Les elevages ont ete realises dans deux types d'eau : une eau marine et la meme eau stabulee dans un vivier. Les eaux d'origine sont caracterisees par la diversite des peuplements et par le faible volume des cellules. Seul un faible pourcentage de bacteries (de 1'ordre de 1%) est cultivable sur le milieu de culture 2216E. L'addition d'antibiotiques, la presence des larves et des algues modifient profondement l'environnement des bacteries, ce qui se traduit par une evolution des peuplements : diversite reduite, augmentation de leur volume cellulaire, changement de metabolisme, resistance au chloramphenicol. L'analyse taxonomique ne fait pas apparaitre de souches caracteristiques de l'une ou l'autre des eaux, ni a l'origine ni au cours de l'elevage. Cependant des differences dans les potentialites de reponse a l'addition de substrat organique et de productivite sont identifiees selon l'origine des eaux. Ces differentes etudes ont permis de mettre au point un modele qui illustre l'influence de la productivite propre a chacune des deux eaux, des densites bacteriennes initiales, des apports de matiere organique et des phenomenes de grazing.

Larval rearings were performed with seawater either freshly pumped or after storage in a pond. Concerning bacterial populations these waters were characterized by a high diversity, little sized bacteria whose few of them (1%) were cultivable on Zobell medium. In the tanks, the bacterial ecosystem was modified by the presence of antibiotics, larvae and algae: the diversity was lowered, the cellular volume was increased by a factor 2 to 5, 100% bacteria were cultivable and chloramphenicol resistant, and their metabolism changed. Taxonomic screening showed no differences among the two waters neither at the beginning nor at the end of the rearings. However, depending on the water (stored or not) differences were identified for bacterial physiology and productivity: bacteria from the pond were more organic matter dependent. From these results we have developed a mathematical model which takes into account the influence of the respective productivity of both waters, the initial bacterial concentration, the organic matter content (seawater content and larval excretion) and the bacterial grazing by the larvae.