Analyse et modelisation des parametres d'exploitation du stock de coquilles saint-jacques (pecten maximus, l.) en baie de saint-brieuc (manche ouest, france).

The population dynamics is developed upon data on the age-composition of catches collected since 1974 by a sampling technic at two levels (fishing day, vessel). A temporally stratified sampling is carried out. A cohort analysis is realized through these data ; assumptions for the coefficients of natural and indirect fishing mortality are set up. Catchability models for all age groups are fitted. Catchability of the most abundant age groups (2 and 3) is an increasing function of their abundance and of the fishing fleet mean horse power. These functions present a maximum asymptotical value. Their adequation is explained by the adaptations of dredges and by the population sedentary life allowing the concentration of fishing fleet on the most abundant areas. A structural model of production based on an equilibrium state is carried out. The most likely translations of required parameters are tested by reconstitution of yields per recruit or annual landings. Low values of instantaneous coefficient of natural mortality (0,15 and 0,20) and high values of terminal rate of exploitation (0,50) optimize the reconstitution. Isopleths of yield and fecundity per recruit are presented for different values of the recruitment abundance and the fishing fleet mean horse power. At present, parameters of fishing pattern control (nominal fishing effort, age at first catch) give an equilibrated production near to optimum as well as an acceptable fecundity. The present decrease of the stock is explained by interannual variability of the recruitment and by coexistence of several métiers* which causes a constant increase of vessels horse power. The sampling technic used since 1986 for directed stock assessment is analysed. These prospections are necessary for the annual determination of total allowed catches. The bay of Saint-Brieuc is divided into six spatial strata ; sampling rate for each stratum is defined by the corresponding commercial landings. The estimation of abundances needs calculation of the dredge efficiency through diving. Relative precision of abundance and biomass for each age group is 25 to 30 %. An increase of the samples number would give a small gain of precision. On the other hand, an increase of the diving number would allow a significant decrease of bias, due to the definition of efficiency as a rate estimator. Probable existence of high values of natural mortality between 1986 and 1988 is studied. Natural mortality was perhaps higher during last winter period of these years. Nevertheless, it is impossible to dissociate natural mortality from components due to the exploitation (not declared landings). It is not demonstrated that this eventuality would have provoked a fishing intensification. The annual management through short terms projections is limited. Spatial variability of growth and interannual variability of recruitment should be taken into account in the roodclization system in order to take advantage of the knowledge of the equilibrated structural analysis. Growth shows a spatial variability higher than the temporal one. No used method of comparison of growth curves is really perfect, but growth seems quicker in southern bay. Estimations of the gTowth parameters are different from those obtained by prcsious studies. Spatial variability of growth should be taken into account when a rotation of fished areas is considered or when an extensive aquaculture is developed. Study upon the recruitment is realized through an empirical model which explains its interannual fluctuations as a function of spawning biomass of the stock, of temperature and of fishing effort developed during first winter of the cohort life. An incorporation of the model to a management system with mean terms projections is carried on. Temperature seems to be the most important factor if spawning biomass reaches an adequately high level for a good reproduction.