households had no toilets at all, 39.9% an outside pit latrine, and 18.6% an outside VIP toilet.

Water for domestic purposes (cooking, drinking and washing) was obtained from a variety of sources, depending on the area and water availability:

• rivers, streams: 60.6%;
• rain collected at the homestead: 51.1%;
• boreholes: 32.6%;
• communal standpipe or tap: 18.6%;
• dam: 3.0%;
• private tap at the homestead: 2.7%;
• water supplied by the government, or brought in privately for sale: 1.5%.

For those who collect water away from the homestead, 47.5% take more than twenty minutes on a return trip, 39.2% from ten to twenty minutes, and 13.3% less than ten minutes.

Many people interviewed complained about a poor or inconsistent water supply, and thus looked towards the proposed project (canals) as a dependable water source. Some communities have initiated the construction of a dam in their area, through the MOAC; they have apparently been told to wait for the Usuthu Smallholder Irrigation Project - to provide their water for the future.

Public Health

Although it is officially recognised that HIV/AIDS is reaching pan-epidemic proportions in Swaziland, few respondents to the socio-economic questionnaire recognised it as a health problem in the area (11.0%). Rather, malaria, diarrhea, Acute Respiratory Infection and skin infections were cited as the most prevalent illnesses.

The majority of people in the study area (92.8% of the sampled households) use a clinic when sick; at Sidvokodvo, Siphofaneni, St Phillips, Sithobela, Ebholi, Ndzevane, Lubuli, Gucuba or Big Bend. A major complaint was distance traveled to the nearest clinic (two-thirds indicated that they live over 8 km away), particularly as most people have to walk, take public transport or hire a vehicle in emergency cases. Other complaints centered around the hours of operating (not open at night or over weekends), difficulty in obtaining a ticket for attendance, the cost of services, the inadequacy of drugs and equipment and, to a lesser extent, the attitude of staff.

When ill, 73.4% of the households stated that they seek help from a hospital (such as at Siteki or Manzini), 42.2% from a traditional healer, 23.6% from a health motivator, 23.2% obtain over-the-counter medicine from a shop or chemist, while others rely on home treatments or "prayer".

Cultural heritage

Graves are situated in communal grave sites or at individual homesteads. When asked whether a household has a grave, over half the respondents to the socio-economic survey (53.3%) said that they had none. Of those with graves, one third indicated that their household only had one, 14.8% that they had two, a further 15.6% that they had three, and the remaining 36.9% more than three.

Natural hazards

The Usuthu, Mhlathuzane and Mhlatuze rivers undergo rapid changes in water levels during flood events, which primarily occur over the early to mid summer months. This prevents river crossings for weeks at a time.

In addition to floods, the area is susceptible to hail storms (which destroy crops), and electric storms. Lightning is a considerable hazard to villagers who seldom have protectors for their homes and often find themselves exposed to the elements while walking.

Home | National Projects | Environmental Legislation | Policies and Action Plans | Documents

This page was last updated on 15 November 2001