by
G.W. Coulter
Department of Ichthyology and Fisheries Science
Rhodes University, P.O. Box 94, Grahamstown 6140, R.S.A.
Lake Tanganyika pelagic fishes are a major resource, but they are subject to wide variation in abundance. Such fluctuations are typical of clupeoid fishes and are due to fluctuations in recruitment which are independent of stock size. Fishing pressure has reduced the population size of large predators but the population of L. stappersii has increased. Predation by L. stappersii stongly influences purse seine catches of Stolothrissa but is probably less important in the inshore traditional fishery. There is a marked seasonality in the vertical mixing of the lake and thus in plankton and fish production. Nutrient replenishment (and thus production) is greater in the south of the lake and occurs earlier there than in the north. Stolothrissa appears not to migrate much in the lake. The apparent inverse relationship between L. stappersii and Stolothrissa abundance may provide material for making catch predictions.
The pelagic fishes of Lake Tanganyika represent a major resource, but the exploited stocks appear prone to wide annual variations in abundance which could impose severe restraints upon development and management. This paper examines some of the causes underlying stock fluctuations, in particular those determined by physical and chemical features of the lake environment.
Our knowledge of the life history and population ecology of the Tanganyika clupeids is still very limited. They seem, however, to exhibt many characteristics of clupeoids observed elsewhere. It may, therefore, be useful to briefly review some of the general characteristics of clupeoids to indicate what we might expect to find in Lake Tanganyika.
In general, clupeoid populations, which characteristically have a short life cycle and high replacement rate, show an inherent tendency to fluctuate in abundance, even in the absence of fishing. They tend to respond to short-term changes in the environment that affect recruitment (usually a matter of larval survival) and feeding. A prominent characteristic of clupeoid populations is, therefore, a wide variation in recruitment which, over wide ranges, is completely independent of size of spawning stocks (Murphy, 1977).
It has been shown in many cases that the size at maturity, fecundity of adults, and survival of larvae are dependent on food availability. For larval survival, the timing of the availability of correct sizes of food particles appears to be critical. To compensate for variability in abundance of suitable food, these fish are commonly serial spawners.
A major concern with many clupeoid fisheries is that they are potentially vulnerable to recruitment failure. Adult stocks fluctuate despite the fact that recruitment is generally not dependant on the numbers of spawners, and that populations are resilient to heavy fishing which often takes a high proportion of young. Experience of stocks with such characteristics indicates that moderate exploitation rates will probably cause increasing oscillation, but when exploitation takes a major part then the relative magnitude of oscillations tends to diminish (Ricker, 1975).
Another effect is from the altered levels of natural predation caused by fishing. Clupeoid fisheries are often associated with large by-catches of predators which are more susceptible to fishing pressure. Relaxation of predation then causes changes in the prey populations. By removing a large part of natural predation, the factors controlling the maximum size of the prey population will shift from limitation by cropping towards limitation by resources (i.e. availability of planktonic food). A clupeoid population might then exhibit the characteristics of resource-limited systems. However, when that state is reached, density independent models would not apply.
Prior to commercial fishing, which began 25–30 years ago, the pelagic fish populations consisted of approximately equal biomasses of small clupeid species (Stolothrissa tanganicae and Limnothrissa miodon), and 4 species of much larger Lates predators. The clupeids are short-lived (about 1 year), highly fecund and highly productive (P/B ratio about 4). Three of the Lates species are long-lived (> 10 years), and L. stappersi lives < 5 years (Coulter, 1981).
Fishing pressure decreased the abundance of large predators, resulting in an increase in the clupeids. (After 7 years initial fishing in Zambia, Lates catches were less than half of initial catches, whereas clupeid catches had increased two-fold; Coulter, 1970). The smallest Lates species, L. stappersi has increased to become an important predator under exploited conditions, and it shows cycles of abundance of several years. Year to year abundance of the dominant clupeid Stolothrissa in the purse-seine catch now oscillates markedly, apparently in relation to cycles in the L. stappersi population (Pearce, 1985; Roest, 1985).
In the natural state, the effect of oscillations in clupeid abundance upon total pelagic fish biomass would tend to be damped because of the presence of a large biomass of long-lived predators. When most of the long-lived element among the predators is removed by fishing, natural oscillations in clupeid biomass would be expected to become more apparent and should also increase in absolute terms as clupeid abundance increased. Since the fishery is now largely dependent upon Stolothrissa, which is virtually an annual fish, it has become vulnerable to factors that annually influence Stolothrissa production. These factors include some that are influenced by natural fluctations in the lake environment and are therefore not directly attributable (i.e. in the same year) to fishing.
Prior to recruitment, clupeid abundance is probably determined chiefly by larval and fry survival depending on the regulation of their plankton food supply. Plankton production appears to depend very largely upon nutrients that are made available by vertical mixing with the nutrient-rich hypolimnion, and by regeneration in the upper oxygenated layer of the water column. Fairly direct relationships should exist between nutrient supply mechanisms, plankton production, and the production of the planktivorous clupeids. Nevertheless, annual variation in survival may be considerable, as is usual in clupeids.
Once recruitment to the traditional or purse-seine fisheries takes place, mortality will be determined to some extent by fishing. Purse-seines catch mainly mature fish, but traditional methods catch a large but unknown proportion of immature fish. According to Roest (1983), 11% of the total catch is made by purse-seines and the remainder by small-scale and traditional effort. However, we should note that even in areas that are intensively fished (by purse-seine), natural predation accounts for most clupeid mortality. For example, in Burundi, Henderson et al (1973), concluded that nearly 80% of the variation in clupeid abundance was due to variance in the Lates species, while purse-seine fishing effort and clupied abundance were not significantly correlated.
Abundance of Stolothrissa in the purse seine catch now appears to be strongly influenced by cyclical changes in natural predation (Pearce, 1985; Roest, 1985). This is not necessarily also so in the traditional fishery, which is larger but operates in inshore regions catching mostly immature Stolothrissa and Limnothrissa. Predation pressure may be less there, and environmentally induced factors more important in determining survival and population biomass. Unfortunately, no comprehensive catch per unit effort data are available for the traditional fishery to show whether oscillations of a similar nature to those in offshore stocks occur.
There is a warm wet season in September-April when winds are mainly light, and a dry cool season of strong southerly winds in May-August. In the wet season the lake is well stratified, with a strongly defined thermocline at about 50m. In the dry season, the south wind cools the epilimnion, mainly by its evaporative effect, causing vertical mixing. The thermocline tilts downward towards the north. Mixing probably occurs to at least 150m in the southern basin, and the hypolimnion upwells at the south end. Warming takes place when the wind ceases, and stratification is restored in September-October, but the thermocline remains weak and superficial in the north basin until October-November.
The hypolimnion is rich in plant nutrients, and vertical mixing involving the upper hypolimnion is crucial to the supply of nutrients to the surface for phytoplankton production. The relative effectiveness of various kinds of mixing mechanisms have been investigated (Coulter, in prep). It was concluded that the thermocline acts as a diffusion floor during most of the wet season, but vertical transfer occurs as the thermocline breaks down in the south basin during the dry season, and by diffusion across the weak thermocline in the north basin in September-November. Application of a nutrient recharge model to south basin conditions indicates considerable nutrient replenishment of surface waters there in the dry season, but later diffusive transfer in the north basin is probably of much less magnitude.
The models of vertical mixing and nutrient recharge are consistent with plankton and fisheries data, in that phytoplankton maxima correspond to the timing of vertical transfer which is earlier at the south than at the north. Also clupeid catch maxima follow those of the plankton, and there is similarly a lag between the south and north (Coulter, 1977).
Figure 1 indicates the timing of clupeid peak catches and their relative magnitudes at each end of the lake, using the average purse-seine catch per month over 9 years in Burundi (1964–72) and 6 years in Zambia (1967–72). The overall catch ratio was 2.5 at the south to 1 at the north.
Figure 2 illustrates Stolothrissa modal length frequencies in the catches of a purse-seine made at a number of points along the Tanzania sector in October 1974 (based on data in Chapman, Bayona and Ellis, 1974). It indicates that Stolothrissa has a different growth pattern along the longitudinal axis of the lake at this time. The modes are probably those of the dominant annual cohort, and it would seem from the decrease in length towards the north that the cohort appears earlier at the south. These size distributions along the lake axis in October are consistent with usual Stolothriss size modes in the September catch maximum in Zambia and the November-December maximum in Burundi.
From the foregoing, the following conclusions are indicated:
Strong seasonality in clupeid catches may be expected. Annual Stolothrissa production will vary in different parts of the lake in terms of timing, relative magnitude and location, following a predictable pattern. The clupeids, or at least Stolothrissa, probably do not migrate much in the lake and consist of stocks of intergrading size structure. The south basin is more productive than the north. Stolothriss catch levels in any year will depend partly on the conditions in the lake environment that determine recruitment success and the production of planktonic food, and partly on the relative abundance of predators (chiefly L. stappersi). In the case of the former, the magnitude of vertical mixing between May and August is probably crucial. Instability arising from the annual turnover of the Stolothrissa population and from its cyclical interaction with L. stappersi will be expressed in oscillations in the total catch.
It might be speculated that material for catch prediction is provided by the apparent inverse quantitative relationship between Stolothrissa and L. stappersi and the regular periodicity of their oscillations. Catch predictions might also be made several months in advance of seasonal maxima by surveys of young of the year, or perhaps more readily by correlations with environmental parameters, especially in the period May-August.
Chapman, D.W., J. Bayona and C. Ellis. 1974 Preliminary analysis of test fishing and limnological sampling in Tanzanian waters of Lake Tanganyika. UNDP/FAO Project FI: DP/URT/71/012, Working paper 12.
Coulter, G.W. 1970 Population changes within a group of fish species in lake Tanganyika following their exploitation. J.Fish Biol., 2:329–53
Coulter, G.W., 1977 Approaches to estimating fish biomass and potential yield in Lake Tanganyika. J.Fish Biol., 11:393–408
Coulter, G.W., 1981 Biomass, production and potential yield of the Lake Tanganyika pelagic fish community. Trans.Am.Fish. Soc., 110:325–35
Henderson, H.F., G.W. Coulter, M.J. Mann and G. Wetherall. 1972 Predator-prey relationships and the pelagic fishery potential in Lake Tanganyika with particular reference to the purse-seine fishery of Burundi. Draft ms. on file with FAO Fisheries Department, Rome. 43p.
Murphy, G.I. 1977 Clupeoids, p. 283–308 In: J.A. Gulland (ed.) Fish Population Dynamics. London, John Wiley.
Pearce, M.J. 1985 Some effects of Lates species on pelagic and demersal fish in Zambian waters of Lake Tanganyika. FAO/CIFA Workshop Document SAWG/85/WP2.
Ricker, W.E. 1975 Computation and Interpretation of Biological Statistics of Fish Populations. Bull.Fish.Res. Board.Can., 191:382p.
Roest, F.C. 1985 Predator-prey relations in northern Tanganyika and fluctuations in the pelagic fish stocks. FAO/CIFA Workshop Document SAWG/85/WP1.
Fig.1 Mean monthly purse seine catches of clupeids from the northern (Burundi) and southern (Zambia) regions of Lake Tanganyika. (Burundi data from 1964 to 1972; Zambia data from 1967 to 1972)
Fig.2 Modal length frequencies of Stolothrissa tanganicae sampled from purse seine catches along the Tanzanian coastline in October 1974. (Data from Chapman, Bayona and Ellis, 1974)
