Lakes, reservoirs and wetlands cover a total area of about 7.8 million km2 and provide a rich environment for inland capture fisheries. Production from the world's inland capture fisheries has grown steadily to over 11.6 million tonnes in 2012, with almost 95% of the catches from developing countries. The sector is composed primarily of small‐scale fishers and provides employment to approximately 61 million people, of which 50% are women. Inland capture fisheries yield is only 6.3% of the global reported fish production (capture fisheries and aquaculture). However, it is proposed that this is an underestimate and actual yield may be several times greater. The apparent low proportion of fish provided by inland capture fisheries globally does not reflect adequately the importance of inland capture fisheries in today's society. A lack of accurate information renders analysis of existing data sets difficult and makes decision‐making problematic. Is the sector in serious trouble because of the multiple uses of and threats to inland water ecosystems? Is the sector stable and increased production due only to better reporting? Or, is the sector growing?

In order to obtain an appropriate and reliable method for the mass production of Clarias gariepinus (Burcheli) fingeriings. experiments on artificial reproduction and pond rearing were carried out in the Republic of Congo in 1987-1991. Reproduction could be Induced throughout the year using common carp pituitaries (3 mg kg'' female). The average relative fecundity of females varied between 1.3% ± 0.3 (SEM) and 14.37o ±1.3 (SEM). The average hatching percentage of the eggsvariedbetween28.4%±4.5(SEM)and59.1%± 3.7 (SEM) respectively, in the dry and rainy season.The fluctuation of the relative fecundity as well as the fluctuation of the hatching percentage follows the seasonal fluctuation in rainfall and temperature. The use of a net cage made of mosquito netting (1 m'. 0.5 mm mesh size) and the roots of water hyacinth (E. crassipes) as an egg incubator proved to be reliable. In ponds not protected against frogs an average of 5.0 ± 2.9 (SEM) fingerlings m"^ per 40 days cycle were obtained, whereas in ponds completely surrounded by a wall of aluminium roof plates (0,8 m high) the average production was 32.3 ± 3.3 (SEM) fingeriings m"-per 40 days cycle, when they are stocked with approximately 100 larvae m'. Increasing the larval stocking density or extending the rearing period did not improve the final production. The main causes of low production in unprotected ponds are competition for food resources due to the presence of phytophagous frog larvae and cannibalism among the fingeriings of Clarias gariepinus. An analysis showed that the system is labour orientated, technically reliable and economically feasible when the fingeriings can be sold for US$ 0.07 apiece.

From 1999 to 2011 some regions in the Lake Tanganyika Basin experienced humanitarian crises that displaced hundreds of thousands of people to neighboring countries. When relative calm returned to the region in 2008, an influx of displaced peoples and refugees returned to the lake seeking their ancestral fishing grounds. Well-meaning non-governmental organizations and United Nations-organizations donated fishing equipment to these returning people to aid their livelihood opportunities. These fishing programs, however, increased uncontrolled fishing effort on Lake Tanganyika beyond that of previous levels, resulting in decreasing fish catches. Increased monitoring of the fishery, therefore, became essential. In 2009, as a result of uncontrolled fishing effort due to the influx of returnees, inefficient national efforts to monitor their fisheries, and the observed decline of fishery resources, the four countries bordering Lake Tanganyika (Burundi, Democratic Republic of Congo, Tanzania and Zambia) established a regional coordinating body called the Lake Tanganyika Authority to implement fisheries conservation and management measures in compliance with the Convention on the Sustainable Management of Lake Tanganyika. To inform and enable a fisheries monitoring program on Lake Tanganyika, the Authority conducted a lake-wide fisheries frame survey in 2011 to inventory the number of fishermen, fishing units and fisheries infrastructure around Lake Tanganyika. When comparing the 2011 frame survey with data from a similar survey conducted in 1995 (the two most extensive studies to date on Lake Tanganyika), results revealed troubling trends in fish capacity, including: an increase in illegal fishing gear, a doubling of the total number of fishermen and fishing units, and a decline in catch rates since 2002 (based on Burundi data which has been consistently collected). This article analyzes the trends of the Lake Tanganyika fishery, including: fishing effort, the changing uses of gear, and trends in employment in the fishery. Because of the observed increase of fishing capacity (e.g. the numbers of vessels, licenses and fishermen), this article addresses whether an effective fishery management program can be implemented on the lake. Past management efforts have been made from within the basin by the individual countries (Burundi, Democratic Republic of Congo, Tanzania and Zambia), transnational organizations (Food and Agriculture Organization), and the Lake Tanganyika Authority. Using current notions of fishery management on large lakes in the region and ideas from a case study from Gambia, West Africa, this study suggests that effective fishery management on Lake Tanganyika requires the adoption of a formal Monitoring, Control & Surveillance system, community surveillance, an improvement in licensing systems, and a limitation in the number of fishermen and fishing units.

Larval fish drift and distribution patterns were studied in the River Lohajang, a tributary of the River Jamuna, Bangladesh, during the monsoon seasons (June ‐ October) of 1992, 1993 and 1994. Larval fish drift peaks with the water levels of the River Jamuna. In all three years, the Indian carp, Catla catla, Labeo rohita L. and Cirrhinus mrigala, were associated with the first pulses, but were absent from drift samples after September. Hilsa species and other non‐identified fish species were found throughout the monsoon season. The highest larval fish densities were found in the upper strata of the river, near the shore. The drift densities of these fish became evenly distributed through time. Water turbidity probably influenced the vertical and horizontal distribution, and the absence of diel patterns. The most important species in the larval drift were Hilsa spp., followed by Labeo rohita L., Glossogobius giuris L. and Rasbora daniconius. Among the major carp, Labeo rohita was the dominant species, contributing 80‐90%. The impact of flood control programmes on the reproduction of riverine and larval fish drift is discussed.

A simulation model for the production of the Nile tilapia (Oreochromis niloticus niloticus L.) for mixed-or mono-sex culture and for poly-culture with the African cat¢sh (Clarias gariepinus Burchell 1822) or African snakehead (Parachanna obscura Gˇnther 1861) is presented. The model is based on an exponential decay model used in population dynamics and follows an individual-based approach. The model consists of a tilapia sub-model representing population dynamics, growth and recruitment and a predator sub-model representing the predation process and population dynamics and growth of the used predator. The model was calibrated with data on mixed-or mono-sex culture of Nile tilapia and for poly-culture with the African cat¢sh or African snakehead obtained in Congo Brazzaville and validated with similar data from the Philippines, Thailand and the Ivory Coast. The model visualized major underlying processes in tilapia farming and aspects for further improvement of the model; growth is one of the most sensitive input parameters and should be quantitatively related to feeding level and feed quality; length at ¢rst maturity and quanti¢cation of the recruitment of Nile tilapia has a relatively large in£uence and recruitment should be related to the length of the females; preyp redator relations are too coarse and should be more ¢ne-tuned with the relation between prey size and predator size. Incorporation of these features would provide the basis of a model that can serve as a predictive and decision-making support tool. Aquaculture Research, 2005, 36, 455^471 Simulation of Nile tilapia culture in ponds G J de Graaf et al.