With the closure in December 1965 of an impressive rockfill dam on the upper Pangani River some 50 km south of Moshi, the filling of Nyumba ya Mungu (NYM) reservoir began. As part of the Pangani Basin Development Plan, its primary purpose was to provide hydro electric power, store water and, at the same time, opportunity for irrigation and fisheries development. Taking about two and a half Years to fill and with an area of 180km2 at top water level, in the late 1960,s NYM was East Africa’s largest man-made river lake. Notwithstanding the significant role it has played in other areas, it was the fisheries function for which the reservoir was to become internationally renowned. NYM became a spectacularly good example of the positive fishery potential presented by a new tropical river lake.

Initially very fertile and with high inputs of solar energy, NYM exhibited a prolific production of microscopic life offering rich supplies of primary foods for any fishes able to take advantage of them, and the extensive shallows of the new lake environment. Now, whereas the Pangani basin has a rather restricted fish fauna overall, it contains four types of endemic tilapias and NUM was fortunate in harbouring at least one and probably two of them since its inception. Essentially small particle feeders, tilapias graze on algae, bacteria and fine detritus. In rivers tilapias spawn in sheltered backwaters before releasing swarms of young to grow on floodplain shallows. They are thus well suited to habitats offered by river lakes and, and, as familiar and widely accepted food fishes, can provide an ideal fishery resource. By the end of the 1960s a flourishing Tilapia fishery was established at NYM. News of profitable fishing spread rapidly, and fishermen were drawn to the lake from many parts of East Africa, bringing their skills, gear and boats. At its peak an estimated 3,500 fishermen were active on the reservoir amongst a population of 25,000 who settled in 26 ‘fish-rush1 villages around its perimeter. Computed yields for 1970 reached an incredible 28,500 tonnes, about 1,900 kg/ha, an order of magnitude elsewhere to be expected only from managed fish ponds. Gill nets were the primary gear since drowned scrub and woodland prevented a widespread use of beach seines. Catches of large fish were sold either to outside traders in fresh and iced fish or to locals who processed them before resale to fishmongers.

During a detailed study at NYM in 1974, it was clear to me that the boom was over. In part this was due to a natural ecological phenomenon, namely the exhaustion of sequential production peaks amongst the biological components of the sunny-side or grazing pathways of energy flow, which are initiated when a river is converted to a lake and accompanied by nutrient releases from hitherto unflooded land. At the same time there were other contributory factors – the extraordinarily high fishing pressure exerted by man and the birds roosting among emergent branches of drowned trees; a massive encroachment of the shallow end by bulrush swamp; and fully-operational drawdowns superimposed on the seasonal drydowns, to produce an overall drop in lake levels. Tilapias are resilient fishes however which, when confronted by environmental pressures, respond by breeding earlier in life and at smaller sizes, thereby maintaining population numbers. In 1974, although large tilapias up to 2kg and 50cm were still caught, 64 per cent of fishermen’s landings were less than 20cm in length. Interestingly two non-indigenous tilapias, a legacy of previous stocking in the Pangani basin, had also made an appearance. The commonest, originally from Lake Victoria, inhabits open water, feeding on plankton, a life style complimentary to that of the inshore dwelling endemic species which browse the algal films on the submerged surfaces of grasses, drowned scrub and trees.

Between 1974 and 1983 annual yields of 2,000 to 5,000 tonnes had been estimated, and during field work at NYM in 1984 Dr. L Nhwani of the Tanzanian Fisheries Research Institute confirmed the remarkable observation that the catch had been dominated by the ‘Victoria’ species. This remained the situation when I visited the reservoir in late 1994. Both experimental and fishermen’s catches comprised mostly very small fish. 10-14cm long, many sexually mature; 80 per cent belonged to the ‘lake1 species. This switch from ‘Pangani’ to the ‘Victoria’ tilapia is difficult to account for, although I had noticed similar events much earlier, in small Tanzanian dams. At NYM it could be that, as drowned woodland rotted and stumps were removed, protected feeding grounds for the endemic tilapia were removed. Moreover, with the obstructions cleared, seines had become a major fishing gear since they could now be shot way out in open water and then pulled safely ashore. The latest statistics indicate that up to 1,000 fishermen remove about 3,000 tonnes per year, more than twice the preimpoundment prediction I made in 1965; this may be an underestimate, judged by the baskets of fish seen daily in Moshi Market. Larger surviving riparian villages now have a permanent air – with schools, dispensaries, shops, bars, netball pitches. Water birds abound and crocodiles are rare. The endemic tilapias of Lake Victoria, as a result of overfishing and the spread of Nile-perch, are now rare. Earlier consignments of at least one of them were established in Government ponds at Malya, Iringa and Korogwe and introduced into reservoirs during the 1950s and 60s. It occurs to me that if or when the Nile-perch population and fishery eventual crashes, a ‘many-generations-on’ reservoir stock could provide the source for re-introduction into the Great Lake.
Roland Bailey