APPENDIX F
SOCIO-ECONOMIC ISSUES FOR PLANNING IN SUPPORT OF FISHERIES MANAGEMENT

by

D. Gréboval
Inland Fisheries Planning, Development and Management in East (Central/Southern) Africa
P.O. Box 1250
Bujumbura, Burundi

1. INTRODUCTION

The fisheries of Lake Victoria have undergone radical changes over the last decade mainly because of the rapid proliferation of the Nile Perch (Lates niloticus) and its effect on the lake's ichthyomass. While the ecological disruptions brought about by this introduced predator drew worldwide attention, the fact that the fisheries of Lake Victoria were in the process of becoming the most important of Inland Africa remained mostly unnoticed outside of the three countries concerned.

Today, the single most important issue which these countries are facing is the urgent need to bring these fisheries under proper management in order to avoid expected resource over-exploitation and the rapid dissipation of the huge socio-economic benefits brought about by the new fisheries regime. The present paper addresses some of the main socio-economic issues related to the management of these new fisheries.

For Kenya, Uganda and Tanzania the fisheries of Lake Victoria presently represent a most valuable source of food and much needed animal protein. The lake has become the main source of fish production in all three countries, with annual production estimates for the last three years ranging from 300 000 to 400 000 t. Compared with the estimated average production of 100 000 t observed under the pre-Nile perch regime (Table 1), an average surplus production of about 150 000 t per year was caught in the lake over the last 10 years.

Other very important benefits were also derived in terms of employment, financial returns to the industry and reduced consumer prices for fisheries products within a wide area all around the lake. According to Reynolds and Gréboval (1988), the value of the catch in constant 1987 US$ increased from US$ 25 million in 1975 to US$ 72 million in 1987, reflecting the magnitude of the benefits derived from the new fisheries regime since the beginning of this decade. For the most part these benefits accrued to small-scale fisherfolk whether previously engaged in the fisheries sector or brought into the industry by its profitability. Indeed, with the exception of the Nile perch fisheries of Kenya where industrial processing and marketing companies play a significant role, the fisheries of the lake remain almost exclusively exploited by small-scale fisherfolk at all levels from harvesting to marketing. It is now estimated that over 350 000 fisherfolk derive their livelihood from these fisheries.

The major reason why such large benefits have been derived so far is that riparian fisherfolk and consumers did adjust very rapidly indeed to the changing composition of the fisheries resource base. It follows from relative consumer preference and relative exploitation cost, that the present resources of Lake Victoria are a lot more valuable than the resources under the former regime. Indeed the latter was characterized by the predominance of Haplochromis species which constituted an estimated 80% of the lake's biomass (FAO, 1973) and remained nearly unexploited due to very low demand for these species.

The implication of this evolution is, however, that both high demand for table fish (especially if few immediate substitutes are available) and low production costs are also the key factors leading to the systematic overexploitation of unmanaged fisheries. Indeed, following the rapid and sustained expansion of fishing effort since the early 1980's, the two major fisheries based on Lates niloticus and Oreochromis niloticus are becoming heavily exploited. With effort still reported to be increasing in these two fisheries, they might very soon be entering a phase of progressive overexploitation. As this process unfolds the rent -- that is the value of production over and above what it costs to produce -- will progressively be dissipated and the huge benefits derived from the new fisheries regime would have been only short-term gains to the riparian fisherfolk and consumers.

The way this process of overexploitation will unfold will depend of course of the bio-economic characteristics of each of the three major fisheries (Lates, O. niloticus and Rastrineobola argentea) and their interaction; and on whether or not the three States concerned can introduce and enforce management regulations which will at best bring the process to a halt or at least slow it down while further information is gathered for the elaboration of more effective management schemes.

2. RECENT EVOLUTION AND SOCIO-ECONOMIC CHARACTERISTICS OF THE FISHERIES

Indeed little information is currently available on the characteristics of the new fisheries of Lake Victoria and their interrelated dynamics. The need for further biological and stock assessment investigations has been pointed out in a number of instances, especially with respect to the three key species already mentioned. More recently, the need for socio-economic investigations has also been stressed (CIFA, 1988; Reynolds and Gréboval, 1988).

While a large number of articles have been published on the recent evolution of the fisheries of Lake Victoria (of reviews in: Reynolds and Gréboval 1988, and University of Leiden, 1986), these have mostly focused on the decline of Haplochromis stocks and have unfortunately been based more on secondary sources of information than direct observation. Nowadays, however, a clearer picture of what happened to the fisheries and the resources on which they are based is emerging, which allows for a sketchy comparison of the new and former fisheries regime of Lake Victoria.

2.1 The Pre-Nile Perch Fisheries Regime

From the inception of fisheries in Lake Victoria until the mid 1960s the fisheries exploited mainly Oreochromis esculenta and variabilis. The fisheries based on these species collapsed through overfishing, forcing the fishermen to exploit the smaller and less valuable Haplochromis and Rastrineobola.

Until the end of the 1970s, the resource base of Lake Victoria was characterized by the predominance of Haplochromis stocks. These cichlids did constitute, according to the scientists involved, quite a unique case material for the study of lacustrine ecology and evolutionary biology (e.g., Greenwood, 1984; Fryer, 1984; Coulter et al., 1986). From a socio-economic point of view, however, this resource had very little value and remained a food of last resort all round the lake. In spite of various efforts undertaken in the seventies through marketing campaigns and the creation in Tanzania of a trawl fishery and related industrial processing which did prove quite unprofitable, the fishery remained characterized by a very low level of exploitation and poor marketability (Dhatemwa, 1982; Nyholm and Whiting, 1975). Furthermore, it is doubtful that haplochromines could have withstood heavier fishing pressure because of the trophic specialization of the many species concerned (Goodswaard and Witte, 1985). The fish biomass of Lake Victoria also consisted of more valuable species groups: Oreochromis/Tilapia, Bagrus, Synodontis, Clarias, Protopterus and Barbus.

Total catch during the late 1960s and 1970s were relatively stable as indicated in Table 1. Average production was over the period 1968–79 of 100 000 t per year. The stability of the global figure does, however, shadow a progressive reduction of the real value of the catch resulting from overfishing. This phenomenon has been well documented starting with Graham (1929). Among the species especially affected over the last two decades are Labeo victorianus, Oreochromis esculentus, Schilbe mystus, Alestes jachsonii, Alestes sadleri and Barbus. The key factor in the gradual collapse of these fisheries has been the systematic overexploitation of most valued inshore stocks though the intensification of gillnet fishing and the progressive decrease in mesh size used (e.g., Cadwalladr, 1969; Mann, 1970; Marten, 1979). As a result, over half the total catch from Lake Victoria consisted of little valued Haplochromis and Rastrineobola argentea in the later part of the 1970s.

Given the general socio-economic characteristics of the fisheries, it is likely that this trend would have continued if the overall fisheries situation had not been dramatically transformed by the rapid expansion of the Nile perch stocks after 1978–79. The socio-economic characteristics of these fisheries are not known in any details as only a few investigations of the sort were ever conducted. However, past reports of the CIFA Sub-Committee for Lake Victoria do provide basic information. In this section only the characteristics of the fishery which bear some relation to its management are considered.

Until the creation of the Nyanza Fishing & Processing Company which started operating up to four trawlers in the mid seventies, the fisheries remained solely exploited by small-scale fishermen. The basic fishing unit is composed of wooden canoes and simple gear which do not constitute a sizeable investment. Since the introduction of synthetic fibre nets in the early 1950s, only gradual changes occurred in the technologies used in the harvesting sector such as: the replacement of dugouts by planked canoes; a slowly increasing motorization rate which nevertheless remains marginal; a gradual decrease in mesh size and higher specialization by target species in the gillnet fisheries; and a move away from traditional fishing techniques such as traps and weirs used to catch once relatively abundant anadromous species, and toward beach seines, mosquito-nets, scoop and ring nets used in the Haplochromis and R. argentea fisheries.

Table 1

Lake Victoria-Total nominal catches of the artisanal fisheries by country, 1968–79

^a Uganda's total annual catch assumed to be less than the 1973 magnitude, following the deterioration of the fishing industry during the abnormal economic period 1973–79

Source: CIFA (1982)

It is to be noted that this evolution implied higher investment and operational costs per fishing unit whereas gross benefits derived from the fishery were actually decreasing throughout most of the sixties and seventies. While the number of fishing units did increase during the sixties, it seems to have remained quite stable over the seventies and might even have slightly decreased. Butcher and Collaris (1973) reported the total number of fishing units to be 12 142 in 1972, while in the late seventies estimates range from 11 000 to 12 000 (CIFA, 1981; CIFA, 1983). According to the same sources, the number of fishermen operating in the 1970s is estimated at around 50 000. Roughly half of the boats are owner operated, with 80 % of the fishermen deriving their income primarily from fishing. A definite trend towards operating on a thoroughly commercial basis is thus to be noted, especially in Kenya. But subsistence fishing remains nevertheless significant, especially in Tanzania where a large proportion of fishermen are engaged in complementary agricultural activities.

This evolution from an originally subsistence/part time mode of exploitation is explained by two main factors, namely the scarcity of agricultural land and of alternative employment opportunities around the lake as well as by the specialization of some riparian ethnic groups in fishing activities. In such context, one can expect fishermen to remain actively involved in fishing activities even if it becomes less and less productive and profitable-as mobility out of an economically overexploited fishery depends on opportunity costs of labor and capital (Panayoutou, 1982), and since these appear limited in the case of Lake Victoria.

This is why the intensity of fishing effort remained especially important in Kenya as compared to the two other countries, leading to more intense overexploitation. Indeed the shortage of agricultural land and alternative employment is higher in the Nyanza Gulf area than elsewhere around the lake and the Luo ethnic group is highly specialized.

The process of overexploitation in an open access fishery is characterized by the dissipation of all resource rent-that is any revenue over and above the cost of production inclusive of minimal returns to both labor and capital. In such situation intramarginal rents can still be derived by those units which are more performant, i.e., those which are generally but not necessarily using more capital intensive techniques.

The fishery of the lake in the 1970s is quite typical of this situation with most of the fishermen involved in the predominant gillnet fisheries having presumably become progressively impoverished, and with few better equipped fishing units possibly earning above-normal profits especially in more recent fisheries such as light fishing for Rastrineobola argentea, or because of seasonal involvement in several fisheries. While the author is not aware of any comparative cost and earnings study having been conducted during this period, the impoverishment hypothesis explains the relative stagnation of fishing effort over the 1970s and the slow development of alternative but more capital intensive techniques such as light fishing.

As far as demand is concerned, one notes first that the price of “table” fish¹ does not decrease with decreasing fish size. Actually and as often noted in other African fisheries, a better price can be derived from fish size corresponding to individual serving. This is one of the factors which reinforced the progressive use of smaller mesh size in the gillnet fishery.

Another demand-related factor is the lack of readily available substitutes for “table” fish. First, the price of meat is generally a lot higher than that of fish due to the fact that it is itself relatively scarce all around the lake. Then there is a high qualitative difference between “table” fish and small size species such as Haplochromis/Rastrineobola argentea which implies that these are not substitutes. Finally, the area of Lake Victoria is relatively isolated from other fish production centres, with the noticeable exception of the Uganda shores between Masaka and Jinja. This explains why fishing pressure has always remained less in Uganda than Kenya and Tanzania and why the exploitation of lesser quality fish like Haplochromis and R. argentea has remained marginal in this country. Indeed the very large quantities of Nile perch and tilapia caught in Lake Kyoga were direct substitutes to Lake Victoria's higher quality fish and as such were largely marketed on the same markets.

On other markets around the lake, the relative price inelasticity resulting from the relative scarcity of close substitutes and growing demand related to population growth resulted in higher prices for decreasing quantities of quality fish. The characteristics of demand also explain the pernicious tendency of overexploiting major species through the use of small mesh gillnets. Increased landings of less valued fish may have slowed down this process through its impact on total demand and price but the related and increasing use of beach seines and mosquito nets also had a detrimental impact on high value fish stocks through the capture of juveniles from these species (e.g., Marten, 1979).

¹ Fish served whole or in chunks rather than used as a (sauce) ingredient to a dish as it is generally the case for small pelagic or juveniles

2.2 The Evolution and Characteristics of Lake Victoria's Fishery in the Late 1980s

It has already been over 10 years since the dramatic upsurge of the Nile perch stock. The evolution of Nile perch into a resource base supporting a major and thriving industry was not anticipated by those who have been so outspoken in their opposition to its introduction. But the development of the Nile perch fishery was extremely rapid, and also to be expected, given the socio-economic context.

Even in Kenya where Nile perch is less appreciated by the riparian population than elsewhere around the lake, Nile perch catches rose from around 1 000 t in 1978 to over 4 000 t in 1979–80, reaching nearly 23 0000 t as early as 1981. In spite of this development, heavy predation by Nile perch and a continued intense fishing effort greatly reduced the abundance of haplochromines, while subsequently affecting more valuable species such as Clarias, Protopterus, Bagrus, Barbus, Synodontis and Schilbe.

Judging from catch record the impact of Nile perch on these species has been more drastic (so far) in Kenya than in Tanzania and to a lesser extent Uganda. This follows from the rapid proliferation of Nile perch having originated in Kenyan waters. The evolution of species composition over the period 1974–87 is given in Table 2 for Kenya. It denotes an extremely rapid evolution of the fisheries towards what would eventually prevail with a lag of 3–4 years throughout the lake.

The evolution of the ichthyomass of Lake Victoria is closely following a pattern already observed for Lake Kyoga where Lates along with O. niloticus were introduced in the mid 1950s. Following the virtual disappearance of haplochromines and numerous native species similar to these encountered in Lake Victoria, the fisheries of Lake Kyoga rapidly evolved into a three species fisheries dominated by Lates, O. niloticus and R. argentea (Ogutu Ohwayo, 1985). R. argentea became the main prey of Lates and so far remains virtually unexploited due to the lack of market for this species in Uganda.

Table 2

Kenya - Lake Victoria waters, annual catches (t) by species, 1974–85*

* Source: Statistical Bulletins, Department of Fisheries, Kenya

A similar pattern is apparently occurring in Lake Victoria with Lates, O. niloticus and R. argentea having become rapidly predominant. Unlike in Lake Kyoga, however, Lates does not only feed on R. argentea but younger Lates also prey heavily on the freshwater shrimp Caradina. As a result of species interaction, the abundance of Caradina has been shown to have increased dramatically in recent years (Wanink et al., 1988).

The evolution of species composition and their relative abundance is not only related to interspecies competition but to economic forces. Data presented in Table 3 and the general evolution of major fisheries since the late 1970's provide some insight as to the market forces at play. Among these are:

- The Relative Heterogeneity of Consumer Preferences Around the Lake

O. oreochromis can be viewed as the exception, being widely considered as prime quality fish all around the lake and fetching a higher price than other endogenous tilapiine species -- even if compared to the once highly priced O. esculentus. Lates, at high level of production, fetches a higher price in Uganda than in Tanzania and in Kenya. This is easily understandable if one considers that Lates is found in other Uganda Lakes and that large quantities coming from L. Kyoga have been marketed around the Lake Victoria shores of Uganda for decades. Among other formally major species, one notes, for example, that species such as Bagrus and Clarias are fetching a higher price in Uganda and Tanzania than in Kenya. A reverse situation is observed for Protopterus and Haplochromis.

- The Elasticity of Demand at Low Level of Production

It is implied that while the price of major species do rise as production decreases, it would either stabilize or decrease as marketed quantities are reduced to very low levels. This is a relatively surprising phenomenon which explains some of the data presented in Table 3. For example, Lates in Tanzania was fetching a relatively low price before significant quantities became available. Similarly, the price of key species like Bagrus, Barbus, Clarias or Protopterus appears to have risen relative to the price of the other species when production decreased. But as production diminished to very low levels, the relative prices of these species decreased or remained stable. There are a number of explanations for this phenomenon, such as: the reluctance of wholesalers to deal with low volume products and the difficulty for retailers to find the few consumers who would be ready to pay higher prices for scarce products as marketing structures for such products do not exist. Risk aversion added to the fact that few people may be ready to buy these products at premium price seem to be the key factors.

- The Relative Stability of Nile Perch Prices at High Level of Production

The rapid increase in abundance and landings of O. niloticus in the 1980s has more than compensated for the loss of other high price species. Meanwhile important and ever increasing quantities of Nile perch were being landed without affecting very much its price relative to that of O. niloticus. It is not surprising that Nile perch found a ready market round the lake. This market already existed in Uganda and Lates was not completely foreign in other countries as small quantities were already being landed in the seventies from Lake Victoria and other lakes in the three countries. Furthermore, Lates has supported flourishing fisheries everywhere it is found in Africa including Lake Albert, Lake Tanganyika and Lake Turkana even if the situation of Lake Turkana is significantly different as the lake region is populated by ethnical groups which traditionally do not eat fish. What is surprising, however, is the rapidity with which the Nile perch market expanded. Indeed one would have expected Nile perch to find a ready market equivalent to the quantities of average to low value species supplied under the previous regime. Allowing for population growth this could have represented a market of about 100 000 t for the whole lake at prices between 75 % and 50 % of that of most highly priced species. But one could hardly have expected that the market could expand in a period of 3 to 4 years to absorb a supply 2 to 3 times higher without much effect on prices. Indeed while production more than doubled in Kenya between 1981 and 1986, relative prices only decreased by about 30 %. In the two other countries where Nile perch is even more appreciated, relative prices decreased even less. This demonstrates the popularity of Nile perch in a wide area around the lake, the existence of a huge demand for averagely priced table fish in three countries, and a remarkable ability of fisherfolk and small entrepreneurs to adapt to new marketing situations especially when large profits are to be made.

Table 3

Comparative ex-vessel price indexes in Kenya, Tanzania and Uganda

¹ Source: Actual prices from Department of Fisheries, Kenya. Statistical Bulletins

² Source: Annual reports, Regional Fisheries Office, Mwanza and Mara Regions. Prices from Mara Region (1978) and Mwanza Region (1981–86)

³ Source: Kanyike (1972) for 1972 date reflection consumes preferences in the Entebbe region; 1987 data collected by the author

⁴ Price includes other miscellaneous species for Tanzania

- The Absence of Market for Small Size Fish in Uganda

Under the previous fisheries regime, Uganda was already singled out as the only riparian country where Haplochromis found only a very limited market, and R. argentea hardly any market at all. This follows from the basic unacceptability of these products in either fresh or sundried forms in most of the country (a limited market for R. argentea exists only in the West Nile region). While ample harvesting opportunities exist for the development of a R. argentea fishery of the type developed in Kenya and Tanzania, the primary market of the small quantities now being landed in Uganda is for animal feed.

The evolution of the fisheries at harvesting level appears to have been mostly linked to the rapid evolution of demand. Indeed few adjustments were required on the part of small-scale fishermen as Nile perch is best caught by gillnets. Although the small mesh nets in current use were initially damaged by Lates, and typically lasted no more than six months on Lake Victoria, these were rapidly replaced by larger mesh nets. The profits realized thereafter more than compensated for any losses incurred at the very initial stage of Lates proliferation. The intensification of fishing for O. niloticus and R. argentea also followed from previous involvement in these fisheries.

Similarly, the changes which occurred on the lakes did not require significant changes in processing technologies as sun drying and fish smoking were and remain predominant. Actually it is at marketing level that major adjustments had to be made in order to distribute ever increasing amounts of fish throughout a larger and larger area.

It has been shown that for the overall fisheries sector, these adjustments were made without any major displacement or adverse impact (Reynolds and Gréboval, 1988). There is no doubt that consumers gained tremendously from the changes having affected the fisheries so far, with huge amounts of fish having been made available at more affordable prices throughout a large portion of the three countries concerned. As prices globally remained quite high due to extremely rapid market expansion, fisherfolk from the harvesting to the distribution sub-sectors also derived tremendous benefits from the new fisheries regime. So much that fisherfolk have nicknamed the Nile perch “the savior”. The fisheries were and remain almost exclusively operated by small-scale rural fisherfolk. Fisherfolk previously engaged in the fisheries appear to have gained as much if not more than newcomers who predominantly came from the lake shore as well. There is no evidence that investment cost increased at first, and the use of better equipped units generally reflect progressive reinvestment of profits made in the fisheries.

In the late 1970s total fleet size was estimated at about 11 000 canoes involving about 50 000 fishermen/crew and may be three times as many people in the secondary sector (CIFA, 1983). In 1987, nominal effort has been roughly estimated to have risen to about 19 000 canoes (Reynolds and Gréboval, 1988). Keeping the same figures for fishermen/crew per canoes and primary to secondary employment ratio, primary and secondary employment figures may have risen to 95 000 and 285 000 people respectively. Overall this would mean that 180 000 additional jobs have possibly been created so far.

In this context little case can be made, in our opinion, of commonly advanced arguments on the adverse socio-economic impact of Nile perch proliferation and the changes which it induced. It is nowadays common knowledge that the fish is widely accepted, and highly priced all around the lake in spite of the huge quantities landed in recent years. As far as local displacement is concerned, far from destroying opportunities, Lates has created them by nearly doubling total employment in the fisheries. While a case has often been made of the deforestation resulting from increasing fuel wood demand to smoke increasing amounts of fish, the main cause of deforestation in the area as elsewhere in Africa remains the ever increasing demand for household use. The issue nevertheless warrants further and urgent attention because of its impact on the environment and on the cost of fish processing.

Displacement due to increasing industrial fishing and processing activities has remained marginal. As far as industrial fishing is concerned there are presently about twenty small trawlers fishing for Nile perch in Lake Victoria and nearly all based in Tanzania. More will be said in the next section about the desirability of such investment. However the catch of these trawlers remains very marginal. As far as industrial processing activities are concerned these are all based in Kenya (Kisumu, Nairobi, Mombasa) where fourteen processing plants process Nile perch fillets as a major or secondary activity. Production capacity was estimated in 1987 at about 25 000 t annually, that is about 30 % of the Nile perch production of Kenya for this year (Tettey, 1988). The long-term desirability of having built such a capacity is an open question. Since the early eighties, however, it broadened the Nile perch market and did contribute to stabilize ex-vessel prices by removing excess supply from the shore area. Furthermore the largest part of this production is marketed in Kenya where Nile perch is now one of the cheapest sources of animal protein. In Nairobi and other large cities, Nile perch skeletons have even become quite a delicacy for people of lesser means.

Exports outside East Africa are limited to those of Kenya which represented 11 500 t Nile perch (wet weight) in 1987. This corresponds to less than 5 % of the lake's production for that year. From observations reported in Reynolds and Gréboval (1988) it can be assumed that an additional 5 % to 10 % of the total catch of Lake Victoria is informally exported outside the three countries. This mostly involves exports of sundried R. argentea to Zaire, Zambia, Zimbabwe and Rwanda; as well as some exports of smoked Lates and O. niloticus.

For management purposes, major changes which occurred at socio-economic level since the late 1970s concern:

1. the profitability of fishing activities which remains very high in all aspects of the fisheries industry (Reynolds and Gréboval, 1988);

2. the wide expansion of the market base which is no longer limited to riparian communities and nearby urban centres;

3. a continuation of the trend toward more commercially oriented exploitation even if subsistence fishing/processing/marketing activities are still significant;

4. the possibility that a number of newcomers to the fishing industry could have higher opportunity cost and may therefore leave the fishery if and when it becomes far less profitable;

5. the development of a small industrial fishing and processing capacity which may induce localized conflicts with the interest of small-scale fisherfolk if catches are allowed to decrease very significantly.

In spite of these changes, the sector remains fundamentally characterized by small-scale exploitation taking place in a relatively similar socio-economic framework, with little fundamental change in technologies, techniques and practices compared to the former regime.

3. MAJOR MANAGEMENT ISSUES

3.1 The Stakes

The benefits derived from the fisheries of Lake Victoria have reached unprecedented levels over the last decade. But without proper management, further potential rent could be dissipated within the next decade as the Nile perch stock is already believed to be overexploited with definite signs of overexploitation for highly fished areas like the Nyanza Gulf (FAO, 1988).

The stakes are extremely important. The ex-vessel and market value of total catch were estimated respectively at US$ 72 million and US$ 180 million in 1985 when about 260 000 t were landed (Reynolds and Gréboval, 1988). Now that total catch is reported to have reached about 400 000 t, it is probable that over the period 1978/80–1990/92 the 1985 figure of 260 000 t may represent an average.

If it is the case, it follows that over an expansion period of about 12 years the value of the catch may be estimated at over US$ 2 billion. Compared to the average yearly production of about 100 000 t prevailing under the previous fisheries regime, the value of the net increase in catch over the same period could be estimated at US$ 1.25 billion.

Without proper management, the downfall of the fishery may follow a similar pattern with catches stabilizing at their pre-Nile perch level in 10 to 15 years. Accordingly, the value of the net increase in catch over a period of about 25 years would have been over US$ 2 billion¹. This would have constituted extremely important benefits, even if additional benefits are progressively dissipated in the absence of adequate management.

If the 1985 level of production in weight and value is sustainable, the stakes of proper management may be the additional yearly production of up to 150 000 t of fish valued at around US$ 100 000 million, together with 100 000 to 200 000 jobs.

Now the central question is whether the present resource base can substain high yields. As noted earlier the abundance of O. niloticus has increased tremendously in recent years and fish size remains quite large in spite of heavy fishing pressure. Because of marked difference in habitat preference, the predation pressure from Lates on O. niloticus has been and should remain very low. As in Lake Kyoga, the proliferation of Nile perch has led to the equally important development of R. argentea which abundance is reported to be still increasing². A similar development has been noted for Caridina (Wanink et al. (1988); Ligtvoet (1989) which together with R. argentea and insect larvae constitute the main source of food for Lates. With increasing prey stocks at very high level of predator abundance the prospects of medium terms sustainability from a predator-prey point of view seem very likely. Even if the ecosystem has still not stabilized, it is therefore likely that it can sustain a fairly large stock of Lates.

The parallel already drawn with the evolution of Lake Kyoga's fisheries can serve as a basis for estimating very indicatively what the MSY of Lake Victoria could be. As shown in Table 4 Lake Kyoga was able to sustain over the period 1961–82 an average yield of 82 000 t composed essentially and about equally of Lates and O. niloticus. The catch of Lates has significantly decreased since the mid 1980s due to heavy fishing pressure and diminishing water level, but catch and catch rates were reported by the Ugandan Fisheries Department to have been increasing again since 1988.

Taking a conservative MSY estimate of 50 000 t or a yield of 18.5 t/km² (9.25 t/km² for each species) for Lake Kyoga, and applying it to the very near shore area of Lake Victoria³, Reynolds and Gréboval (1988) estimated quite conservatively that an annual production of 120 000 t⁴ for Lates and 60 000 for O. niloticus could be sustainable. Together with R. argentea which MSY is likely to be very high especially if Lates predation is reduced, total MSY may be roughly estimated to be in the range of 200 000 to 300 000 t. This represents an average lake-wide yield of 3.6 t/km² which is not particularly high for a relatively shallow lake of this type and only about half the yield of Lake Tanganyika which supports a similar predator-prey fishery. Higher MSY figures could mean that the stakes are even higher than previously indicated.

¹ If inputs are valued in economic terms rather than in financial terms - as they should in this context - the net benefits derived from the Lake Victoria's surplus production would likely be higher than half the net value indicated. Indeed labor represents a large part of input cost, and labor (opportunity) cost can be estimated to be close to zero as unemployment and under-employment are very high all around the lake

² Wanink (1989) estimated the 1982 biomass of R. argentea at 187 400 t and reports a four-fold increase in biomass between 1982 and 1987 in surveyed areas. Extrapolation to the whole lake give an estimated biomass of 750 000 t for 1987

³ It is assumed that the major productive waters and/or fishing grounds are confined to the 20 m or less depth zone for Nile perch (an area of some 12 700 km² or 18 % of the lake) and to half this area (6 350 km²) for O. niloticus

⁴ In comparison, Ligtvoet (1989) estimated the standing stock of Lates in 1987 at about 480 000 t for the whole lake

Table 4

Lake Kyoga: Total annual catches (t) 1961–82 (all species)¹

¹ Source: Uganda Fisheries Department

² Rounded to nearest 100. As noted in the text, there has been a drastic drop in the Lake Kyoga Nile perch catches since 1983. Recent observations indicate that the most likely cause of this decline has been overfishing, using small-mesh seine nets and that the Nile perch stock is now recovering (personal communication by Uganda Fisheries Department based on draft terminal report of IFAD/World Bank Lake Kyoga Project)

³ Mean 1961–75: 53 226 t

⁴ Mean 1961–82: 82 245 t

Within the next few years, it is likely that Nile perch catch will start falling and lead to the use of mesh size smaller than the 7–9 inch (177–228 mm) mesh which is now commonly used all around the lake. If this occurs one would also expect a certain proportion of effort to be rapidly transferred to the O. niloticus fishery. As the two fisheries are predominantly exploited by gillnets, and as the fish are relatively good substitutes, it is likely that any overexploitation process will affect both fisheries simultaneously. But it is difficult to know which one will be most affected. Significant transfer of fishing effort on R. argentea is likely to occur only at a later stage. Even if this fishery is undergoing a very rapid expansion in Kenya and Tanzania - a trend which is likely to persist as the fishery remains largely underexploited and as additional effort is therefore unlikely to significantly affect the profitability of already operating units for some time - a transfer into this fishery implies a significant change in equipment and may not necessarily be as profitable as remaining in an overexploited fisheries if prices for Lates and O. niloticus do increase significantly.

Judging from the little information which does exist on fish markets around the lake, it is likely that Lates and O. niloticus prices would increase significantly and jointly if the production of Lates or both species falls significantly. Indeed even if purchasing power is limited, these two species have few immediate substitutes and the price of Nile perch is likely to be driven up by the forces of the nowadays numerous and diversified markets on which it is sold. Market prices are therefore likely to reinforce the overexploitation process as higher prices would to a certain extend compensate for lesser catch rates.

While the whole process of rent dissipation might take well over a decade - period during which large benefits will continue to accrue - the fisheries under relatively open access would eventually stabilize at a presumably much lower level of production which may be close to the level observed in the sixties and early seventies. But such an evolution can hopefully be avoided through the introduction of proper management measures.

3.2 The Management Framework

There are presently very few management-related regulations controlling fishing effort on Lake Victoria and the few that exist are seldom enforced. According to van Densen (1989) these concern essentially:

• minimum fish size for Lates (48 cm) and tilapia (28 cm) in Uganda;

• minimum mesh size in the tilapia fishery of 10.2 cm in Kenya and 7.6 cm (in bays) for Tanzania;

• the prohibition of night fishing (mostly for security reasons) and of beach seining in Uganda;

• a closed season from 21/3 to 1/8 for beach seines and mosquito seines in Kenya;

• and closed areas such as for trawling in bays in Tanzania and for seining in Kisumu in Kenya.

Practically, few of these regulations are strictly enforced but notable exceptions are the ban on beach seining in Uganda and to a lesser extent, seining restrictions in Kenya. It is to be noted that all these restrictions would affect gear use and other qualitative components of effort and not nominal effort per se. It reflects the generally accepted conclusion that it is indeed the gear used and the fishing practices that have led to severe overexploitation under the previous fisheries regime (Marten, 1979). As effort has significantly increased over the last decade, this may no longer be entirely true. Even if overexploitation will eventually take the form of decreasing mesh size in the gillnet fisheries, it is doubtful that further investment into these fisheries is at all required. For other socio-economic reasons as well, this also concerns trawling and the expansion of industrial processing.

It is therefore likely that regulations should be introduced to control both the quantitative and qualitative dimension of fishing effort. Among the range of regulations which one may consider, priority should be given to those which are more readily acceptable and in a related way more easily enforceable. But both acceptability and enforceability depend greatly on how committed the three governments are in avoiding severe overexploitation. Whereas the stakes are a lot higher than they have ever been for Lake Victoria, imposing generally unpopular regulations is politically a difficult task and requires enforcement means far in excess of those presently available. However, the governments concerned should bear in mind that it is a lot easier to introduce and enforce regulations in a climate of profitability and relative prosperity than it would be in a few years when the fisheries become overexploited.

Also to be noted is the fact that while the actual effectiveness of any regulation depends on acceptability and enforceability, these two factors are in the final analysis greatly influenced by the government's own dedication to fisheries management. Some regulations can be made more readily acceptable through proper popularization and consultation with concerned communities. Similarly, more appropriate but less enforceable regulations should not necessarily be put aside by assuming that present enforcement capabilities may not be strengthened.

In further determining the framework of management, a major issue is whether or not managers can effectively control the very forces leading to overexploitation, especially in the case of gillnet fisheries for which mesh size reduction may eventually become the key factor. The engine of effort expansion is profitability. Obviously, the higher the profitability, the higher the incentive to enter, especially if entry cost is limited in terms of investment cost or otherwise. In the expansion phase of a fishery, one could therefore limit effort expansion either by controlling effort directly or by controlling profitability. It is actually best to control both simultaneously. Indeed the more profitable a fishery is, the more difficult it is to control effort directly as the pressure for entry is high and more difficult to resist.

Because effort/profitability control schemes are difficult to enforce, effort would likely continue to expand under such management schemes; but at a much reduced rate as compared to free entry. Furthermore as long as profits remain slightly above normal, the incentive to reduce mesh size is not as strong and mesh size regulations can be more easily enforced. Mesh size regulations should actually be introduced early in the process, i.e., well before mesh size becomes a binding constraint to “normal” profitability.

Simultaneously controlling (or constraining) effort expansion and reducing profitability should slow down the process of overexploitation preserving both the livelihood of fishermen and a higher supply to consumer's for years to come. Another advantage of this approach is that, if mesh size reduction becomes the only way for fishermen to ensure the “normal” profitability of their activity, the other constraints imposed on the fishery can be adjusted by simultaneously reducing effort and increasing profitability.

In this way, governments would also slow down the pauperization process which generally goes hand in hand with overexploitation. The logic is that whatever the government takes when profits are high, it can slowly give back when profits all to a low level and pressure to ignore key regulations builds up. But the strict control of effort would nevertheless remain the only way to guarantee high returns to fishermen on a sustained basis.

Mesh size regulations would be required in any case, but applied by themselves, these are bound to be far less effective. Indeed as effort is allowed to expand more rapidly at higher levels, a faster post-peak reduction in catch rates and profits can be expected, leading to a faster and less easily controllable process of mesh size reduction. This evolution is likely to be reinforced in small-scale fisheries like those of Lake Victoria as mobility in and out of the sector is thought to be quite asymetric. Furthermore, mesh size regulations become limiting at very low levels of profitability and thus more difficult to enforce because of the social pressure. This situation very much prevailed under the former fisheries regime, making mesh size regulations ineffective or politically unenforceable. The result has been the systematic overexploitation of the gillnet fisheries.

To avoid a similar evolution, the three governments concerned should act promptly and preferably in a concerted way to introduce simultaneous controls of all three components: nominal effect, profitability and mesh size. Short of acting promptly on all three fronts with much increased enforcement means and ample concertation with the communities concerned, it is believed that the downturn of the gillnet fisheries would lead to a fast and similarly uncontrollable process of overexploitation.

While greater benefits could have been derived by acting earlier, it can be argued that it is still timely to control entry into the gillnet fisheries. Indeed, even if the fisheries are more or less at peak-production, profits are still very high and will continue to be so for some years, attracting additional effort.

3.3 Suggested Immediate Management Regulations

An attempt is made below at defining an indicative set of regulations and measures for immediate consideration. It contains key elements of a short-term strategy aimed at simultaneously controlling nominal effort and selected qualitative components of effort in a direct or indirect manner.

Introducing too many regulations and measures may lead to difficulties with respect to both enforceability and acceptability. It is believed, however, that the same means of enforcement can be applied to enforcing several regulations and that the socio-political factors which generally limit enforcement effectiveness may partially be overcome by a multidimensional set of regulations and constraints.

- Mesh Size Regulations

Minimum mesh size should be introduced for both the Lates and O. niloticus fisheries. While a number of factors would have to be considered in defining these minima, it is recommended that they be set provisionally at the lower level of the mesh sizes commonly used nowadays, i.e., 6 inches (152 mm) for Lates, 4 inches (101 mm) for O. oreochromis. This way the regulation is not likely to become lineating short-term, giving time to the fisheries departments of the countries concerned to popularize the regulation and to strengthen their enforcement capabilities.

Restrictions on the sale of illegal mesh size around the lake can only be contemplated if gillnetting for other species is marginal. It is so now, but some species will recover as the abundance of Nile perch decreases. This has already been noted for the Nyanza Gulf (Siwo, 1988).

- Licensing of All Fishing Units by Main Fishery

Owners should be required to obtain a licence for each fishing unit/boat which they own and for each main fishery in which the unit operates. Canoes should carry official markings and operators should be required to carry proof of proper licencing at all times. While some requirements of the sort exist in all three countries, the system should be strengthened and fully operationalized so as to enable effective monitoring of effort in each major fishery (immediate priority) and eventually to control entry (short-term priority).

It is further suggested that gillnetting licences be given for only one target fishery. In the present context, this is a key element to enforcing mesh size regulation and will allow for control of mesh size independently of whether a unit has been fishing or not.

- Banning of Gear

Uganda should continue to strictly enforce the prohibition of beach seining as it is targetted primarily on juvenile Lates and O. niloticus in the country. In Tanzania, and even more so in Kenya, beach seines (and mosquito nets) are also used in the R. argentea fishery. Imposing more restrictions on the use of this gear in these two countries warrants further immediate investigation, taking into account that other open water gear such as ring nets and scoop nets are not only less destructive but reportedly more efficient and increasingly used in Tanzania (Ligtvoet, 1989).

Further investigations may lead to the recommendation of other regulations which will complement the above and existing ones. The recommended set of regulations should be relatively effective in achieving their purpose if penalties are swift, involving both confiscation and legal pursuit. As far as acceptability is concerned, licensing would be readily accepted if licence fees remain reasonable relative to profitability. Mesh size restrictions are generally well understood by fishermen, but would only be readily accepted if thoroughly applied. Gear prohibition can be made more readily acceptable by giving fishermen time to fully depreciate their gear and providing incentives to switch to other gear. It is also more readily enforceable than most other regulations. Mesh size regulations and licensing are about equally difficult to enforce, necessitating regular on-shore and on-water control over extensive areas and for a large number of fishing units. It is believed, however, that these can be effective even if strict enforcement is limited to key landing areas where effort is concentrated.

3.4 Suggested Complementary Management Measures

These measures are based on socio-economic considerations and concern economic incentives or disincentives which can influence effort allocation. In the short-term, these should aim at discouraging a further and economically wasteful expansion of fishing effort in the gillnet fisheries by reducing the profitability of these fisheries in general and relative to others.

- Avoiding Further Government Support to Gillnetting

This can take the form of Government discouraging new development projects that would directly or indirectly result in effort expansion in the Lates and O. niloticus fisheries. Especially concerned are projects which require direct government approval. New entry into these fisheries at small-scale level can also be discouraged by limiting cheap credit for gillnetting and avoiding the subsidized importation of gillnets through projects of other government programmes.

- Providing Incentives to the Transfer of Effort Away from Gillnetting

Through development projects and financial incentives, such as lower taxation, transfer grants or special credit schemes, the government may induce transfer away from gillnetting into long lining and open water pelagic fishing. The introduction of lift nets by some Lake Tanganyika fishermen in the southern part of Lake Victoria may provide further scope for the development of the pelagic fishery.

- Raising Taxes

A variety of taxes and fees are levied in all three countries: from registration fees, landing and marketing taxes to partial taxation of inputs. As noted by Reynolds and Gréboval (1988), these do not amount to a significant portion of total revenues, especially at harvesting level. Raising taxes at this level is not only unpopular but would affect statistical recording if applied to landings and increase “parallel” trade in any inputs it is applied to, which is not desirable. If taxes are to be raised, it should rather be done by progressively raising registration/licence fees at both harvesting and marketing levels. It should further be noted that these fees are now set at very low levels and that the fisheries administrations have direct control over these fees, whereas other taxes are generally levied at decentralized levels.

- A Ban on Further Development of the Trawler Fleet

Although trawler's catch remains very marginal, the number of trawlers operating in the fishery has increased significantly in recent years. In Tanzania, 13 trawlers are now operating compared to 3 to 5 during the 1970's; 5 trawlers operate in Kenya, and in Uganda a new Sino-Ugandan joint venture is soon to start operating 2 pair-trawlers. It was shown by Reynolds and Gréboval (1988) and in FAO (1988) that trawling does not make economic sense as it does provide far fewer socio-economic benefits than canoe-based operations.

This is especially the case as trawlers and canoes do exploit the same stock of Nile perch. Even if the catch of trawlers remains marginal, their operations will increasingly conflict with those of small-scale fishermen, as catch rates decrease.

- Discouraging Further Investment in Industrial Processing

This should be considered on a country by country basis. In Kenya, where industrial processing has developed rapidly, it is believed that the proportion of Nile perch catch going to existing factories is already very high. As the catch decreases, this proportion is likely to increase almost proportionally with significant adverse effect on local processors, distributors and consumers. In the longer term only long-established companies may continue to be profitable (their investment being fully depreciated). However, more recent and any additional investment into this activity are likely to be unprofitable. In other countries, there may be room for the development of a small industrial capacity if it can be shown that these would remain profitable at higher ex-vessel prices and if these do not compete directly with small-scale processors and distributors on their own market (a situation likely to occur in the case of the newly created processing plant based in Jinja, Uganda).

3.5 Other Management Issues

The management of Lake Victoria's fisheries would require on the part of the three countries a determination to act promptly and in a concerted way. An issue which may be raised is whether or not more in-depth studies are required before determining a management plan. Taking into account that the information base is sufficient to qualitatively assess the dynamics of the system and its likely reaction to management measures, it is felt that an interim management plan could be prepared on the basis of a thorough analysis of existing information. Complementary research priorities could be determined concurrently as a basis for fine-tuning the plan at a later stage.

Enforcement is a key issue to the elaboration and implementation of such a plan. Without additional enforcement means it is doubtful that the fisheries of the lake could be effectively managed. The strengthening of enforcement capabilities would require additional staff and logistical support as well as complementary training at field and managerial levels.

It is to be noted, however, that the approach to enforcement is as important as the means themselves. Full consultation between all parties is a prime requirement, and fisherfolk should not only be made to know and to understand the regulations but should also be progressively associated in their elaboration and implementation. For example, each country could establish a management council which, on a consultative basis, would associate fisherfolk representatives and local leaders in the elaboration and implementation of management measures. Local management committees could also be established to monitor the implementation of management measures and to ensure the fair and thorough application of the regulations.

The elaboration and implementation of a management plan of the sort described in this document is a fairly complex task. The three countries concerned may take advantage of the experience acquired elsewhere and seek external assistance to strengthen their management planning and enforcement capabilities. However, experience has shown the key constraint to fisheries management to be of a political nature as the introduction of restrictive regulations is bound to be initially impopular. For this reason, a key priority should be for the relevant fisheries institutions to organize the concertation and lobby for the political support required to enact and implement appropriate management measures.

4. REFERENCES

Butcher, D.A. and J.C.G. Colaris, 1975. A sociological survey of the fishermen population around Lake Victoria. Report prepared for the Lake Victoria Regional Fisheries Research Project. Rome, FAO, FI:DP RAF/71/242/4

Cadwalladr, D.A., 1969. A discussion of possible management methods to revive the Labeo victorianus fishery of Lake Victoria, with special reference to the Nzoia River, Kenya. Occas.Pap.Fish.Dep., Uganda (2):1–4

Committee for Inland Fisheries of Africa (CIFA), 1982. Report of the first session of the Sub-Committee for the development and management of the fisheries of Lake Victoria. Mwanza, Tanzania, 12–14 October 1981. FAO Fish.Rep., (262):73 p.

Committee for Inland Fisheries of Africa (CIFA), 1983. Report of the second session of the Sub-Committee for the development and management of the fisheries in Lake Victoria. Rome, 6–7 October 1983. FAO Fish.Rep. (301):20 p.

Committee for Inland Fisheries of Africa (CIFA), 1988. Report of the fourth session of the Sub-Committee for the development and management of the fisheries of Lake Victoria. Kisumu, Kenya, 6–10 April 1987. FAO Fish.Rep., (388):112 p.

Coulter, G.W., et al., 1986. Unique qualities and special problems of African Great Lakes. Environ.Biol.Fishes, 17(3):161–83

Dhatemwa, C.M., 1982. The utilization of Haplochromis spp. in Uganda. FAO Fish.Rep., (268) Suppl.: 102–6

FAO, 1973. Lake Victoria Fisheries Research. Regional: Kenya, Tanzania, Uganda, Interim report. Rome, FAO/UNDP, FI:DP/RAF/65/049:44 p.

FAO, 1988. Kenya - Tanzania - Uganda Lake Victoria Fisheries Development Reconnaissance Mission. Report of the FAO/African Development Bank Cooperative Programme. Rome Investment Centre No. 81/88 AF-KEN-URT-UGA 20 FAO.

Fryer, G., 1984. The Conservation and Rational Exploitation of the Biota of Africa's Great Lakes. pp. 135–54. In: A.V. Hall (ed.) Conservation of Threatened Natural Habitats. South Africa National Scientific Programmes Report 92. CSIR. Pretoria.

Goudswaard, P. and F. Witte, 1985. Observations on Nile Perch Lates niloticus (L.) 1758, in Tanzanian Waters of Lake Victoria. FAO Fish.Rep., (335):62–5

Greenwood, P.H., 1984. African Cichlids and Evolutionary Theories. pp. 141–54. In A.A. Echelle and I. Kornfield (ed.) Evolution of Fish Species Flock. University of Maine at Orono Press, Orono.

Kanyike, E.S., 1972. The Present-day Fishery of the Uganda Waters of Lake Victoria. Occas.Pap.Fish.Dep.Uganda, (4) :2–6

Ligvoet, W., 1989. Stock Assessment of Nile Perch (Lates niloticus) in Lake Victoria. Paper presented at the HEST/TAFIRI/FAO/DANIDA Regional Seminar on Lake Victoria Fish Stocks and Fisheries. Mwanza, Tanzania, January – February 1989.

Mann, M.J., 1970. A Resume of the Evolution of the Tilapia Fisheries of Lake Victoria up to the Year 1960. Ann.Rep.E.Afr.Freshwat.Fish.Org., (1969):21–6

Marten, G.G., 1979. Impact of Fishing on the Inshore of Lake Victoria (East Africa). J.Fish.Res.Board Can., 306(8):891–900

Nyholm, E. and P.S. Whiting, 1975. Fresh Water Fish Economic Studies in Kenya, Uganda, and the United Republic of Tanzania (1967–71). A report prepared for the Lake Victoria Fisheries Research Project. Rome, FAO, FI:DP RAF/65/049/2:77 p.

Ogutu-Ohwayo, R., 1985. The Effects of Predation by Nile Perch Lates niloticus (Linné) Introduced into Lake Kyoga (Uganda) in relation to the Fisheries of Lake Kyoga and Lake Victoria. FAO Fish.Rep., (335)18–41

Panayotou, T., 1982. Management concepts for small-scale fisheries: economic and social aspects. FAO Fish.Tech.Pap., (228):53 p. Issued also in French and Spanish

Reynolds, J.E., and D.F. Gréboval, 1988. Socio-Economic Effects of the Evolution of Nile Perch Fisheries in Lake Victoria: a review. CIFA Tech.Pap. (17):148 p.

Siwo, J., 1988. Nile Perch Fishery in Lake Victoria. In Technical reports presented at the project seminar on improved utilization of Nile perch, Kisumu, Kenya, 28–31 March 1988. Rome, FAO, GCP/KEN/055/NET: 93–103

Tettey, E.O., 1988. The Economics of Nile Perch Processing and Distribution. Report prepared for FAO Project GCP/KEN/055/NET, Improved utilization of Nile perch, May 1988. Abidjan, INFOPECHE, 33 p. (mimeo)

University of Leiden, 1986. The Decline of Lake Victoria's Cichlid Species Flocks: survey of the press publications. Zoologish Laboratorium, Leiden.

van Densen, W.L.T., 1989. Biological and Management Aspects: summary of discussions. HEST/TAFIRI/FAO/DANIDA Regional Seminar on Lake Victoria's Fish Stocks and Fisheries. Mwanza, Tanzania, January – February 1989.

Wanink, J.H., 1989. The Ecology and Fishery of Dagaa, Rastrineobola argentea (pelegrin) 1904. Paper presented at the HEST/TAFIRI/FAO/DANIDA Regional Seminar on Lake Victoria Fish Stocks and Fisheries. Mwanza, Tanzania, January – February 1989

Wanink, J.H., W. Ligtvoet and F. Witte, 1988. HEST/TARIFI Research in Lake Victoria: some preliminary results and their relevance for fishery management. Report presented at the National Seminar on Fisheries Policy and Planning, Dar-es-Salaam, 2–4 May 1988 (mimeo).