The pelagic fishery of Lake Tanganyika provides employment to about 1 million consumers and protein to many more people living around the lake. It is mainly based on three commercially important fish species namely two clupeids Stolothrissa tanganicae and Limnothrissa miodon commonly known as “dagaa” and a perch Lates stappersii. A declining trend of the perch both in its composition and abundance in the pelagic fish landings is partly tied to local over-fishing and climate change. There are three important periods in the exploitation process identified as: (1) a traditional fishery period, the pre-1975 period marked by low catches of Lates spp. and a dominance of clupeids; (2) an industrial fishery period (1975–1978) marked by high catches of Lates stappersii and high total landings; and (3) an artisanal fishery period (post 1984) with relatively low catches and high dominance of clupeids especially Stolothrissa tanganicae. Population growth and refugee influx from the politically volatile Burundi and Democratic Republic of Congo both caused high fish protein demand, particularly for dried clupeids. Also changes in the rural economy caused by drought in the area are resulting into high exploitation pressure on the pelagic resources. Concerted efforts to prevent/reduce exploitation pressure on the pelagic fish resources should take into action a lake wide management strategy where by management issues in all the riparian countries should be harmonized. More effort should be geared towards preventing the use of beach seines and small meshed nets subsequently ensuring a sustainable utilization of the pelagic fish resources.
Lake Tanganyika's pelagic fish community is basically comprised of six endemic species; namely Limnothrissa miodon, Stolothrissa tanganicae, Lates stappersii, L. angustifrons, L. mariae, and L. microlepis. Of the six species, three are the main targets in the three fisheries: traditional, artisanal and industrial. These are the two clupeids—Limnothrissa miodon and Stolothrissa tanganicae (locally called “dagaa”), and the perch, Lates stappersii (“mgebuka”) which predates clupeids. Although a lake-wide increase in catches and catch per unit effort, as a result of increased efficiency of artisanal units, has been reported (Aro and Mannini, 1995; Mölsä et al., 2002; Sarvala et al., 2006), local over-fishing has also been reported (Bayona et al., 1992; Shirakihara et al., 1992; Mulimbwa, Centre de Recherche en Hydrobiologie (C. R. H.), Bujumbura, Burundi, pers. comm.).
The pelagic fishery employs about 1 million people, both in the fisheries and post harvest sector around the lake (Mölsä et al., 1999). There are about 13000 fishermen in the Tanzania sector of the lake with half of them operating from Kigoma. The demand for animal protein in the area is high and the importance of fish, especially clupeids, in the daily meal of Kigoma people is high. Kigoma is a refugee stricken region and has a relatively high population growth rate (4.8%) (Tanzania Bureau of Statistics, Kigoma Office); consequently fishing pressure, especially on clupeids, has been increasing during the last decade.
The pelagic fisheries of Lake Tanganyika is characterized by daily, seasonal and long-term fluctuations (Mölsä et al, 2002) affecting both the economic condition and food security of the many people depending on it.
The demand for fish protein and consequently fishing pressure in Kigoma is likely to increase (Chitamwebwa and Kimirei, 2005; Sarvala et al., 2006). While the Tanganyikan fish stocks are already heavily targeted and expected climate change might furthermore cause a negative impact on these stocks, the need to understand the causes for the fluctuations in fisheries is vital (O'Reilly et al., 2003; Sarvala et al., 2006; Verburg and Hecky, 2007). This paper therefore assesses the situation using long-term data series available from the Kigoma area and proposes/suggests measures to be taken into consideration that are important to the sustainability of the fisheries.
Fishing on Lake Tanganyika is done at night by light attraction using pressure lamps. There are three main types of pelagic fisheries in Lake Tanganyika: traditional (scoop-net), artisanal (lift-net) and industrial (purse seine). Artisanal fishery is the most important type in the Tanzanian sector of the lake, both in terms of number of units and landings; catching mainly the clupeid, Stolothrissa tanganicae. Although the efficiency in catching different species (S. tanganicae, Limnothrissa miodon and Lates stappersii) differs from one method to another, all methods land a considerable proportion of each species. Catch and effort statistics from the Tanzanian side of Lake Tanganyika are fragmentary and sometimes unreliable. This is probably because of the bad data collection system and lack of enough skilled manpower. In this study, most reliable catch data have been acquired from Fisheries Department in Kigoma, separated into different species (1967–1978 and 1984–2003) and total catches (1967–1991), and from Tanzania Fisheries Research Institute (TAFIRI) Kigoma Centre (1976–1978 and 1981–1986). The data from TAFIRI were recorded from purse seiners at one landing beach in Kigoma. Another source of data was the now defunct Kigoma Uvuvi Limited (1985–1988, also from purse seine) itemized by species and by months; likewise, there are gaps in this data set too. However, catch composition data for 2003 were collected from three landing beaches in the Kigoma Township. Relative abundance expressed as metric tons (m.t) per fishing unit per night and individual species composition were calculated. For these analyses catches of the clupeids (S. tanganicae and L. miodon) were grouped as “clupeids” and juveniles of L. stappersii were grouped with their adults.
Figure 1a and Figure 1b present species composition and total catch data, respectively, from Kigoma region of Lake Tanganyika. They indicate that the contributions of the different species in the pelagic fisheries were over 70% and 65% “clupeids” during 1967–1974 and after 1983 respectively, and 60% Lates stappersii during 1975–1978 (Figure 1a). Contribution of ‘big’ Lates species (L. mariae, L. angustifrons and L. microlepis) to the fishery was always comparatively lower and showed a general total catch decrease over the past four decades. However, they contributed about 33%, 15%, 19%, 24% and 16% of the landings during 1968, 1972–1973, 1975 and 1994 respectively. Table 1 presents nightly catch rates for the industrial fishery (purse-seine) as recorded around Kigoma area from 1981 to 1987 by TAFIRI–Kigoma. The dominance and importance of L. stappersii from the table can not be over emphasized.
Figure 2 indicates different catch compositions from two, industrial (purse seine, Figure 2a) and artisanal (Figure 2b), fisheries respectively. While L. stappersii dominated in the industrial, clupeids dominated in the artisanal catches. The data in Figure 2a were recorded, by TAFIRI–Kigoma, from the then existing purse seiners, while those in Figure 2b are from Fisheries Department in Kigoma. This data is composed of both the artisanal and industrial catches. The artisanal fisheries contributed most of the clupeids catch which was by far the highest as compared to the industrial catches. From Figure 2, also, it can be observed that the industrial fishery was catching “big Lates spp.” with relatively higher efficiency than the artisanal fishery; however, their contribution faded and became negligible from 1984 onwards.
Figure 3 presents catch and effort development for Kigoma region. It indicate a more or less exponential growth of both the number of fishers, canoes and catch from 1967–1977. The catches fell from 77,000 mt in 1974 to about 40,000 m.t between 1977 and 1983. It then rose to the highest catch ever recorded in the area (11,5000 m.t) in 1985 after which it declined gradually to 64000 m.t in 1991. The number of canoes fell exponentially to minimum in 1991, the figure corresponding to that of 1967; however, the number of fishermen remained high.
Different fishing methods/gears have had a significant impact on the Tanganyika's pelagic fisheries catches and species composition over the past four decades of exploitation. The dominance of a particular species during one particular time was dependent on the dominance of one or more fishing gear type/method. The period before 1975 was dominated mainly by a traditional fishery, whereby traditional gears like Lusenga–scoop net, and beach seines were prevalent. The traditional methods were used in the inshore waters catching exclusively clupeids and juveniles of Lates stappersii; clupeids catches dominated this period. Low catches of the fast swimming perch may be the result of a slow scooping process. However, during early 1970s, lift-nets using catamarans, were introduced to rural areas, from Burundi, so that fishermen would also exploit the offshore abundant fish stocks (Colaris, 1973; Stride, 1976; Andrianos, 1977). As the fishing methods were being improved/changed, the catches also gradually increased; this also attracted more fishermen and fishing units as evidenced by Figure 3.
The period between 1975 and 1978 was dominated by catches of L. stappersii; this period also saw elevated catches of big Lates spp. (Figure 1). Artisanal and industrial fisheries using catamarans and purse seine units respectively were in use during this period. The use of new methods and boats may be responsible for change in composition and increase in catches although the purse seine fishery never developed much (Chitamwebwa and Kimirei, 2005). These purse seiners fished in offshore waters and were relatively more efficient in catching the fast swimming perch, which is an offshore, deep water species, hence its dominance (Figure 2a). It spawns offshore and so most of its population is found offshore, unlike the clupeids that spawn in the inshore areas. Actually, Limnothrissa miodon spends almost all of its entire life in the inshore waters (Coulter, 1991). The continued decrease of other Lates spp. was a result of high fishing intensity since they are long-lived and can not withstand high exploitation pressure unlike L. stappersii which is the smallest and a relatively more resilient species (Coulter, 1976, 1991).
The period after 1984 is classified under artisanal fishery, because it has more of its characteristics (e.g. high clupeids catches), especially when a combination of artisanal and purse seine data is considered (Figure 2b). Until 1987, there were about 6 (mean ± SD = 5 ± 2) purse seiners in the fishery, which were reduced to 2 by 1988 and only one intermittently fishing purse seine unit by mid 1990s (Mölsä et al., 1999; Chitamwebwa and Kimirei, 2005). However, from Figure 2, the artisanal fishery contributed more importantly to the total landings. The lift-net fishery became efficient to the level of competing out purse seiners (Petit and Kiyuku, 1995; Roest, 1992; Lugiko, Surveillance and Control, Western Zone, Kigoma, Tanzania, pers. comm.). This may explain the dominance of short-lived clupeids in the catch.
There was a steady decrease in the total catch from 83100 m.t in 1985 to around 11000 m.t in 1994, then a sharp increase to a peak in 2001 (92000 m.t) and a declining trend thereafter. The clupeids have maintained their dominance ever since; contributing between 5 and 6 times that of L. stappersii (Figure 1b, Mölsä et al., 2002). The fisheries catches situation is even worse to date where the decreasing catches are characterized by daily, seasonal and long-term fluctuations that are partly due to predator-prey relationships existing between clupeids and Lates stappersii and local over-exploitation (Roest, 1988; Bayona et al., 1992; Mölsä et al., 2002; Chitamwebwa and Kimirei, 2005; Sarvala et al., 2006). The poor contribution of Lates stappersii, unlike during the period prior to 1975, may be a result of a combination of factors including local over-fishing.
In general, the total catches of the fishery constituting six species were mainly determined by the ‘main contributor species’ of the landings in a particular period. For example, the “clupeids” catches determined the total catches of the periods 1967–1974 and 1984–2003 while Lates stappersii determined the total catches of the fishery during 1975–1978.
Increased regional fish protein demand, especially dried clupeids, caused by high population growth and easy processing methods for clupeids (sun-drying on sand or gravel) could be the driving forces towards high exploitation pressure on this ‘precious’ pelagic resource. The clupeids (S. tanganicae) live for about one year; have high P/B ratio, are highly fecund and consequently have a high turnover rate, features that help them to withstand high exploitation pressure. However, a recovery in the highly fished areas of Zambia, Burundi and Congo has not been realized. Plisnier (1997) suggests that the decline was not only effort increase mediated, but also climate change may be playing an important role. Nonetheless, the effect of fishing pressure resulting into local over-fishing can not be neglected since the current fishing pressure is reported to be high (Chitamwebwa and Kimirei, 2005; Sarvala et al., 2006). The current demand is forcing fishermen to use powerful engines (up to 55 HP) to trudge distant waters; unlike in about two previous decades, as reported by fishermen, when fishing was done in the inshore waters. Fishermen are also forced to use both improved fishing gears and destructive ones (small mesh size nets that catch juveniles) to catch clupeids (I.A Kimirei, Tanzania Fisheries Research Institute, Kigoma, Tanzania, personal observation). Mosquito nets were reported to be in use in Zambia (Aro and Mannini, 1995) and Tanzania is no exception especially now when the clupeids business is lucrative. A codend made of mosquito net was recently observed at Katonga landing beach (I.A Kimirei, Tanzania Fisheries Research Institute, Kigoma, Tanzania, personal observation). This may result in a collapse of the fishery if no stern measures are taken nor proper management measures sought.
Beach seining is prohibited in Tanzania; however, it is still being used in some parts of Kigoma region where the fishing is done at night to avoid the government's hand. For example, beach seines are operated in areas around Malagarasi River and other shallow stretches where beach seining is possible (I.A Kimirei, Tanzania Fisheries Research Institute, Kigoma, Tanzania, personal observation). The new efforts by the Government of Tanzania in combating illegal fishing methods by establishing a patrol and surveillance unit in Kigoma are highly commended. Beach seining destructs spawning grounds for “big” Lates species, critical habitats for Limnothrissa miodon (Coulter, 1991), and all the other hundreds of species of cichlids as well. It is thought that the decline of big Lates spp. in Lake Tanganyika is linked to, apart from heavy exploitation by industrial units, beach seining that resulted in both growth and recruitment over-fishing of these species (Ngatunga, 1992).
Conclusions and Recommendations
From the foregoing discussion, three important periods in the exploitation process of the Tanganyikan fisheries can be identified: (1) a traditional fishery period, the pre-1975 period marked by low catches of Lates spp. and a dominance of clupeids; (2) an industrial fishery period (1975–1978) marked by high catches of Lates stappersii and high total landings; and (3) an artisanal fishery period (post 1984) with relatively low catches and high dominance of clupeids especially Stolothrissa tanganicae. Fluctuations in fish species compositions and abundances have, therefore, emanated from different causal factors. These range from the type of fishery operating in a certain period causing different compositions of the pelagic ichthyomass; increased exploitation that is rooted in increased fish protein demand and improved fishing methods resulting in local over-exploitation.
Despite the fact that the Tanganyika pelagic fishery is moving towards a single species fishery, there is a need for directing research efforts into finding a proper mesh size to be used in the ‘multispecies’ fishery. Other management measures may include, but are not limited to, closures, both spatial and temporal, minimum mesh sizes and size limits where minimum legal lengths are set. The last two have been successful in Lake Victoria. In addition, expansion of the fishery should be limited by allowing a set number number of fishers and fishing vessels through licensing. However, licensing should be used as a means of controlling the fishery rather than for revenue collection. It is also important to harmonize enforcement of management measures in all the riparian countries since the depletion of stocks in one part will consequently affect the others. It is also imperative that all stakeholders (fisherfolks, processors and the community at large) be involved in the management decision making if sustainability is to be achieved. Moreover, new fisheries statistics collection frameworks need to be put in place since the existing ones do not seem to work properly.
The authors would like to acknowledge the assistance from TAFIRI Kigoma, Kigoma Uvuvi and the department of Fisheries in Kigoma for providing historical catch and effort data. Also, thanks are due to the CLIMLAKE Project for research funding, the American National Science Foundation (NSF) through the NYANZA Project (ATM#9619458 and 0223920 and DBI-0353765) for providing subsistence allowance, the GLOW IV organizing of the Aquatic Ecosystem Health and Management Society (AEHMS) committee for travel assistance, and two anonymous reviewers for their constructive criticism.