Previous Page Table of Contents Next Page


2.1 Data- poor yes, but not data-less

Many African fresh water fisheries can be considered as data-poor (Mahon, 1997; Johannes, 1998): management decisions have to be made based on little information on drivers of, pressures on and states of fish stocks. Nevertheless, often long-term series of catch and effort monitoring are available, obtained through Catch and Effort Data Recording Systems (CEDRS) mostly developed in the seventies and eighties. The maintenance of a CEDRS is difficult due to the high costs involved, both in the data collections themselves as well as in maintaining the institutional set-up and the knowledge base. As a result, data collected are often considered too poor to be utilized in formal stock-assessments, while the information from assessment results is often qutance, the value of single figure estimates of long-term sustainable yields (e.g. MSY) based on steady state assumptions of environmental stability is questionable when large changes in productivity of fish resources occur, as a result of environmental drivers such as seasonal, annual and longer-term fluctuations in water level. This has as a number of consequences:

The perceived inaccuracies of monitoring data collected, as well as frustrations over the applicability of standard fishery science, have lead to a severe under utilization of the quantitative and qualitative information present in the catch and effort time series collected. Despite that, data continue to be collected and in quite a number of cases it is now possible to re-construct sometimes long time series of catch and effort (e.g. Zwieten et al., 2002). Much can be gained from maximising the use of existing information and knowledge on the performance of a fishery by evaluating trends and variability in catch and effort time series.

In addition effects of naturally changing or fluctuating environments on fish production should be taken into account to assess the effect of changing fishing effort. In floodplain environments the evidence of environmentally driven fish production is overwhelming (e.g. Lae, 1995; Junk, 1996; Hoggarth et al., 1999a), but in recent accounts attention is drawn to the regulation of fish stocks in small and medium sized lakes by changes in productivity through fluctuating water levels (Kariba: Karenge and Kolding, 1995; Turkana, Kolding, 1992; Chilwa Kalk, McLachlan and Howard-Williams, 1979; Zwieten and Njaya, 2003 and other lakes Lévêque and Quensière, 1988; Talling and Lemoalle, 1998) or even in large lakes as Lake Tanganyika through other, not yet fully understood long-term environmentally driven processes (e.g. Spigel and Coulter, 1996; Zwieten and Njaya, 2003; Sarvala et al., 1999). Changes and fluctuations in environmental drivers as water levels can be important indicators for changes in stocks in an adaptive management context, as water levels are easily measurable, with often long times-series of data present.

Compared to surrounding countries, the Malawian information base for fisheries management is well established. Past monitoring of stocks and of some environmental parameters has yielded useful information that is available. Also specialized studies on all important fisheries in Malawi has provided biological information on a species level useful for stock-assessments (e.g. Kalk, McLachlan and Howard-Williams, 1979; FAO, 1993; Tweddle, 1995; Palsson, Bulirani and Banda, 1999, Jambo and Hecht, 2001, Banda et al., 2002).

2.2 Evaluation of effectiveness of fisheries resource management in highly fluctuating environments

The evaluation of the effectiveness of fisheries resource management is largely dependent on the possibility to perceive changes in indicators for stock abundance and relate these to measures taken. Time trends in fish stocks as reflected in time series of catch rates are the basis for such evaluations. The possibility of measuring success of resource management actions within the appropriate time window depends both on the strength of effect over time and the variance around it (Peterman, 1990; Pet Soede et al., 1999; Densen, 2001; Zwieten et al., 2002). In this paper we will:

In statistical terms, the capacity to detect a trend is determined by the statistical power of the information examined, which in turn depends entirely on the variance of the data, given the number of observations and statistical decision levels. The time series of estimated monthly catch rates from CEDRS surveys of Malombe by major stratum (i.e. East and West Malombe) represent the lowest level of data aggregation for this lake that are used in reports on the status of the fishery. The capacity to detect trends and evaluate the effects of resource management with the aid of these time series is therefore dependent on the variability within these series, given the present sampling and data handling methods in use.

A justification of the research strategy followed in this study and the methodology of analysis of catch, catch rate, effort and water level data (analysis of variance, trend analysis, trend-to-noise analysis, multiple regression analysis) is outlined in Zwieten and Njaya, 2003. A description of the data sets as well as a justification of the aggregation into species groups can be found in Appendix 1. Appendix 2 describes the method of reconstruction of effort data. In Appendix 3 we give an analysis of the error structure of the data, with particular attention to the causes and effects of administrative errors on trend evaluation.

Previous Page Top of Page Next Page