Methodologyin this note is used as a generic term encompassing mechanisms, processes, techniques, approaches and working concepts, all developed and applied when the need arises to collect data in order to manage the economic and social aspects of a given activity, in this context "fisheries".

Looking at fisheries in a global perspective requires that each national fisheries statistical programme will become a component of an international dataset of fishery statistics. Such integration of different datasets (each created through national programmes), is in general feasible when national systems utilize a common regional or inter-regional set of statistical standards, and apply internationally recognized definitions, classifications and codes.

There are a number of issues that affect the quality of international fishery statistics. The first concerns data coverage. In several cases available official data are incomplete in terms of range of variables and sectors covered. Consequently, records in the international database represent not only data returned from countries but estimates that are based on best possible assumptions. A second issue relates to differences in statistical concepts and methodologies used by countries for collecting the data. Methodological variations often make comparability of statistics a difficult task, especially when the degree of accuracy among national data is uneven. A third issue concerns the quality of primary data. Although this is a national responsibility it nevertheless has a strong impact on the overall quality of international statistics. Thus, improving data quality and protecting the credibility of international statistics requires interaction with countries and resources invested in capacity building at national level. With regards to the latter, political willingness of the national authorities is necessary for the adoption of measures deemed necessary to improve data quality.

The diverse economic condition in countries, the varying cost of systematic data collection, and the different amount of resources made available for establishing and maintaining a fishery statistical collection system or for improvements to an existing system, necessitate the adoption of firm priorities.

In general, it is preferable to invest in the improvement of the existing system rather than to set up a new independent data collection system. Where the industry is sufficiently organized with centralised marketing facilities or auctions, it is possible to obtain reliable records of the total catch and value without the need for sampling and recording directly, except for occasional checks. Where no records are available, the collecting system must start with a complete survey to obtain the most basic information on ports, landing sites, numbers and type of vessels, and sample surveys to identify species landed and their utilization.

Data types and variables on which information needs to be collected should be tailored to the structure of national fisheries and an overall strategy is required to decide on the data collection methods. As with other types of statistical series, the methods most commonly used include surveys, based on either complete enumeration (e.g. vessel registers, infra-structural data, aquaculture production units) or on sampling. Complete enumeration is expensive, but may be essential for certain data sets. Sampling is more cost-effective, but caution is required in designing the distribution of sampling effort in time and space.

Experience has shown that a national statistical programme normally uses both census-based and sample-based schemes for data collection. The first applies primarily to industrial and semi-industrial fisheries, where data are directly obtained from the operators themesleves, usually as part of the licensing process. The second approach is used for the artisanal fleets. These are often characterized by high dispersion combined with large size, thus making impossible the collection of information on a complete enumeration basis.

The variables on which information is collected differ according to the main objectives of national fishery policies. Often one or more of the following objectives apply:

1. to produce enough fish for domestic requirements

2. to develop exports

3. to improve the socio-economic conditions of fishers

4. to promote expansion of fisheries

5. to formulate indicators relating to abundance, fishing pressure, gear selectivity and fleet performance

6. to develop aquaculture production


The main topics of any survey methodology would include:

1. design issues in the context of practical constraints

2. use of different data sources and collection techniques

3. total survey error

4. survey evaluation

5. time series analysis

6. seasonal adjustment

7. data integration

8. estimation and data analysis methods

9. general survey systems development

Until fairly recently the collection and compilation of fishery statistics was heavily biased towards the data required for management of biological resources. With the general acceptance that there is an overcapacity of fishing fleets, a better integrated approach to fishery management has to be developed, taking into account the biological, environmental, social and economic aspects of fisheries. The CWP will be required to develop the concepts and definitions for the parameters to assess these additional aspects of fisheries and to determine how these additional elements can be assimilated in the existing statistical programs. The lack of adequate and reliable capture fishery and aquaculture statistics for many countries, and for developing countries in particular, are issues that CWP may have to address. It could provide guidance on the establishment of cost-effective and sustainable data collection schemes and methodologies, and may also be influential in motivating donor-support for fishery statistical development, projects and training.

Bibliography