by
N'sibula Mulimbwa
CNRS Uvira, Zaïre
c/o B.P. 254
Bujumbura
Burundi
and
Piero Mannini
UNDP/FAO IFIP Project
B.P. 1250
Bujumbura
Burundi
ABSTRACTThis paper gives preliminary results of an investigation into the population dynamics of the three most important pelagic species in the pelagic fishery in the Zaïrean sector of Lake Tanganyika. Stolothrissa tanganicae has a very rapid growth rate, a life span of one year, high mortality (Z = 9∙23 yr-1) and appears to be heavily exploited. Limnothrissa miodon has a slower growth rate, a life span of two years, a lower mortality rate (Z = 4∙67 yr-1) and is less heavily exploited. The predatory Lates stappersii has a life span of about six years and a mortality rate of about 2.09 yr-1. Condition in the two clupeids varies according to food availability and the size of the fish. Gonadosomatic indices and the recruitment patterns suggest that reproduction occurs in pulses which are often determined by hydrological conditions in the lake. All three species are caught by the inshore fishery with Limnothrissa and L. stappersii being taken before they reach sexual maturity. Although all three species have rapid growth rates and high mortality the extent to which they can withstand fishing pressure still needs to be investigated. RESUMECette communication donne les premiers résultats d'une recherche sur la dynamique démographique des trois principales espèces pélagiques des pêcheries pélagiques du secteur zaïrois du lac Tanganyika. Stolothrissa tanganicae présente une croissance très rapide, une durée de vie d'un an, une forte mortalité (Z = 9∙23 an-1) et semble être très fortement exploitée. Limnothrissa miodon présente un taux de croissance plus lent, une durée de vie de deux ans, un taux de mortalité faible (Z = 4∙67 an-1) et est moins fortement exploité. Le prédateur Lates stappersii a une durée de vie d'environ six ans et un taux de mortalité d'environ 2,09 années-1. La situation des deux clupéidés varie selon la disponibilité d'aliments et la taille du poisson. Les indices gonadosomatiques et les schémas de recrutement donnent à penser que la reproduction s'effectue par à coups souvent déterminés par la situation hydrologique du lac. Les trois espèces sont capturées par les pêcheries côtières, Limnothrissa et L. stappersii étant prises avant d'atteindre leur maturité sexuelle. Bien que les trois espèces présentent une croissance rapide et une forte mortalité, la mesure dans laquelle elles peuvent supporter la pression des activités de pêche reste à étudier. |
The clupeids Stolothrissa tanganicae and Limnothrissa miodon and the centropomid Lates stappersii are the most abundant pelagic fish species in Lake Tanganyika and are a valuable resource for the people living around the lake. Their biology has been studied in considerable detail by various authors (summarised in Coulter, 1991) but little has been done on the Zaïrean side of the lake since the early work by Marlier (1957). Recent fishery statistics indicate that the catch is declining in the Zaïrean sector of the lake (Enoki and Mambona, unpublished) which could have serious social and economic implications. However, the biological information that is needed for the development and management of this fishery is inadequate. This paper is a preliminary assessment of aspects of the population dynamics of the three important pelagic species.
Samples were collected every month during 1988 from an artisanal fishing unit. It consisted of two large wooden boats (8 m long and 1∙4 m wide) and a smaller one (4∙8 m × 1∙1 m). The three boats were attached to each other and equipped with a lift net which was 20 m deep with a mouth of 256 m2. The net mesh was 4∙4 mm (stretched) except for the lowest four metres where it was reduced to 3∙3 mm. The net was set just below the boats 30 minutes after the lamps had been lit and was drawn up 90 minutes later, i.e. two hours after lighting up. Kerosene pressure lamps were used to attract the fish; two large ones pointed straight down into the water while eight smaller ones illuminated a wider area of the lake's surface. Fishing was carried out both inshore (500 m from shore, 50 m deep) and offshore (7km from shore, ≈ 150 m deep) in the Uvira area. Sampling was done on 12–16 nights per month from a total of 187 inshore and 204 offshore hauls.
A total of 61676 Stolothrissa, 22326 Limnothrissa and 23412 L. stappersii was measured (standard length) and grouped into 2 mm, 4 mm and 10 mm length classes respectively. The parameters of the von Bertalanffy growth function (VBGF) were estimated by means of the ELEFANI computer programme (Pauly and David, 1981). Shepherd's length composition analysis (SLCA) (Shepherd, 1987) and the Projection Matrix method (Projmat) (Basson et al., 1988) were also applied to the L. stappersii data.
Total mortality (Z) was calculated by means of length-converted catch curves (Pauly, 1983) whilst natural mortality (M) was estimated by Pauly's (1980) empirical relationship between L∞, K and average surface water temperature which was assumed to be 25°C. The length-weight relationship and Fulton's condition factor index (CFI) (Ricker, 1975) were estimated from individual length and weight measurements. The gonads from 2459 female and 3078 male Stolothrissa and from 1941 female and 2887 male Limnothrissa were weighed and used for the estimation of the gonadosomatic index (GSI). The seasonal pattern of recruitment was inferred by the backward projection of each length-frequency distribution using the VBGF parameters estimated earlier by the ELEFAN routine.
L. stappersii was the most pelagic of the three species and few were caught by the inshore fishery (Fig. 1). Limnothrissa, on the other hand, was largely confined to the inshore waters whilst Stolothrissa occurred in both areas. The catches of Stolothrissa and L. stappersii were greatest during the six months from August to January (Fig. 2). This is when recruitment to the stock takes place as well as seasonal water mixing and the effects of algal blooms on the productivity of the ecosystems (Hecky, 1991).
The length-frequencies of the Stolothrissa samples were generally polymodal with little evidence of modal progression (Fig. 3). Most of the juveniles (30–40 mm long) appeared in the inshore waters whilst the bulk of the adult population was found offshore throughout the year. This seems to confirm Coulter's (1970) observation that the smallest fish are pelagic, then move into coastal waters where they are caught at about 35 mm in length and then move offshore again once they reach 50 mm in length.
The seasonal cycle of the GSI was similar in males and females in both inshore and offshore waters (Fig. 4a). The latter give the best indication of the breeding condition of the fish since they breed in open waters. The GSI was highest in the first half of the year (February-July) with a clear peak in June. The GSI decreased after that to reach its lowest level in October. This suggests that Stolothrissa breeds during the early part of the year as proposed by Poll (1953). The breeding pattern can be explained by the fact that phyto- and zooplankton is most abundant at the end of the dry season (around September) and is therefore available to the young fish.
The length-weight relationship was w = 0.3955 * 10-5SL3∙24 (n = 1099). Their condition factor varied in relation to their length and seemed to decrease as they grew up to 40 mm in length after which it begins to increase with increasing size (Fig. 5a). One explanation for this is that the growth rate of the young may be very fast and not isometric, i.e. the value of the length-weight coefficient b is significantly different from 3. Another explanation is that the smaller fish feed mainly on phytoplankton and only change to zooplankton when they reach 50 mm in length (Coulter, 1991). This transition could be marked by a change in condition. A final possibility is that the body weight was being affected by the development of the gonads.
Seasonal changes in condition were relatively small although it declined slightly between March and September (Fig. 6a). The improvement in October may have been caused plankton blooms at this time. The cause of the small increase in March is uncertain although zooplankton was more abundant in the lake at that time (Mulimbwa, unpublished).
The growth of the two cohorts found in the samples was slightly different, with the following parameters:
| L∞ (mm SL) | K (year-1) | Φ' | |
| January cohort | 108 | 1.90 | 2.35 |
| August cohort | 109 | 2.29 | 2.43 |
The differences between them had very little effect on the resultant growth curves (Fig. 7a). Both cohorts originate at a time when food was most abundant but August is at the beginning of that period and January the end of it. This might explain why the growth performance of the August cohort was better than that of the January one because they can utilise the higher plankton densities to be found from September to December (Hecky, 1991).
The average life span was about 12 months and only very occasionally were specimens older than this, i.e > 95 mm in length, found in the samples. According to Ellis (1971) the length at first maturity was about 64 mm FL in males and 75 mm FL in females which, when corrected to standard length, corresponds to an age of 5 and 6 months respectively. The rate of total mortality was 9∙23 yr-1 and the rates of natural and fishing mortality were 3∙65 and 5∙58 respectively. Since Z is approximately equal to the production/biomass ratio (Allen, 1971) it is obvious that Stolothrissa must have a high turnover. Although it is heavily-exploited these rates seem to be very high and may have been over-estimated.
The mean length of recruitment to the artisanal fishery is 40 mm and the fish are then about 3 months old and a high proportion of the fish are recruited before they are sexually mature. The pattern of recruitment indicates that there is a main pulse lasting about 5 months in which about 75% of recruitment takes place. This is supported by changes the GSI of fish caught offshore and tends to confirm Poll's (1953) suggestion that these fish have a six-month spawning season.
According to Coulter (1991) Limnothrissa spends most of its life in coastal waters and this is confirmed by the length-frequency distributions of the fish collected in this study (Fig. 9). It is only when they reach a length of 80 mm that some of them become pelagic. The reason for this is that they become piscivorous from this size and prey upon Stolothrissa which are more numerous offshore (Poll, 1953; Henderson, 1976; Bashirwa cited in Coulter, 1991).
Seasonal changes in the GSI are not as pronounced as they are in Stolothrissa (Fig. 4b). Apart from the high GSI values for inshore females February and March there was no consistent pattern which suggests that this species may breed throughout the year. A moderate increase in breeding activity may occur in the wet season (October-May) as suggested by Matthes (1965–66) and Ellis (1971). In contrast to Stolothrissa, reproduction in Limnothrissa does not seem to be linked to the seasonal cycle of water mixing and consequent plankton production. This might be explained by its wider trophic niche and less specialised feeding habits which make it less dependant on the availability of specific types of food.
The length-weight relationship was w = 0.3993 * 10-5SL3∙28 (n = 1116). Condition increased with length up to 80 mm when it began to level off (Fig. 5b). This is when they become sexually mature and begin to spawn. Condition in some of the largest individuals improves again, possibly because they have become piscivorous. Condition changes during the year, however, and seems to be highest at times of plankton abundance (Fig. 6b) which is unexpected in a fish which seems otherwise to be less dependant on seasonal cycles of plankton.
The von Bertalanffy growth parameters were L∞ = 140 mm SL and K = 1∙19 yr-1 whilst the growth performance (Φ') = 2∙36. The average longevity was about 18 months by which time the fish had grown to about 120 mm (Fig. 7b) and very few fish reached their third year of life. The rate of total mortality was estimated to be 4.67 year-1 with natural mortality and fishing mortality rates of 2∙35 and 2∙32 respectively. This gives an exploitation rate (E) of 0∙5 which means that the Limnothrissa stock is not heavily exploited at present.
The length of 50% maturity was 75 mm in both males and females (Ellis, 1971) which corresponds to an age of about 7 months. Limnothrissa recruit to the artisanal fishery when they are about 35 mm in length or 3 months old. The recruitment pattern differs from that of Stolothrissa (Fig. 8b). Recruitment took place throughout the year except for a three month period when it was very low and there were two moderate pulses separated by about 5 months. These pulses could not be confirmed by the variation of the GSI during the year (Fig. 4b).
Large specimens of L. stappersii were seldom encountered during this study because they are rarely caught in the artisanal fishery but some are taken by the industrial fishing vessels which are more effective in catching these fast swimmers. This species is pelagic throughout its life but migrates into coastal waters in its second year (Coulter, 1991). The length-frequency distributions of the Zaïrean samples suggest that there is an additional period when they come inshore (Fig. 10). In the northern waters of the lake young L. stappersii tend to remain in coastal waters until they reach a length of 65–70 mm before moving offshore. This might be explained by the fact that shallow shelving areas are more extensive in this region.
The results of the three methods that were used to estimate the growth of this species were very similar (Table 1). The smallest size at which sexually mature fish have been found is 150 mm (males) and 170 mm (females) (Chapman and van Well, 1978; Ellis, 1978). Since they recruit to the artisanal fishery at a length of 45 mm, or 3 months of age, the stock could be seriously affected by this fishery. This is important because the artisanal fishery in the Zaïrean waters, fishing within 7 km of the shore, is able to catch significant quantities of L. stappersii (Fig. 1), in contrast to the situation in the Kigoma area (Chapman and van Well, 1978). As they grow larger the fish move into deeper water where they are recruited to the pelagic fishery (Roest, 1988).
The maximum longevity appears to about six years. the total mortality rate ranges from 2∙02 to 2∙16 year-1, depending on the method used to determine growth (Table 1), with natural mortality ranging from 0∙79–0∙91 and fishing mortality from 1∙18–1∙24. The exploitation rate was therefore around 0∙58–0∙60 which means that the stock is relatively heavily exploited and vulnerable to overexploitation in the future. The mortality rate varied considerably between 1973 and 1988 and increased very slightly (Fig. 11). The catches in the industrial fishery declined during this period as well.
There was a well-defined recruitment pattern and 70% of the recruitment took place during a three-month period (Fig. 8c). This peak follows the main Stolothrissa peak by about three months and suggests a close relationship between the predatory Lates and its main prey species.
Although much remains do be done, this is the first attempt to study the biology of the pelagic fish on the Zaïrean side of the lake. All three species have complex inshore and offshore movements at various stages in their lives. All of them have rapid growth rates and high mortalities which should enable the stocks to withstand intensive fishing pressure. However, it would be unwise to make easy and optimistic assumptions about their ability to do this. The L. stappersii population may already be subject to excessive fishing pressure whilst there is some doubt about the reliability of the mortality estimates for Stolothrissa.
The reproduction of Stolothrissa and L. stappersii seem to be linked to the production dynamics of the lake which is reflected by pulses of recruitment. These relationships are not fully understood and may vary from one part of the lake to another; real understanding can only be achieved through a simultaneous lake-wide research programme.
Allen, K.R., 1971. Relation between production and biomass. J. Fish. Res. Bd. Can., 28: 1573–81.
Basson, M., A.A, Rosenberg and J.R. Beddington, 1988. The accuracy and reliability of two new methods for estimating growth parameters from length-frequency data. J. Cons. Int. Explor. Mer, 44: 277–85.
Chapman, D.W. and P. van Well, 1978. Observations on the biology of Luciolates stappersii in Lake Tanganyika (Tanzania). Trans. Am. Fish. Soc., 107: 567–73.
Coulter, G.W., 1970. Population changes within a group of fish species in Lake Tanganyika following their exploitation. J. Fish Biol., 2: 393–408.
Coulter, G.W., (ed.), 1991. Lake Tanganyika and its life. London; British Museum (Natural History) and Oxford University Press.
Ellis, C.M.A., 1971. The size at maturity and breeding seasons of sardines in southern Lake Tanganyika. Afr. J. Trop. Hydrobiol. Fish., 1: 59–66.
Ellis, C.M.A., 1978. Biology of Luciolates stappersii in Lake Tanganyika (Burundi). Trans. Am. Fish. Soc., 107: 557–66.
Hecky, R.E., 1991. The pelagic ecosystem. In G.W. Coulter (ed.), Lake Tanganyika and its life. London; British Museum (Natural History) and Oxford University Press.
Henderson, H.F., 1976. Notes on Luciolates based on a study of length-frequency diagrams from the ring-net fisheries of Lake Tanganyika. Notes on the large size of Limnothrissa in the catches of the ring-net fishery in Tanzania. FAO Rept FI:DP/URT/71/012/29: 1–6.
Marlier, G., 1957. Le Ndagala, poisson pélagique du lac Tanganyika. Bull. Agric. Congo Belge, 48: 409–22.
Moreau, J. and B. Nyakageni, 1992. Luciolates stappersii in Lake Tanganyika. Demographic status and possible recent variations assessed by length-frequency ànalysis. Hydrobiologia, 232: 57–64.
Pauly, D., 1983. Some simple methods for the assessment of tropical fish stocks. FAO Fish. Tech. Pap. 234: 52pp.
Pauly, D. and N. David, ELEFANI, a BASIC program for the objective extraction of growth parameters from length-frequency data. Meeresforschung, 28: 205–211.
Poll, M., 1953. Poissons non Cichlidae. Résultats scientifiques de l'exploration hydrobiologique du lac Tanganyika (1946–1947). Inst. Roy. Sci. Nat. Belgique, 3(5A): 1–251.
Ricker, W.E., 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Bd. Can. 191: 382pp.
Roest, F.C., 1988. Predator-prey relations in northern Lake Tanganyika and fluctuations in the pelagic fish stocks. CIFA Occas. Pap. 15: 104–29.
Shepherd, J.G., 1987. A weakly parametric method for estimating growth parameters from length composition data. In D. Pauly and G. Morgan (eds), Length-based methods in fisheries research. ICLARM Conf. Proc., 13: 353–62.
Table 1. The growth parameters and total mortality (Z) of Lates stappersii in the Zaïrean waters of Lake Tanganyika, 1988 estimated by three different methods.
| Method | L∞ (mm) | K (yr-1) | Z (yr-1) |
| ELEFAN I | 462 | 0.46 | 2.16 |
| Shepherd's (SLCA) | 480 | 0.38 | 2.02 |
| Projmat | 456 | 0.45 | 2.16 |
Figure 1. The proportion (% by weight) of Stolothrissa tanganicae, Limnothrissa miodon and Lates stappersii in (a) inshore catches and (b) offshore catches in the Zaïrean waters of Lake Tanganyika, 1988.
Figure 2. The monthly catch (kg) of pelagic fish in the inshore and offshore waters in the Zaïrean sector of Lake Tanganyika, 1988.
Figure 3. The length-frequency distributions of Stolothrissa tanganicae collected every month from January to December 1988. The shaded portion of the histogram represents fish caught inshore and the unshaded part those caught offshore.
Figure 4. Variation in the mean gonadosomatic index (GSI) in (a) Stolothrissa tanganicae and (b) Limnothrissa miodon, January to December 1988. The solid points denote offshore fish, the open points inshore ones.
Figure 5. The relationship between condition factor (K) and standard length in (a) Stolothrissa tanganicae and (b) Limnothrissa miodon.
Figure 6. Seasonal variation of condition factor (K) in Stolothrissa tanganicae and Limnothrissa miodon, January – December, 1988.
Figure 7. Growth curves of (a) Stolothrissa tanganicae (Note: the upper curve = August cohort, lower curve = January cohort), (b) Limnothrissa miodon and (c) Lates stappersii (Note: the three curves obtained with the different methods of length-frequency analysis are indicated)
Figure 8. The recruitment pattern in (a) Stolothrissa tanganicae, (b) Limnothrissa miodon and (c) Lates stappersii.
Figure 9. The length-frequency distributions of Limnothrissa miodon collected every month from January to December 1988. The shaded portion of the histogram represents fish caught inshore and the unshaded part those caught offshore.
Figure 10. The length-frequency distributions of Lates stappersii collected every month from January to December 1988. The shaded portion of the histogram represents fish caught inshore and the unshaded part those caught offshore.
Figure 11. The instantaneous rate of total mortality (Z) of Lates stappersii in northern Lake Tanganyika, 1973–1988. Drawn from data in Moreau and Nyakageni (1992) with additional data from thios study. [Editor's note. This relationship has a correlation coefficient r ≈ 0.44 which is not statistically significant]
