The limnological studies completed and underway have been undertaken primarily to provide a base for developing the potential of the fishery of Kainji Lake. Limnological studies of the sort so far accomplished should be regarded as sources of a tentative hypothesis in considering fishery problems rather than a firm foundation for explaining the present and future condition of the fishery. In particular it must be borne in mind that the knowledge of the physiology and behaviour of freshwater tropical fishes is generally rather more poorly known than that for temperate and marine species and it is thus difficult at present to anticipate the responses of the local fauna to such conditions as low oxygen concentration, thermal gradients and similar physical characteristics of the water. Nevertheless it seems useful to suggest some of the more salient implications of the work reported to the problems of fishery development as a guide to further work at Kainji.
The potential catch or yield of fish from any body of water is fundamentally limited by the supply of food available to the fish. This is in turn limited by the quantity of inorganic nutrients in the lake and/or influx of organic and/or inorganic materials to the lake from its tributaries. While there is considerable variability both in the efficacy with which these materials are utilized in various lakes and in the components of the total fish production which are harvestable, a rough estimate of potential fish production and of fish yield is possible by using well established coefficients of transfer of food from the primary photosynthetic producers through generalized food chains to the fishes.
In order to obtain such an estimate the following assumptions were made:
Tropical lakes with surface temperatures in the range 20 to 30 C have an expected gross rate of primary production in the range 2–6 g Carbon/m2/day (Tailing, 1965); Kainji Lake has an expected value of 2–3 g Carbon/m2/day (Section 5.2.2).
Net production is about half of gross.
Between 5 and 15 percent of the net primary production may be transferred to secondary production, i.e., herbivore production.
Further transfers along food chains occur with a similar range of efficiency though these are expected to be of somewhat greater efficiency owing to higher utilization rates of animal food,
Harvesting of the fishes of Kainji will continue to be divided more or less equally among herbivorous, microphagous (eating small animals) and predaceous species.
That the Kainji fishery is reasonably efficient in harvesting a variety of fishes, cropping from 20 to 30 percent of the annual production of the fish community.
Using these assumptions1 and standard conversion factors from carbon to dry weight (x2) and dry weight to fresh weight (x5 for fishes), the tropical lakes above should yield fish as fresh weight of from 25 to 750 kg/ha/year, while Kainji might range from 25 to perhaps 450 kg/ha/year. Comparison of these figures with catch data from other tropical lakes suggests that these figures are somewhat optimistic as they tend to fall in the lower part of the range calculated.
It is worth noting that by very approximate calculation the zooplankton standing crop in Kainji Lake may be about 10 kg/ha (Cladocera, Copepoda, Rotifera), As only the uppermost 5 m are regularly sampled, this probably underestimates the actual standing crop somewhat. As the zooplankton produce several generations per year, the production of the zooplankton may be from 10 to 20 times the standing crop (Mann, 1966) suggesting 100 to 200 kg/ha/year for the production of the zooplankton as actual weight or 20 to 40 kg/ha as dry weight. This figure is very low compared with the estimates of herbivore production as obtained from the primary production figures (400 to 2 000 for Kainji), perhaps indicating that the planktonic community is of less importance in the overall biological economy of Kainji Lake than suggested in Chapter 5.
Studies by Ryder (1965), Jenkins (1967) and Regier (unpublished) suggest that the total dissolved solids and the mean depth of a lake are the most important limnological indices of potential and actual catch of fish from a freshwater lake. Data for a few lakes in Africa have been assembled by Regier, Ryder and Cordone (unpublished). Assuming that Kainji belongs to the same family of lakes as those examined, the figures for Kainji Lake (Sections 4.1.2 and 4.5.4) suggest a potential catch of around 15 to 30 kg/ha or equal to the lowest estimate from section 6.1.1. The morphoedaphic estimate, however, is based on actual catches assuming that these are close to potential.
While the above estimates may be useful as order-of-magnitude estimates in connexion with planning, they are of very low precision and provide little guidance for evaluating management strategy. It should be noted that the shallow lakes of the interior delta of the Niger, Lakes Faguibine and Do-Niangaye, yield about 45 to 65 kg/ha (Joint Consultants' Report, Vol. 6, Part 8), While these lakes presumably are both higher in total dissolved solids and lesser in mean depth, thus presumably rather more productive than Kainji, many other factors also influencing catch are probably more similar to Kainji than Kainji is to the other lakes used for reference in the above calculations.
By far the greatest area (69 percent) of the lake falls in the middle section where the Niger is split into two channels. As this portion is also shallow, has a large fraction of drowned trees and Echinchloea marsh, and the greatest expanse drawdown land suitable for agriculture, it seems very likely that more than 75 percent of the annual potential catch of fish will come from this central basin. While the catch survey now underway shows that already 60 percent of the canoes engaged in fishing are on this portion of the lake, it is also the area farthest from the main administrative, marketing and distribution centres at Yelwa and Kainji.
Owing to the short period of stratification and relatively little enrichment of deep water, upwelling at the downwind margins of the lake is unlikely to be significant in the distribution of production in the lake, even though such movements of the water should be appreciable.
Much greater significance is likely to follow from concentrations of nutrients in the bays of the Swashi, Kpan, and other rainy season tributaries to the lake. These are also likely to be important migratory routes for flood-spawning fishes.
Stirring of the bottom over the former Foge Island during low water when storm winds are frequent (Section 4.5.3) is another likely source of enriched food supplies, or at least, greater vulnerability of bottom organisms.
It was noted in Section 5.2.3 that small clupeid fishes seem to be distributed throughout the lake, concentrating during daylight hours at depths to 20 metres. Numerous echoes of large fish have been noted at depths near this layer. Though these echoes have not been identified there is reason to believe that they are of adult Lates, both as they show morphological adaptation to predation at low light intensities (a reflective layer in the retina of the eye) and as they are reported to be captured most successfully at such depths in other lakes during daylight (Lake Rudolf and Lake Tanganyika). While such species are better fished at night commercially, deep trolling for Lates and and other species may become a valuable adjunct to the development of the game reserve and tourism even though rather special boats and equipment are required.
Stratification of the lake from February to May is likely to impose additional depth restrictions on fish distribution owing to the poor oxygen conditions in the layer below the thermocline. While a number of the species in the lake are adapted to low oxygen through supplementary air-breathing, these species tend to be inshore or surface-active species.
Many of the species of the Kainji Lake appear to migrate into flooded areas for spawning. A notable feature of Kainji Lake is the wide dispersion of the flood period over time. Plow in the small lateral streams begins to rise in May and June soon after the start of the rains. It is also noted that catches of tiger fish (Hydrocynus) below the spillway of the dam increase at about this time. The main flood begins to rise in August and it is likely that the majority of the flood spawners of the river started migrations toward the latter part of August and September. The black flood of the Niger maintained high water levels in the river until April, while the local streams drop in December. The lake level is delayed relative to the river cycle by about one month. As there is little information available concerning the stimuli which initiate spawning behaviour in these species, it may be expected that several generations may be required to accomplish readjustment of the timing of spawning to the somewhat altered flood regime.
In comparing the seasonal characteristics of the river and of the lake, it appears that the seasonal cycle in the lake is rather more complex than in the river, particularly with respect to temperature. There appears to have been only a single short cool season in the river, beginning in mid-December when temperatures dropped rapidly from around 30°C to 22°C. The higher temperatures were reached again by February, and remain more or less at this level for the rest of the year. While the temperatures at the surface of the lake remain high except during the same period, temperatures below one metre show a double seasonal cycle (Fig. 12). Most other factors show a single seasonal cycle. The combined effects of the numerous factors affecting growth of fishes may, however, lead to a more complex pattern of growth of the fishes owing to distinct differences in phase among the annual cycles of water level, illumination, nutrients, stratification, and the several floods. Thus, while conditions in the lake have been described in this paper as highly seasonal for a tropical system, it would be well to reserve judgement regarding seasonality of fish growth until more direct information is available.
It is usually assumed that the visibility of fishing gear to fishes can be an important factor in gear efficiency. The local fisherman are quite aware of this and anxiously await the onset of the turbidity of the white flood. The change in underwater visibility is quite pronounced in the lake, varying by a factor of about 10 (Section 4.4). Turbidity, which increasing scattering of light, is more effective in reducing underwater visibility than light absorption reduces the intensity of light. This is particularly the case where image resolution is necessary as in distinguishing the netting of large mesh gill nets.
Net colour is also thought to be significant although the evidence is less certain that colour plays an important role. In this case, it is necessary to match the net colour to the apparent colour of the background. The latter may depend very much on the angle of view. It is worth noting that both the intensity and the colour of the underwater light is markedly altered by the white flood, making a red net appear lighter and a blue net darker than would be the case when the turbidity is reduced.
It is also interesting to note that red seems to be a frequent “signal” colour among the fishes of Kainji. Particularly noteworthy is the similar colouration of species of Alestes, Citharinus, and Hydrocynus, all of which are more or less silvered but counter-shaded (dark above) with a patch of red along the ventral surfaces and the lower portion of the caudal fin. Apparent dominance of red pigment to dominance of red light in the Niger flood waters would appear to be connected.
In projecting the present productivity of Kainji Lake into the future, the most likely source of gradual change is change in the nutrient input to the lake with changes in agricultural practice in the drainage basin. With an increase in the use of nitrate and phosphate fertilizer, the most deficient nutrients in the local soils, an increase in these nutrients may be expected in the lake. These are also the nutrients that appear to be most critical to the productivity of the lake itself. Such development will almost certainly reduce the amount of dry season burning, an uncertain factor in the present productivity of the lake (Section 4.5.7). There is little doubt, however, that the overall effect will be to increase the nitrate concentrations, while phosphate will increase more slowly (phosphate tends to be more tightly bound to the soil). Owing to the low level of nutrients in Kainji Lake, such development may be expected to increase the fish production of the lake significantly though such increases may be accompanied by changes in the species composition.
It is also likely that such development may increase the amount of soil transported by runoff with the effect of increasing the turbidity of the local floods. The latter effect is difficult to evaluate, but is less likely to affect the productivity of the lake itself.
The present yield of Kainji Lake is likely to be in the neighbourhood of 20 kg/ha/annum, with most of the total catch obtained from the large central basin of the lake. This estimate is preliminary and of only temporary utility. Much remains to be learned about the relation between the physical and biological character of the lake and optimal fishing strategies.
