Make your own free website on Tripod.com

Abstract TD80

Coenen E.J., Paffen P. & E. Nikomeze, 1998, Catch per Unit of Effort (CPUE) study for different areas and fishing gears of Lake Tanganyika. GCP/RAF/271/FIN-TD/80 (En): 100p.

SUMMARY AND CONCLUSIONS

1. FS, CAS and CPUE results

The 1995 simultaneous FS revealed the presence of about 17000 vessels active in fishing and operating from about 800 landing sites. Traditional units are the dominant fishing type followed by liftnets and beach seines. Uvira (north-west coast, Zaire) is the area where the most dense fishing effort (principally liftnets and traditional units) is present when expressed as number of fishing units per km of shoreline. Next come Moba (south-west coast, Zaire) and East Coast and Mpulungu (Zambia) with a majority of traditional units. Least dense effort areas include Bururi and Makamba (Burundi), Rukwa (Tanzania) and Nsumbu (Zambia).
Converted in "traditional effort units" (TEUs), results show that especially the north and south ends of the Lake are subject to the heaviest fishing pressure, respectively by liftnets and by industrial/traditional units. In between the heavily exploited north and south ends there is a decreasing effort gradient from north to south.
Previous total annual catch estimates for the Lake have most probably underestimated the contribution of Zaire (estimated at about 90000 tonnes) to the total annual Lake catch which in recent years might have approached 200000 tonnes. Moreover, it is believed that the present period is one of reduced annual fish yields due to changed environmental conditions (cause: increased air temperature -- reduced wind speeds -- reduction of upwellings and other hydrodynamic phenomena -- reduced primary production -- lower fish yields).
Recent CAS estimates per country indicate that:

•

Burundi maintained its fish yield level of about 21000 tonnes in 1995. However, in 1996 - due to numerous closures of the Lake because of security reasons - the annual fish catch estimate dropped to about 3000 tonnes (1/7 of previous years) resulting in a more than doubling of the price of fish and in an increased importance of the traditional subsistence fishery.

•

Tanzania recorded lower annual fish yields in 1994-95 of about 55000 tonnes compared to 72000 and 80500 tonnes in 1992 and 1993, respectively.

•

Zambia estimated in 1994 a total annual fish yield of about 127OO tonnes (9100 traditional/artisanal and 3600 industrial). Because a continuous catch monitoring system is not in place, except for the industrial fishery in Mpulungu, no annual total catch estimates were available for 1993 and 1995-1996. Although the Mpulungu industrial effort in 1994-1996 remained almost constant, the industrial CPUE showed a declining trend, form 877 kg/fishing trip in 1994 down to 535 kg/trip in 1996. This is an indication of local overfishing by the industrial units in the pelagic fishing grounds around Mpulungu, especially of the Lates stappersii stock, the dominant species in the catches.

•

Zaire has no CAS monitoring system in place for its part of Lake Tanganyika. Based on extrapolated fishing effort counts (1995 FS), a possible annual fish yield of 90000 has been estimated. Some local CAS estimates for the industrial fisheries in Kalemie and Moba are presented.

More detailed CPUE estimates, especially as to the clupeid species composition, were obtained during the fish biology sampling programme (SSP, 7/93-6/96) for different fishing gears and fishing areas.

•

Liftnet catches show an increasing CPUE trend from north to south, mainly caused by L. stappersii dominating the catches as from Kipili southwards at the cost of Stolothrissa tanganicae. Strange enough, the use of liftnets in the extreme south in not very popular because these units are apparently not very safe during rough weather and windy conditions. Some liftnet CPUE correlations were found between monthly total and/or species CPUEs of adjacent fishing areas in the north (Bujumbura, Uvira, Karonda, Kigoma). An attempt was made to correct liftnet CPUEs for different numbers of fishing lamps used per liftnet unit in different areas around the Lake.

•

Detailed CPUE characteristics are also presented for beach, kapenta, chiromila and purse seines.

2. CPUE: a measure of abundance?

Starting from catch and effort estimates (originating from CAS, FS and fish biology sampling data), CPUEs for different fishing gears, used in different fishing areas of Lake Tanganyika and measured over different time spans (month, trimester, year, 3 years SSP period) were presented. Emphasis was put on new data from LTR's SSP period (7/93-6/96) which were not yet analysed and presented in earlier Technical Documents. The latter period is the most important one because an important number of data were collected in different disciplines during this period. The acoustic surveys with R/V Tanganyika Explorer were also started during this period.
CPUE results might prove to be a tool for comparison with the magnitude and characteristics of the spatial and temporal biomass estimates obtained during the acoustic cruises. CPUEs do indeed reflect - in certain cases - the magnitude of abundance of fish stocks. However, as mentioned before, in the case of Lake Tanganyika, the CPUE results are not de facto a measure of fish abundance and should therefore be handled with extreme caution. And this because of several reasons: the species in question are mostly fast swimming, shoaling and migratory species and in - the case of S. tanganicae - with a very short longevity and thus a fast turnover of the stocks. On top of that, they have a patchy distribution and the CPUE estimates originate from fishing units using light attraction that concentrates the fish towards the fishing gear used. CPUE estimates might thus not reflect the real natural abundance of the species considered.
CPUEs have another disadvantage: they are averages of catch observed over a certain time span and in a specific geographical area and have therefore levelled out certain variations which might have occurred during the time span in question. For example, it has been observed that - for the same fishing type and in the same fishing area - daily unit catches change e.g. from a few kilograms today to nearly one tonne the day after.

3. The influence of environmental conditions on CPUE

Apart from the effects of the fisheries and the prey-predator relationship (the latter which previously was thought to have a dominant effect on prey and predator abundance) on fish stock abundance, we have seen in chapter 5 that a number of other factors are very important in the regulation of species abundance, especially for the clupeids. They include intrinsic biological factors but also a number of environmental factors which are probably as important if not more important in regulating species abundance and distribution in the short and long term.
Environmental factors include temperature, wind speed, upwelling, turbidity, hydrodynamic phenomena, etc. and are all interrelated. Some effects of these factors on clupeid spawning success, larval survival, recruitment, etc. were already presented in chapter 5. An example is the situation in recent years for Lake Tanganyika: higher air temperatures (possibly influenced by El Niño events (ENSO) in the Pacific Ocean), lower wind speeds (especially the dry season south-east winds), less tilting of the Lake volume, reduced hydrodynamic phenomena (oscillation amplitude, internal waves, upwelling, etc.), lower turbidity when less upwelling, lower primary production through lower nutrient availability, lower food availability and recruitment (less and smaller eggs, reduced survival of larvae, etc.), reduced fish population size and catches, etc. Although the environmental interrelationships might have been presented a bit simplified in previous example, it does reflect the tremendous effect of environmental conditions on the fish habitat and thus on the CPUEs of the exploited fish stocks.
Related to the above, and also influencing fish stock abundance and CPUE estimates, it was also observed that:

•

small Lates stappersii (up to 10 cm) are found together (as observed in Bujumbura fish biology samples) with shoaling clupeids displaying a similar non-predator behaviour,

•

adult Lates stappersii, a high visual predator, favour areas without windy conditions, when turbidity is low and visibility high,

•

clupeids favour areas with windy conditions, with high turbidity and thus low visibility (disadvantage for its predator L. stappersii), often coinciding with higher nutrient availability and primary production.

4. Future monitoring of catch/effort fishery statistics

The LTR project, in collaboration with the 4 riparian countries, is now preparing a Fisheries Management Plan for the fisheries on Lake Tanganyika. The proposal will be based on the results of the multidisciplinary research activities executed during recent years on Lake Tanganyika.
Apart from specific measures (e.g. standardisation, limiting or banning of certain practices, promotion of other practices or activities, etc.), LTR will also propose and support the execution of a continuing Monitoring Programme. In this way, the necessary data to follow up the evolution of the Lake characteristics, including catch/effort and other fishery statistical parameters can be obtained. This will not only measure the impact of the implemented management measures but will also allow to adjust (cancel, introduce other measures, etc.) the management measures already in place. The Lake, its environment and fish stocks are indeed subject to very rapid changes which demands a continuous monitoring and assessment of implications of the measures in place.
Part of the Monitoring Programme will consist of the follow up of the catch and effort evolution in different areas of the Lake. It is therefore recommended that:

•

the riparian countries, in collaboration with LTR or any other project coming into effect to implement the management/monitoring programme, should reinstate, sustain and even reinforce their efforts to execute adequate and continuous catch/effort surveys (continuous CAS, FS minimum every 2-3 years),

•

these countries continue their efforts to adopt similar and standardised methods for collecting catch/effort statistics and at least produce compatible reporting outputs of their annual fisheries statistics (according to the adopted recommendations of the Fisheries Statistical Co-ordinators Meeting, see Coenen 1994b). The increased computerisation and use of standardised fisheries statistical software packages can only support this not avoidable trend,

•

these countries increase their efforts to create (in case they do not yet exist) or reinforce existing fisheries statistical units, not only competent in the planning and follow up of the execution of FS, CAS and other surveys but also in data analysis and presentation of the results obtained,

•

these countries give their full support to the Management Plan and Monitoring Programme to be implemented, especially in the field of fisheries statistics,

•

the planned Monitoring Programme, especially the fisheries statistics component, and within the limits of the available budget to maintain the programme, would consider:

•

to give its full support to the riparian countries in the execution of their CAS and FS surveys,

•

give assistance in the training and development of their respective fisheries statistical units,

•

support efforts for standardisation of fisheries statistical strategies/methods/outputs and for regular meetings between the fisheries statistical co-ordinators of the respective countries,

•

support the execution of complementary continuous CAS surveys as was done in the case of the industrial fisheries in Kalemie and Moba (Zaire),

•

maintain the collection of additional catch/effort (CPUE) statistics for specific gears as was done in combination with the fish biology sampling during the 1993-96 SSP period. The intensity and frequency will of course depend on the planned fish biology monitoring sampling programme but should preferably maintain - for each type of fishing gear to be monitored - a frequency of sampling 4 units every week (minimum 4 units every 2 weeks). As during the SSP, for each unit sampled, place and time, unit characteristics (type, number of lamps, hauls), total catch and catch per species estimates (using catch subsamples) are to be determined.

•

types of fishing units sampled should not only include liftnets, industrial units, kapenta seines, beach seines but should also include traditional fishing units as they constitute the dominant type of fishing on the Lake, followed by liftnets and beach seines.