by
John D.R. Bayona,
Tanzania Fisheries Research Institute
P.O. Box 9750 Dar-es-Salaam,
Tanzania.
ABSTRACTLimnothrissa miodon, a clupeid which largely occupies inshore waters, is capable of moving into pelagic waters where it is prone to industrial exploitation at sizes between 60mm and 175mm in fork length. The influx of the species in the pelagic zone usually reaches a peak between October and November. Time series trends in the species abundance and distribution were evaluated based on the available catch and effort data from the industrial fisheries for the period 1974 through 1987. A low but stable abundance pattern was demonstrated despite a drastic increase in the level of exploitation. However, an exceptional increase in abundance of the species in the pelagic zone was observed between 1977 and 1982, accounting for 0.5% to 11.6% of the total industrial landings and coinciding with a drastic decline in abundance of Stolothrissa tanganicae, its competitor. Possibly this exposed L. miodon to tense predation by Lates mariae, resulting into negatively correlated oscillations in abundance of the two species. The lack of independent catch statistics for artisanal exploitation of the species and the increasing effort which generally yielded poor harvests from the lake are, among the major bottlenecks underscoring the importance of stock assessment to the optimization of exploitation of fish stocks. RESUMELimnothrissa miodon, clupéidé qui occupe principalement les eaux littorales, est capable de se déplacer vers les eaux pélagiques où, quand il atteint des tailles comprises entre 60 et 175 mm (longueur à la fourche), il fait l'objet d'une exploitation industrielle. L'arrivée de l'espèce dans la zone pélagique atteint généralement une pointe entre octobre et novembre. Les tendances qui ressortent des séries de données chronologiques concernant l'abondance et la distribution de l'espèce ont été évaluées d'après les données disponibles concernant les captures et l'effort de pêche industrielle pour la période allant de 1974 à 1987. Le schéma mis en évidence a montré une abondance faible mais stable malgré une augmentation spectaculaire du niveau de l'exploitation. Toutefois, un accroissement exceptionnel de l'abondance de l'espèce dans la zone pélagique a été observé entre 1977 et 1982, représentant de 0,5 à 11,6 pour cent du total des débarquements industriels et coïncidant avec une baisse non moins spectaculaire de l'abondance de Stolothrissa tanganicae, son concurrent. Cette situation a peut-être exposé L. miodon à une prédation intense de la part de Lates mariae qui a entraîné des oscillations opposées de l'abondance des deux espèces. L'absence de statistiques séparées des captures de l'exploitation artisanale de l'espèce et l'augmentation de l'activité de pêche, qui a généralement donné de médiocres récoltes, figurent parmi les principaux goulots d'étranglement, ce qui montre l'importance de l'évaluation des stocks pour l'optimisation de leur exploitation. |
Among the six economically important fish species which have been the focus of research in the Tanzanian sector of Lake Tanganyika, Limnothrissa miodon (Lumbo) is the least documented. The only available information on the species biology, abundance and distribution is based on the work by Ndugumbi et al. (1976), Henderson (1976) and Chapman (1976). Observations by these authors emerged from the analysis of both the species monthly length frequency and catch and effort data of the industrial fisheries, available for the years 1974 and 1975.
The information available on the species growth, abundance and distribution, permits a general comparative review among different sectors of the lake. In the Tanzanian waters, the species mainly appears in the pelagic habitat only at sizes above 60mm in fork length. Individuals of sizes smaller than 60mm are assumed to be distributed in other habitats of the lake if not the inshore area. Spawning is assumed to take place inshore and is continuous throughout the year with the maxima in December to February and August to September. This has resulted into lack of clearly defined recruitment time (Ndugumbi et al. 1976). In the Zambian sector of Lake Tanganyika, however, the species mainly occupies the inshore habitat where it is efficiently adapted to feeding on a broad range of planktonic food (Matthes, 1968; Coulter, 1991). It spawns close to the inshore area during the rainy season, November to May (Matthes, 1968; Pearce, 1985). Adult Limnothrissa of one year old and approximately 100mm in length, appear in the pelagic zone where they become piscivorous, feeding on Stolothrissa tanganicae (dagaa). This off-shore movement of large sized Limnothrissa is seemingly a lakewide phenomenon which is attributed to feeding on smaller abundant dagaa (Henderson, 1976).
In Tanzanian waters, adults of 120mm which were reported by Ndugumbi et al. (1976) to grow to a maximum size of 175mm show a higher growth rate (K=0.0765) than individuals of less than 120mm (K = 0.056). This difference is linked to a switch from other types of food to the piscivorous diet in the off-shore waters (FAO, 1978; Coulter, 1991). Apparently, this observation is different from the growth of the same species in Zambia where constant growth rate for all sizes of fish (K = 0.076) was reported (Pearce, 1985). The difference in growth rates of the species in the two sectors of the lake is an indication that the Zambian stock flourishes in a nutrient richer environment with abundant food items than is the case in Tanzania.
In Zambian waters, L. miodon is a very preponderant species, contributing equally to the clupeid or “Kapenta” catch with S. tanganicae. Pearce (1988) reported that this species seems to have sustained a stable population, free of predator-prey induced fluctuations. Although L. miodon is less abundant in Tanzania than in both Zambia and Burundi sectors of Lake Tanganyika no attempt has yet been made to study the long term spatial distribution of the species in view of the changing levels of the exploitation and predation.
The objective of this paper is to evaluate the impact of both exploitation and predator-prey interactions to abundance and distribution of L. miodon in the Tanzanian waters and to justify the need for continued stock assessment.
Estimates of catch per unit of effort (CPUE) as an index of relative abundance for L. miodon together with other species under the industrial fishery for 1974 through 1987 were developed from the landed catches and actual number of hauls made. Catch and effort data were collected by researchers and technicians at Kigoma, during the operational phase of UN/FAO Fisheries Research and Development Project (FI:DP/URT/71/012) and under the current government following the establishment of Tanzania Fisheries Research Institute (TAFIRI) in 1980. Catch data for L. miodon, S. tanganicae and Lates stappersii were collected by subsampling standard fish boxes whereas the big predators, L. mariae, L. microlepis and L. angustifrons, were enumerated and measured without any subsampling. The lengths were transformed into units of weight (kg) by use of appropriate conversion factors, well elaborated by Bayona et al. (1992).
Information on effort and which includes the number of hauls, active number of fishing nights and fishing vessels was obtained from the daily recorders' data sheets. The analysis presented in this paper is restricted to the use of number of hauls. Note that from a correlation matrix of catch, fishing nights, number of hauls and number of vessels, Bayona et al. (1992) reported a significant correlation between catch and hauls. Therefore these authors recommended the use of number of hauls as the best measure of effort in developing good indices of relative abundance. It is argued that, fish abundance may be closely related to the number of fishing trials/hauls as fishing is done at night when the day time schools of fish are already dispersed into upper layers. This assumes that the light intensity or candle power of pressure lamps for fishing vessels is or remains uniform (Bayona et al., 1992).
Both catch and effort information for 1979 are not used in evaluating trends in fish abundance. This is because of the war situation between Uganda and Tanzania as well as a cholera outbreak which dictated the closure of fishing in the lake for over seven months. The only available data for September through December 1979 coincides with a period of high fish abundance in the pelagic zone and therefore it is not representative of the actual situation during the rest of the year.
The size distribution of L. miodon in the pelagic zone is assumed to conform with the published length frequency data for 1974 and 1975 by Ndugumbi et al. (1976) and it is the only collection available on L. miodon under the industrial fishery. Nevertheless, length frequency data of the species was collected from the artisanal fishery during September 1992 as a point sample in an attempt to validate some observations by Ndugumbi et al. (1976).
Table 1 indicates time series trends for the industrial fishery. This fishery is characterized by a drastic increase in fishing effort between 1974 and 1987 followed by continuously declining catches after three years of exploitation. The number of fishing vessels increased from 1 in 1974 to 5 in 1987. Despite this small number of vessels which exploited the fishery, the effective effort in terms of number of hauls and boat nights increased sharply. Number of hauls increased from 138 in 1974 to 1417 in 1987. The number of hauls in 1980 was more than double the effort level in the previous years. This sudden increase in fishing effort (hauls) was explained by Bayona et al. (1992) as being due to skippers attempting to search for more fish, following a steady decline in industrial catches from 1977. Fig. 1 indicates a relative increase in effort between the number of hauls and boat-nights, portraying a steeper increase for hauls. Industrial catches are generally low, contributing less than 1 % to the total catch (artisanal and industrial harvests combined). The maximum of 607.7 tonnes was realized in 1976, falling to the minimum of 160.1 tonnes in 1987. Fig.2 indicates that catches for the clupeids declined almost interminably to very low levels by 1987. However, catches of Lates species have been relatively high with a peak in 1976 and followed by a gradual decline to their original level by 1987.
Total fish catches declined most sharply between 1974 and 1975, then less sharply between 1976 and 1980. Fig. 4 indicates that between 1981 and 1987 fish abundance remained more or less stable, L. stappersii being the dominant species. In the pelagic zone the abundance of L. miodon in relation to S. tanganicae is shown in Fig. 5. As shown by the figure catches of L. miodon were very low but stable between 1974 and 1976 and between 1982 through 1987. These two periods represent initial low exploitation level and a subsequent high exploitation level respectively. This observation suggests that the long term distribution in abundance of L. miodon in the pelagic habitat is not directly affected by the level of effort. However, L. miodon occurred in abundance in 1977 through 1981. Table 2 indicates that S. tanganicae which in 1974 constituted over 50% of the industrial catch, declined to a very low level in 1978 and it constituted 1.2% of the total catch. Abundance of this species has remained low and somewhat seasonal. However, catch composition for L. miodon increased from 0.5% in 1977 to 3.2% in 1980. In 1979 fishing was conducted for four months and during that period L. miodon composed 11.6% of the catch. This exceptional increase of L. miodon in the pelagic zone which coincided with a drastic decline in the abundance of S. tanganicae, is indirectly related to the exploitation level of the latter species.
Bayona et al. (1992) considered the level of fishing which increased by almost three fold between 1978 and 1980 (Fig. 1) to have been the major cause for both damping of oscillations for S. tanganicae and sustaining a prolonged decline in abundance of this species in the Tanzanian sector of the lake (Fig. 6). This is assumed to be over and above the species fluctuations which are due to environmentally induced recruitment and survival success as reported by Coulter (1976; 1988) or predator abundance (Henderson, 1976; Pearce, 1988). Therefore the paucity of S. tanganicae in the pelagic zone is likely to have reduced competition for food, thereby attracting the off-shore feeding movements of its competitor, L. miodon. Note that the diet of L. miodon is similar to that of S. tanganicae although it is more diversified to include insect larvae (Matthes, 1968). This probable cause of the species influx in the pelagic zone is irrespective of the usual movement of L. miodon to the pelagic habitat in pursuit of recruits of S. tanganicae for food as suggested by Henderson (1976).
Fig. 3 shows a trend of the seasonal distribution in abundance of L. miodon and S. tanganicae. A synchronous peak in abundance of the two species in the pelagic zone is observed between october and November. However, L. miodon remains fairly stable and less abundant during the first half of the year, compared to the second half of the year. This trend is complementary to Henderson's suggestion that adults of L. miodon increase in the pelagic zone around October-November in pursuit of S. tanganicae for nourishment.
Fig. 7 depicts somewhat usual abundance oscillations between L. miodon, the prey and L. mariae, the predator. These negatively correlated oscillations are restricted to a period of high abundance of L. miodon in the pelagic zone and during the period of low abundance of S. tanganicae (1977–1982). There is a possibility that the decline to critically low levels of abundance for S. tanganicae compelled its predators to switch to alternative abundant prey species.
Chapman et al. (1974) and Coulter (1970) noted a positive correlation between a abundance of L. mariae and S. tanganicae. This prompted Roest (1988) to suggest that L. mariae depends, almost entirely, on S. tanganicae for food. Evidence is also available that L. mariae feeds actively on both S. tanganicae and L. miodon as the reported percentage occurrence of these prey items in the stomach of L. mariae is up to the tone of 72.8% (Coulter, 1991). This species selection of its food was also considered by Coulter (1991) to be governed by prey availability rather than preference. The observations support the possibility that L. miodon was prone to tense predation by L. mariae following its increased abundance in the pelagic zone. This feeding relationship is likely to have triggered the observed predator - prey oscillations in abundance of the two species.
In the Zambian sector of the lake, fluctuations in the predator - prey abundance have been reported for S. tanganicae and its principal predator L. stappersii. These oscillate at periodic intervals of 3 years (Pearce, 1988) as opposed to the interval of 6 to 8 years observed in the Burundi sector of the lake (Herman 1977; Coulter, 1988). The interval of the species abundance oscillations in Tanzania remain un-established (Bayona, 1988). However, abundance of L. miodon both in Burundi and Zambian waters is reported to have remained stable and free of predator-prey induced fluctuations because the species is less vulnerable to predation in its inshore habitat (Roest, 1988; Pearce, 1988). The observed fluctuations in abundance between L. mariae and L. miodon in the Tanzanian sector of the lake are therefore considered to be un-usual and are more linked to tense predation which erupted after the decline of S. tanganicae.
Time series trends in abundance and distribution of L. miodon demonstrate paucity and stability of the species in the Tanzanian waters of Lake Tanganyika. This is in part, due to dominance of the species in the inshore habitat, usually a very narrow strip, where exposure to predation is minimal. However, the observed increase in exploitation level is purported to have indirectly affected the usual distributional pattern in space and time of the species. Consequently, individuals were exposed to tense predation, portraying predator-prey induced fluctuations in abundance in some years.
However, there is total lack of catch statistics of fish amendable to artisanal exploitation to assess the species yield levels and catch composition. Worse still, the industrial exploitation is characterised by prolonged poor catches of fish accompanied by escalating effort levels and the ultimate decline in fish abundance, the factors which may antagonize the purpose for sustainable development of the resource. These are among the major bottlenecks to be addressed by stock assessment.
During the late eighties and the early nineties, most fishermen in the Tanzanian sector of Lake Tanganyika expressed their concern for the continuously declining catches of fish, especially S. tanganicae which used to make up for a lucrative fishery. Many industrial fishing vessels are no longer operating because poor catches have rendered them un-economical to run. Even operators of the artisanal lift-net fishery have began to transfer their gears to Lake Victoria in order to tap the alternative abundant species of Rastrineobola argentea at the expense of the decline of catches of S. tanganicae (Ndugumbi, pers. comm.). Fig. 5 indicates that the abundance of S. tanganicae has declined to very low levels, mostly available seasonally for harvesting (October-November). Therefore, there is a general fear that poor catches and a trend of declining fish abundance are advancing into a lakewide problem. This calls for immediate assessment of the stocks so that an optimum level of exploitation can be determined to ensure that the long term yield from the fishery is maximized.
Bayona et al. (1992) noted an increasing trend in annual catches for the combined fisheries (artisanal and industrial) in the Tanzanian sector of the lake (Kigoma and Rukwa Region) during 1974 through 1987. However, the industrial fishery which is restricted in a small area, off Kigoma registered a decline in catches and fish abundance despite the noted increase in fishing effort. This contrasting situation made the authors to believe that Kigoma was facing a problem of local overfishing. Also by assuming that abundance of predator and prey species had stabilized between 1981 to 1986 (Fig. 4) an optimum fishing effort of 1034 hauls was established using a Surplus Production Model (SPM). This level of effort manifested the problem that over 35% of the effective effort during the period in question was effected over and above the optimum level. This situation is, however, no longer tenable to the industrial fishery in Kigoma because a threatening decline in both artisanal and industrial catches seems to extend to almost the entire sector of Tanzania if not the whole lake.
In the case of L. miodon, no estimate of its contribution to the total harvest can be established by the published catch and effort statistics for the lake. Catch statistics taken by the Fisheries Division only combine the two species together (S. tanganicae and L. miodon) as dagaa. There is a need for managing the two species independently which necessitates the collection of catch statistics of the species, independently.
Both abundance and yield assessment techniques for this species should take into account of the species distributional patterns in space and time to avoid under estimation due to the problem of non-accessibility. Fig. 8 indicates the length frequency distribution of L. miodon from the inshore artisanal fishery. Individuals ranging from 8.5cm to 13.5cm in fork length were found distributed in the inshore area but no juveniles or young individuals were observed. However, individuals of similar sizes can be encountered in distant pelagic waters at certain times of the year. This prompted Ndugumbi et al. (1976) to report that part of the adult population of this species usually remain inshore whereas the other moves off-shore.
An appropriate assessment technique under this distribution, for instance acoustic assessment technique which has lately been improved (Bayona, 1990), should be deployed during October to November when adults of this species increase in the pelagic waters (Coulter, 1991). The in-shore population of L. miodon which has been given little sampling attention in Tanzania should also be assessed by use of Length Based Fish Stock Assessment (LFSA) methods described by Sparre et al. (1989). Therefore length and weight data from the artisanal fishery are needed for this analysis.
Bayona, J.D.R. 1988. A review of the biological productivity and fish predation in Lake Tanganyika. In Lewis, D. (Ed.) CIFA Occasional Papers, 15: 1–17.
Bayona, J.D.R. 1990. Fisheries acoustic applications in marine waters and the commercial availability of acoustic systems. MARA Occas. Papers 1990. Dalhousie University Press. 45p.
Bayona, J.D.R., S. Ndaro and B. Ngatunga. 1992. Industrial fisheries in the Tanzanian sector of the Lake Tanganyika: A case of local overfishing in Kigoma. A paper for TAFIRI's tenth year anniversary. Special Edition of TAFIRI Bulletin 1992.
Chapman, D.W. Bayona, J and Ellis, C. 1974. Analysis of test fishing limnological sampling in the Tanzanian waters of Lake Tanganyika in October 1974. Bujumbura; Burundi: FAO; URT/71/012/12 : 1–27, 13 Ann. IDC.
Chapman, D.W. 1976. Acoustic Estimates of Pelagic Ichthyomas in Lake Tanganyika with an Inexpensive Echosounder. Kigoma, Tanzania. Transactions of the American Fisheries Society, Vol. 105 No. 5 : 581 – 587.
Coulter, G.W. 1970. Population changes within a group of fish species on Lake Tanganyika following their exploitation. J. Fish. Biol. 2: 329–53
Coulter, G.W. 1976. The biology of Lates species (Nile Perch) in Lake Tanganyika and the status of the pelagic fishery for Lates species and Luciolates stappersii. J. Fish. Biol. 9: 235–59.
Coulter, G.W. 1988. Production dynamics in Lake Tanganyika. In Lewis, D. (Ed.) CIFA Occasional Papers, 15: 18–25.
Coulter, G.W. 1991. (Ed.). Lake Tanganyika and its life. Natural History Museum Publications. Oxford University Press. London, Oxford and New York. 354 p.
FAO, 1978. Fishery biology and stock assessment. UN/FAO Tech. Report. DP/URT/71/012/1:1–37.
Henderson, H.F. 1976. Notes on Luciolates based on a study of length frequency diagrams from the ring-net fisheries in Lake Tanganyika; and notes on the large size of Limnothrissa in the catches of the ring-net fishery in Tanzania. UN/FAO Report. FI:DP/URT/71/012/29:1–6.
Herman, C. 1977. Fishing in Lake Tanganyika. Present situation and prospects for development. CIFA/77/Symp. 20 Nov. 1977. Bujumbura, Burundi. 9p.
Matthes, H. 1968. Preliminary investigations into the biology of the Lake Tanganyika Clupeidae. Fish Res. Bull. Zambia, 4:39–45
Ndugumbi, Z. Van Well, P. and W.D. Chapman. 1976. Biology of Limnothrissamiodon in Lake Tanganyika. UN/FAO Report, FI:DP/URT/71/012/36:1–7.
Pearce, M.J. 1985. A description and stock assessment of the pelagic fishery in the South-east arm of the Zambian waters of Lake Tanganyika. Report of the Department of Fisheries, Zambia:74p.
Pearce, M.J. 1988. Some effects of Lates species on pelagic and demersal fish in Zambian waters of Lake Tanganyika. In Lewis, D. (Ed.). CIFA Occasional Papers, (15) :69–87.
Roest, F.C. 1988. Predator-prey relations in Northern Lake Tanganyika and fluctuations in the pelagic fish stocks. In Lewis, D. (Ed.), CIFA Occasional Papers.(15): 104–129.
Sparre P., E. Ursin, and S. C. Venema. 1989. Introduction to tropical fish stock assessment. Part 1-Manual. FAO Tech. Paper No. 306/1. Rome. 337p.
Table. 1 Time series trends for the industrial fishery in the Tanzanian sector of Lake Tanganyika, 1974 – 1987.
| Year | Recorded Catch Kg | Average No. of Boats | Number Boats-Nights | Number of Hauls | Catch/Boat Night | Catch/Haul |
| 1974 | 244041 | 1 | 113 | 138 | 2160 | 1768 |
| 1975 | 190427 | 1 | 106 | 253 | 1760 | 753 |
| 1976 | 607747 | 3 | 479 | 931 | 1269 | 653 |
| 1977 | 469649 | 3 | 405 | 865 | 1160 | 543 |
| 1978 | 342604 | 4 | 419 | 784 | 818 | 437 |
| 1980 | 344176 | 4 | 675 | 1431 | 509 | 241 |
| 1981 | 270331 | 4 | 732 | 1334 | 420 | 203 |
| 1982 | 236917 | 3 | 681 | 1034 | 409 | 229 |
| 1983 | 180150 | 4 | 562 | 1093 | 320 | 165 |
| 1984 | 192348 | 4 | 485 | 1025 | 498 | 188 |
| 1985 | 265882 | 5 | 861 | 1394 | 378 | 191 |
| 1986 | 213907 | 5 | 1040 | 1243 | 269 | 172 |
| 1987 | 160078 | 5 | 1307 | 1417 | 202 | 112.9 |
Table.2 Percentage species composition of industrial catches in Lake Tanganyika (Tanzania), 1974–1987
| Species | S. tanganicae | L. miodon | L. stappersii | L. mariae | L. microlepis | L. angustifrons |
| 1974 | 50.9 | 0.3 | 27.1 | 8.2 | 11.4 | 2.2 |
| 1975 | 13.6 | 0.3 | 72.1 | 9.2 | 3.2 | 1.4 |
| 1976 | 17.1 | 0.2 | 64.3 | 5.0 | 12.5 | 0.8 |
| 1977 | 6.1 | 0.5 | 75.6 | 1.4 | 15.3 | 1.1 |
| 1978 | 1.2 | 1.4 | 94.5 | 1.4 | 1.1 | 0.4 |
| 1980 | 2.2 | 3.2 | 92.6 | 1.6 | 3.3 | 9.2 |
| 1981 | 6.7 | 2.2 | 85.3 | 3.6 | 1.6 | 0.5 |
| 1982 | 6.2 | 0.5 | 88.2 | 3.6 | 1.4 | 0.2 |
| 1983 | 1.9 | 0.1 | 93.0 | 3.7 | 1.2 | 0.1 |
| 1984 | 2.6 | 0.5 | 95.2 | 3.5 | 0.3 | 0.2 |
| 1985 | 1.3 | 0.5 | 96.1 | 1.8 | 0.1 | 0.03 |
| 1986 | 1.6 | 0.6 | 96.6 | 1.1 | 0.1 | 0.03 |
| 1987 | 1.7 | 0.7 | 95.5 | 1.6 | 0.4 | 0.0 |
Fig. 1 A trend in fishing effort of the industrial fishery in the Tanzanian sector of Lake Tanganyika, 1974–1987
Fig. 2 A trend in catches of prey and predator species for the industrial fishery in the Tanzanian sector of Lake Tanganyika, 1974–1987.
Fig. 3 A trend in mean monthly catch for clupeids based on the industrial fishery in the Tanzanian sector of Lake Tanganyika, 1974–1987.
Fig. 4 The variation of catch per unit effort of the prey and predator groups of fish based on the industrial fishery in the Tanzanian sector of Lake Tanganyika, 1974–1987.
Fig. 5 Variation in catch per unit effort of clupeids under the industrial fishery in the Tanzanian sector of Lake Tanganyika, 1974–1987.
Fig. 6 Oscillations in abundance between Stolothrissa and Luciolates as based on the industrial catches from the Tanzanian sector of Lake Tanganyika (1974 – 1987).
Fig. 7 Oscillations in abundance between Limnothrissa miodon and Lates mariae based on the industrial catches from the Tanzanian sector of Lake Tanganyika, 1974–1987.
Fig. 8 Length frequency distribution of Limnothrissa miodon from the artisanal fishery of Lake Tanganyika, Kigoma, September 1992.
