Recently there has been a great upsurge in interest in length-based methods of assessing fish populations. The impetus for this has come from at least three sources - the increasing problems of applying the better known age-based methods (especially in tropical areas where fish do not carry easily-deciphered birth certificates on their scales or otoliths); the development of improved methods of analysing length data; and the increased availability of computers, especially desk-top microcomputers, that put within the reach of all the computational power needed to take advantage of some of the new methods.
This interest gave rise to a meeting in Mazara del Vallo, Sicily, organized by the International Centre for Living Aquatic Resources Management (ICLARM) and the Kuwait Institute for Scientific Research (KISR) in February 1985. This meeting brought together many of those working on length-based methods, and the report of the meeting (Pauly and Morgan, 1987) put on record many of the methods being used, some of which were previously available only in the grey literature.
From the meeting report it is clear that length-based methods have come of age, and represent a very valuable set of weapons for stock assessment scientists. They should not be considered as a second best alternative to age-based methods. Most length methods correspond very closely to similar age-based methods. They make the same assumptions about how fish populations behave, and depend to the same extent on the validity of those assumptions. Length methods have the operational disadvantage that they use an awkward time scale. Against this, the size of a fish may be more closely related to matters like food requirements or predation than its age. With the recognition of the usefulness of length-based techniques comes a corresponding need for some general guidance on how to set about using these methods.
In fact there are two needs. One is for technical guidance on how to apply one technique or another. This is now well provided for by the recent FAO manual (Sparre et al., 1989), based largely on the experiences of a number of FAO/DANIDA training courses. The other need is for a more general overview of the various techniques which can provide some guidance as to which methods to use. This report attempts to fill in this latter need. It is set out in four parts.
The first part deals with the collection of data. One of the main advantages of length-based methods is that the data are easy to collect. All that is needed is to go to the fish landing place with a measuring board and pencil and paper; and a major aim of this section is to encourage people to do more sampling and to collect length data from as wide a set of samples as possible in terms of types of gear, areas fished and times of year. To do this efficiently, guidance is given on sampling design, and on methods of recording and compilation.
The second part deals with the heart of length-based methods, i.e., the various techniques that have been used to estimate the vital parameters of fish population dynamics - the growth and mortality rates and the patterns of selection and recruitment. Each of the main methods are described, with their advantages and disadvantages, and enough information is generally given to enable the reader to apply the methods. However, to save space where the methods are complex and have already been described in published form, and especially where computer programs for using the method are available on request, only summary descriptions are given. This particularly applies to the ELEFAN suite of programs, where ICLARM has been very helpful in making these programs readily available to scientists in developing countries.
The third part deals with the actual work of assessment, i.e., bringing together these parameter estimates with other relevant information (especially statistical information from the fishery) in order to assess the state of exploitation of the fishery - is the stock heavily fished or not? - and to provide advice to fishery authorities on the potential for expansion or the need for management. Descriptions are also given for assessing the impact of possible management measures, such as increases in the size of first capture, some of which are particularly well handled by length methods.
Finally, guidance is given on what methods are likely to be most useful under different circumstances. Individual techniques are not equally useful in all fisheries, and in a given fishery only some (or in the worst case, none) of the methods are likely to be useful. This will depend on the pattern of growth and recruitment, the lifespan of the fish, and the selectivity of the fishery. Fisheries can be placed in different types, and a few length samples will show what type a given fishery is, and thus what methods should be considered. This final section also emphasizes the need, whatever method is used, for considering the reliability and precision of the estimates obtained, and the sensitivity of the results to the assumptions made and to the values of the inputs.
