Scientific registration n : 941 Symposium n : 36 Presentation: oral Yield problems in Bigombo Valley Development Project/Tanzania - the use of land evaluation on local, regional and district scale Intérêt d une évaluation des terres à différentes échelles pour résoudre les diminutions de rendement constatées dans les sols du projet de développement de la vallée du Bigombo (Tanzanie) BOJE Gerhild Institute of Soil Science, University of Bonn, Nussallee 13, 53115 Bonn/Germany 1 INTRODUCTION In Ngara district traditional rainfed food crop production is very low also due to unfavourable natural resources: a bimodal rainfall pattern with very unreliable rainy seasons, dominance of Ferralsols (FAO- UNESCO-ISRIC 1990) with low fertility and soil erosion on sloping land. In contrast the advantages of wetland cultivation are: food security by cultivation during the stable dry-season, higher yields due to greater fertility of the Gleysols and Fluvisols as well as lack of erosion on the flat valley bottom (NGARA DISTRICT COUNCIL 1997). A pilot-project was started by Rulenge diocese in Bigombo valley 1975. The local Gleysols changed considerably thereafter. After rehabilitation in 1985, the aims of Bigombo Valley Development Project were: increased food production to achieve self-sufficiency and crop diversification for improved nutrition as well as an income-generating activity for women cultivating vegetables (onions, tomatoes, cabbages). In recent years substantial yield decreases occured. The target of this land evaluation is to identify the causes. Besides the farmer`s vital interest, this information is also relevant for the Rulenge Diocese, Agriculture Development and Improvement Programme (ADIP) and the Agriculture Department of Ngara District Council. 2 LAND RESOURCES OF THE RESEARCH AREA Bigombo Valley Development Project is situated near Rulenge town in Ngara district/nw-tanzania in the border area to Rwanda and Burundi. Its location is approx. 2 44 S and 30 37 E at an altitude level of 1400 m. As physiographic unit it consists of a flat valley floor in roughly N-S orientation flanked by smooth hills on the western and steep hills on the eastern side. Geologically the precambrian basement complex prevails. The Karagwe-Ankolean formation consists of acid sedimentary and partly metamorphic rocks. Ferralsols are the most frequent upland soil types, whereas in the lowland areas Gleysols prevail. While the mean annual temperature ranges about 20 C there is a distinct difference between day- and nighttime temperatures. The highly variable amount of rainfall in the two rainy seasons reaches nearly 1000 mm/a. Flooding and excessive wetness due to poor drainage conditions characterize the hydrological situation of the valley floor which led to the formation of a papyrus swamp (VAN OORT 1987). For lowland soils, classified as Mollic Gleysols (FAO-UNESCO-ISRIC 1990) the typical profile consists of a thick (27-40 cm) greyish-black topsoil and a grey, strongly mottled (hydromorphic) 1
subsoil underlain by a weathering zone in transition to bedrock at approx. 2 m depth (see Tab. 1). Soil textures range from silty clay loam to clay (without any coarse fragments). Analysis data of a representative profile is interpreted with regard to soil fertility according to LANDON (1984): This silty clay loam soil is moderately acid and slightly limiting for sensitive crops only. The weighted average topsoil organic carbon content is still medium (as a relict of former swamp conditions) and so is the related nitrogen content. Consequently the C/N ratio is favourable. Since the organic fraction largely determines the cation exchange capacity, this topsoil has a medium cation exchange capacity. Base saturation is medium also, while among the exchangeable bases magnesium is somewhat high in relation to calcium whereas the potassium content is very low (exchangeable potassium percentage 0.5 %). Tab. 1: Soil analysis data of a representative wetland profile depth horizon sand silt clay C org N C/N cm symbol % % % % % ratio 0-5 Ap1 6.5 53.6 39.9 4.97 0.41 12.1 5-20 Ap2 6.0 53.7 40.3 4.63 0.44 10.6 20-32 Ah/Cg 6.8 52.9 40.3 1.33 0.14 9.6 32-130 Cg1 12.4 52.9 34.7 0.38 0.07 5.8 130-150 Cg2 74.6 19.4 6.0 0.05 0.00 --- 2
depth ph H 2 O Ca Mg K Na Sum CEC base saturation cm (1:2.5) cmol(+)/kg soil % 0-5 5.7 8.2 3.6 0.15 0.4 12.4 25.2 49 5-20 5.9 8.9 5.3 0.16 0.3 14.7 28.4 52 20-32 5.8 5.2 2.8 0.05 0.3 8.4 13.5 62 32-130 5.1 2.6 1.3 0.05 0.2 4.2 8.1 52 130-150 5.7 0.7 0.4 0.01 0.0 1.1 1.8 61 3 FARMER S PROBLEMS IN BIGOMBO VALLEY 3.1 Farming practices At peak time approx. 400 farming families were involved cultivating a plot of 0.1 ha each in the lowland area (commonly associated with approx. 2.5 ha upland). From July onwards beans (Phaseolus vulgaris) are grown on 98 % of the remaining cultivated area. One third of this area is intercropped with maize (Zea mays). A very small fraction (2 %) is used for growing vegetables (onions, tomatoes and cabbages). The crops receive supplementary irrigation during the dry season. After harvest the plots are left fallow. Subsequently the area gets flooded and densely overgrown by swamp vegetation again. 3.2 Constraints for wetland crop production Yield decline since 1991 mainly affecting the main crop beans already led toabandonment of plots. In workshop sessions with Bigombo valley farmers they identified the following major constraints for wetland crop production: Ineffective water management Experienced as No. 1 constraint by farmers is the inefficient water management due to deterioration of the water scheme caused by lack of maintenaince of drainage and irrigation structures. The annual necessary reclamation of the entire valley floor with only hand tools is extremely hard work. While farmers manage to clean the secondary and tertiary channels, the manual excavation of the large outside main drains is rather too difficult. This constraint was recognized already by ADIP manager VAN OORT 1987 who suggested that for this task an excavator should be rented every 3 years. Yet it was not feasible because of lack of funds: while the project administration assumed that the farmers should pay the costs for mechanical ditch cleaning (from funds generated from increased valley crop production), the farmers, sticking to traditionally low crop production remain poor and expect the project/donor to provide this service free. In consequence soil wetness increases and the lower half of the area returned to former swamp conditions. The problem of blocked drainage lines is further enhanced by the large number of farmers involved which makes organization of common work difficult, so does lack of authority in execution. In addition the irrigation system is similarly affected and farmers struggle with ineffective water distribution/water shortage towards the end of the dry season. The farmers request to re-design the whole valley scheme to a shallow channel system which they could handle themselves is contradicted by the local natural wetness conditions. This part of the valley requires strong drainage at the beginning of the dry season to enable dry-season cropping (timely planting). Soil degradation 3
Until now soil degradation refers to chemical characteristics mainly. With assumumption of C/N/P generally correlated as 100/10/1 (LANDON 1984) the main plant nutrient N- and P-contents are still sufficient. It is the third major nutrient, potassium, which first becomes a severe fertility limitation in all lowland soils, also due to leaching by artificial drainage. The pronounced K-deficiency is likely responsible for the sharp yield decrease in beans (having a 50 % higher K-requirement than maize, in HOTTINGA 1984) from 1991 onwards causing symptoms of wilting appearance and spots on the plant leafs, formerly called Usambara mottles (CIAT/FAO 1995). Fertilizer application (mineral or ash from bean residues burnt for heating) would be required but fertilizer use is uncommon among local farmers. 4 RULENGE DIOCESE, AGRICULTURAL DEVELOPMENT AND IMPROVEMENT PROGRAMME (ADIP) Presently reconstruction of the Bigombo water management scheme is the major problem and requires technical aid. In order to draw up a request to donors, the ADIP needs an evaluation of potential land suitability for crop diversification and crop production potential. The evaluation of land suitability (BOJE et al. 1997, method: SYS et al. 1991 and 1993) clearly shows that soil wetness conditions are crucial for crop production in Bigombo valley lowland area: Natural swamp conditions with flooding during part of the crop growing season and poor drainage render the soils unsuitable (order N) for cultivation. By flood control the suitability is improved to class S 3 for all relevant crops. Further water regulation to optimal drainage conditions (class imperfect during dry season) increases suitability to classes S1 and S 1/S2. Soil fertility limitations are less important and would require - besides K-fertilizer - only locally liming for acidity sensitive crops. Climatic limitations for several kinds of crops may be overcome by specially adapted varieties in future. 4
Tab. 2 : Land suitabiliy with optimal water conditions Climatic rating Landscape + soil index factor Land suitability index class beans 83 0.96 80 S 1c maize 57 0.91 52 S 2c,f tomatoes 68 0.83 56 S 2c,f (Irish) potatoes 90 0.96 86 S 1c cabbages 75 0.75 56 S 2c,f green peppers 89 0.83 74 S 2f,c onions 99 0.75 74 S 2f soy beans 80 0.96 77 S 1c sweet potatoes 55 0.97 53 S 2c For the main local food crops beans and maize the quantitative yield potential is also assessed according to the FAO-AEZ methodology (in SYS et al. 1991 and 1993). Compared to good commercial irrigated yield levels, the physically attainable yield of beans (dry) 2.7 t/ha is very good while that of maize (grain) 5.9 t/ha is moderate. This figure for maize yield potential corresponds well with the only official yield data available: with optimal management (early planting, monocropping, optimal water supply, timely weeding) Dutch agriculturists produced 4.5-6.5 t/ha on demonstration plots in 1986. At the same time on neighbouring plots average farmer s yields were only 0.9-1.3 t/ha (VAN OORT 1985). Recent farmers maize yields as discussed in the workshops are still low. Recent yield decrease in beans has unfortunately not been monitored but abandonment of plots is a clear indication that in places it fell short of even a low threshold level. Despite great efforts in project rehabilitation 1985 (by reconstruction of drainage- and irrigationinfrastructure) and start of an agricultural extension service to advise farmers, the expectations of increased and diversified crop production have not been fulfilled since. The main reason is lack of adaptation of new farming practices by local farmers with weitreichende consequences as described in section 3. As the traditional low level land management without inputs is unsustainable and damages the soil resource by depletion, HOTTINGA s prognosis (1984) that in this case shifting valley cultivation would develop, is presently coming true in Bigombo valley. 5
5 AGRICULTURE DEPARTMENT/NGARA DISTRICT COUNCIL: NGARA DISTRICT RURAL DEVELOPMENT PROJECT Future land use planning for wetland areas which occupy 10.6 % of the district area demands soil data from long-term observation. In case of Bigombo Valley Development Project soil data from 1973/before project implementation until 1996/abandonment of plots exists. About 20 years of dry-season agriculture have changed the former swamp soils with an organic surface layer (maybe even Histosols) considerably. Drainage and losses by annual burning have resulted in a dramatic mineralization of organic matter, diminishing the topsoil (0-30 cm weighted average values) organic carbon content from initially 34.1 % (in 1973; 4.7 % in 1984) to 3.6 % in 1996. Simultaneously the cation exchange capacity decreased within the last 12 years by 25 % (from 29.2 cmol(+)/kg soil in 1984 to 21.9 cmol(+)/kg soil) but due to the amount and mixed mineralogy of the clay fraction it will not become a fertility limitation. The soil reaction remained stable over the years as did the sum of basic cations. Regarding diminishing CEC there is a relative increase of base saturation from 38 % in 1984 to 55 % in 1996. According to the FAO legend this changes Umbric into Mollic Gleysols (FAO-UNESCO-ISRIC1990). For the major crop nutrients nitrogen and phosphorus, HOTTINGA (1984) calculated the critical level of supply for annual maize and bean cultivation to be reached at an organic carbon content of 2.6 % which is presently not yet attained. Furthermore the N-content is increased by prolonged bean cultivation by 20 % (from 0.280 cmol(+)/kg soil in 1984). VAN OORT s recommendation (1987) of N- and P-fertilizer application for plots which have been cultivated longer than 5 years was contradicted by a slowing down of the soil degradation process by: - stable soil acidity preventing a faster mineralization of organic matter, - low level crop production and single cropping, - regeneration during the wet fallow period (reduced decomposition and high production of organic matter (approx. 50 t fresh swamp vegetation/ha/a). The first soil fertility limitation is potassium deficiency for which the reserve was exhausted after approx. 15 years of dry-season wetland cultivation. The consequences of yield decrease are either abandonment of the site (shifting valley cultivation) or requirement of higher level management using fertilizer input to replenish nutrient losses. After all, soil changes are only one factor causing yield problems. While the Bigombo valley site still has a high potential suitablility for beans, maize and vegetables, the related crop production potential cannot be attained due to low level land management and its major constraints in inefficient water management, unsuitable crop production methods and continuous soil depletion. For proper utilization of the wetland resource in NW-Tansania it is recommended to consider the longterm experience in the nearby neighbouring countries Rwanda and Burundi, which are much more advanced with regard to sustainable land management in wetland agriculture. REFERENCES 1. BOJE, G., RUECKER, G., SENZIGE, S. and A. SKOWRONEK (1997): Land Suitability for Crop Diversification and Yield Potential of a Drained Swamp Area in NW-Tanzania. (Publication in 1st issue 1998 by Der Tropenlandwirt /Journal of Agriculture in the Tropics and Subtropics, Witzenhausen, Germany). 2. CIAT/FAO (1995): Bridging the Research - Farmer Gap: experience with on-farm research on beans in Tanzania. 4th Regional Bean Workshop, Potchefstroom/S-Africa 1995. CIAT African Workshop Series No. 31. 3. FAO-UNESCO-ISRIC (1990): Soil map of the world. Revised Legend. Reprinted with corrections. World Soil Resources Report No. 60. FAO, Rom, 119 pp. 6
4. HOTTINGA, F. (1984): Bigombo Valley Project, report of a practical year. Rulenge/Tanzania, Unpublished. 120 pp. 5. LANDON, J.R. (Ed)(1984): Tropical Soil Manual. Booker Agric. Internat. Ltd., New York. 450 pp. 6. NGARA DISTRICT COUNCIL/TANZANIA, Ngara District Rural Development Programme (1997): Land Resources Inventory of Ngara District. Vol. I Main Report 56 pp., Vol. II Technical Annex 130 pp., 7 maps., ETC East Africa, Ngara. 7. SYS, C., VAN RANST, E., DEBAVEYE, J. and F. BEERNAERT (1991 and 1993): Land Evaluation I, II, III. International Training Center for post-graduate Soil Scientists, University Ghent, Agric. Publ. 7, General Administration for Development Cooperation Brussels, 273 pp., 247 pp., 199 pp. 8. VAN OORT, W. (1985): Rulenge Diocese Valley Development Project - Research, Observations & Recommendations. Rulenge/Tanzania. Unpublished. 58 pp. Key words: Tanzania, wetland, participatory appraisal, land evaluation. Mots clés : Tanzanie, milieu hydromorphe, évaluation concertée, évaluation des terres 7