The impracticability of managing the complex multispecies resources on a species-by-species basis has already been pointed out by Larkin (1977). Tyler (1982) following this point of view proposes to develop an adaptive management approach based on “species assemblages production units” defined as a “group of trophically coupled resident species”. He proposes to introduce “fish assemblages maintenance goals” in to the management paradigms.
His interesting proposal stems from the assumption that the constitutive emergent properties of a species assemblage will be similar for similar assemblages despite changes in the individual components, and different from the properties of the individual species components. The emergent properties we are interested in are productivity, stability, resilience to fishing,etc.,and Tyler underlines that it may be possible to build surplus production models of assemblages analogous to production models of single species populations. Very recently Ralston and Polovina (1982) successfully used this approach.
It may well happen that trying to “maintain” assemblages structure under fishing pressure is just as impossible as trying to maintain constant biological production under the MSY concept and species interaction within the assemblages may not be a strong structuring mechanism as is usually assumed (Larkin,1982). It remains however that recurrent species assemblages do exist in the tropics as well as in temperate waters. Similar assemblages with analogue species occur in similar environments. Moreover they often support similar exploitation schemes, with similar gears. It is therefore a very logical step to move from the species-by-species to the assemblages approach which offers much more scope for practical implementation of management and development and may be the appropriate scale of aggregation of the data.
This paper provides for a first step towards the approach by mapping tropical demersal fish communities in a few countries of the Gulf of Guinea. Species assemblages are very often definable solely on the basis of geographic distribution and depth (because environmental parameters gradients are usually also related to depth) possibly complemented by bottom characteristics (level, rocky) and sediment types (mud, sand, gravel, etc.). As data are often very scarce in the developing world, this paper tries also to demonstrate that usable information (for management/development) can be generated for an area where knowledge is scarce, if something is known on its hydrography and bottom type and if the species assemblages have been identified in comparable neighbouring areas.
It is necessary to point out that in view of the limited amount of information available and the frequent need for personal interpretation when defining the distributions of species assemblages and individual species, the maps produced in this way might not in some cases reflect the exact average distribution of the species and species assemblages, which is changing slightly from season to season anyway. Indeed, the maps must be considered as working documents only, summarizing one's best knowledge on species assemblages and their distribution; every attempt should be made to check their accuracy by actual survey. It is however felt also that as they are, these maps can already be useful in giving information on species groups availability and distribution. Added to other ancillary information collected in neighbouring areas about seasonal patterns, for instance, or fishing methods, they can be an aid in planning fishery development and management.
