Historically, there seems to be almost no verifiable or scientific information on the cultivation and problems of bananas in the Kagera Region of north-western Tanzania; although there is more in the literature from the adjacent banana areas across the border in Uganda.

A combination study of oral history traditions and archaeological excavations by SCHMIDT (1978) has shown a long history of stable settlement in Kagera. Sites of early Iron Age workings date back to 450-600 BC but the establishment of unbroken settlement is more likely to date to the time of the King known as Rugamora Mahe, around 1650-1700 AD. The oral tradition suggests that food resources were inadequate before his era but he enforced control over the iron-working and re-established forests, thus restoring productivity. The legends are somewhat confusing but he is associated consistently with both food bananas (ebitoke) and ripe juice bananas.

BAKER & SIMMONDS (1951) visited East Africa, including Bukoba, in 1948 and describe the crop as staple and magnificently cultivated. Their description reads as follows:

"Nevertheless, the importance of the banana as a food crop is, or was, very great. Indeed it seems to have been staple wherever rainfall and soils permit, and in the last analysis it is apparently only rainfall that matters, for at Bukoba excellent bananas are, by assiduous application of household refuse, grown on poor, leached sand among boulders. This is remarkable for it implies the utter dependence of many native peoples, spread over a great area, on an alien crop that can be propagated and disseminated only by bulky and delicate planting material".

Apparently, BAKER & SIMMONDS observed no serious pest or disease problems; though they briefly describe the banana weevil, Cosmopolites sordidus (Germar), as the only important insect pest of the crop in East Africa and remark that, it "appears to be a fairly recent introduction into the area and to be still spreading. They reported C. sordidus from Uganda and Eastern Tanzania (then Tanganyika), which implies that they did not see it in Kagera. Some years earlier, HARRIS (1943) reported that bananas at Amani, in eastern Tanzania, had been infested by C. sordidus for over twenty years and in 1942 the infestation was found to have extended to Lushoto, 50 km away. He went on; "This weevil destroyed bananas on a small island in the south of Lake Victoria in 1940 and was subsequently found in Bukoba near the Uganda border, but is not known to occur elsewhere in Tanganyika".

Various reports by entomologists working in Uganda can be found in volumes of the Review of Applied Entomology, Series A. For instance, GOWDEY (1920) reported that C. sordidus had lately become rather important as a banana pest and added that, as the weevil had been collected on Bukassa Island (Sesse Group, Lake Victoria) which had been uninhabited for years, it probably was not a recent introduction. HARGREAVES (1928) reported a widespread outbreak of C. sordidus on plantain and bananas. The same author gave the first information on an attempt to introduce a Histerid predator of the banana weevil, Plaesius javanus Erichs, to Uganda from Java (HARGREAVES, 1935). This was on Kibibi Island, Lake Victoria, but in 1944 attempts to recover the predator were unsuccessful (DARLING, 1946). WHALLEY (1957) described trials of dieldrin to control banana weevils and gave recommendations for its use. As recently as 1970, the banana weevil was rated as the only major pest of bananas in Uganda (MUKASA, THOMAS, INGRAM & LEAKEY in Jameson, 1970).

Thus, from Kagera Region, there is only the indication by HARRIS (1943) that banana weevil was found near Bukoba and the description by BAKER & SIMMONDS (1951), which suggests that there were no problems affecting banana yields in the area. At some point in the decade or so following the 1957 recommendation by WHALLEY for use of dieldrin to control banana weevil in Uganda, the use of that insecticide was promoted in Kagera Region. Local farmers testify that for a few years weevils were low in numbers but that this was followed by a serious decline in the vigour of the banana plants. Sound quantitative records of this decline seem not to be available but NGUNDO & TAYLOR (1973) report their being requested to investigate a decline in banana production in a varietal collection maintained near Bukoba. They discovered large populations of the nematode species, Radopholus similis (Cobb) THORNE, in diseased roots of two varieties. Contrary to a later citation of their work (BUJULU et al. 1986), they did not make any comment on the state of other banana farms, or shambas, in the Bukoba area; although they deduced, from the losses caused by this nematode elsewhere in the world, "that banana production in East Africa is being reduced by R. similis". The only prior report of nematodes infesting bananas in the area appears to be that of WHITEHEAD & LEDGER (1961), who noted that Pratylenchus species were commonly found in banana gardens in East Africa affected with wilt disease. Including their report, the conspectus produced by THOMAS & TAYLOR (1968) of plant nematology in Africa south of the Sahara lists only 21 references mentioning bananas from a total of 567 references cited. None of the other references indicate any serious problems in bananas due to nematodes. A report from Uganda (McNUTT, 1974) mentions that banana weevil had long been considered the only major pest of bananas but in recent years the importance of the nematode Radopholus similis had been recognised.

The banana weevil continued to be regarded as the pest which demanded control measures and dieldrin as 2.5% dust was supplied by the Ministry of Agriculture from 1974 to 1977 (P.T. WALKER, M.J. HEBBLETHWAITE & J. BRIDGE. Project for banana pest control and improvement in Tanzania. Unpublished EEC Report to the Government of Tanzania produced by Tropical Research & Development Institute, London, 1983). The problem of decline in banana yields continued or even became exacerbated and a National Banana Research Project for Tanzania, initially based at the Tropical Pesticides Research Institute, Arusha, was started in 1978. The centre for activities was shifted to the Maruku Agricultural Research Institute, near Bukoba, in 1982 (A.S.S. MBWANA, Five Year research highlights - National banana research project, 1983-1987. Unpublished report to the Tanzania Agricultural Research Organisation, 1987). Working primarily at Maruku, between 1978 and 1981, BUJULU et al. (1986) conducted an evaluation of pesticides which involved assessing both weevil and nematode numbers and the damage that these pests were thought to be causing. The most striking results they obtained, however, show severe "blow down" in the dieldrin treated plots but the differences between treatments (dieldrin was compared with several pesticides and an untreated control) are much less marked when nematode and weevil numbers are compared with damage.

TIBAIJUKA (1979) examined smallholder agriculture in Kagera Region and concluded that the future of the banana industry was bright providing certain improvements were made; these were, the control of banana weevil, the provision to farmers of manure and the improvement of the marketing system. So, it seems that by then lack of nutrients in the soil of the banana shambas was recognised as a limiting factor. Four years later, the same author was writing: "the banana weevil and nematode problems have started a vicious circle in the farming system in the region. By decreasing yields drastically they have robbed the people of their much loved staple food and forced them to work harder but harvest less of the crop" (A.K. TIBAIJUKA Competition between bananas and coffee on smallholders farms in Tanzania: A case study of the Kagera Region pp. 89-101 in KIRKBY, R.A. &: NGENDAHAYO, D. (Eds). Banana production and research in Eastern and Central Africa. Manuscript Report, IDRC-MR114e-132 pp. Ottawa, Int. Dev. Res. Centre, 1985).

Tibaijuka is an economist and it seems that her statements were based on evidence from other sources. Nevertheless, recognition of the severity of the decline in the banana crop led to a technical survey of the crop in the region by Walker and his colleagues (P.T. WALKER et al., op. cit., 1983). They concluded that improved cultivation and reduction in weevil and nematode attack might reasonably be expected to lead to increases in yield. C. sordidus was the most common insect pest seen, although Temnoschoita nigroplagiata (Quedenfeldt) and a few other weevil species also were seen. The nematode picture, however, was unexpected; with the species Pratylenchus goodeyi SHER & ALLEN being identified as much more prevalent than Radopholus similis.

My research programme had its origins in the 1983 report by Walker and his colleagues. An adverse comment on the report has to be that the team mainly went to places within Bukoba District. Of a total of 58 visits to farms, only seven were in Muleba District, only four in Karagwe District and none in Ngara and Biharamulo. Although banana cultivation is of minor importance in most of Biharamulo, the crop in the other three is just as important as the staple crop as it is to the people of Bukoba District. The dense banana cultivations in Ngara, moreover, are isolated by some 70 to 100 km of uncultivated tree and scrub savannah from the other banana shambas of Kagera but are contiguous with those of Burundi.

In addition to their recommendations for research studies, the study team laid down a strategy for rehabilitation of the banana shambas. For severely degraded shambas, drastic action was proposed which involved - uprooting the existing stools; a break period, if possible; and replanting, using pared corms, manure and the pesticide carbofuran. However. the use of the strange-looking pared corms (that is with the outer layers of the corm removed to excise any diseased tissue) instead of the traditional planting of large sword suckers [shown right] was viewed by the local farmers with extreme disbelief. Additionally, the unfortunate results of the use of dieldrin had engendered a strong reluctance by farmers to use any pesticides on bananas.

The Kagera Region of Tanzania

href="\BANANAS2000\BANHIST.HTM"