Introduction

Many attempts have been made to control soilborne plant pathogens and to improve the growth of plants by inoculation of seed or soil with selected strains of micro-organisms; mainly bacteria. More recently, selected bacteria that are expected to degrade xenobiotics have been suggested for the bioremediation of polluted sites (Short et al 1990). Nowadays, risk assessment studies also need to address the fate of engineered telluric soil and non-telluric soil bacteria (Tiedje et al., 1989; Doyle etal, 1991).

Besides increased crop yields claimed by Russian workers (Schroth and Becker, 1990), most of the tentative applications of micro-organisms in agricultural soils have failed. The widely practised inoculation of legume seeds with Rhizobium spp. is one of the few examples of success of application of micro-organisms to improve crop yield (Stacey and Upchurch, 1984). Application of the strain K84, and more recently of the modified strain K1026, of Agrobacterium radiobacter to control crown gall of plants is one of the few examples of a biological control method commercially applied with some success in several countries (Ryder and Jones, 1990). In most cases, the beneficial effects expected from the microbial inoculation are not consistently reproduced under field conditions. The poor survival of the introduced microorganisms in soil is the main explanation for these failures. In fact, the population density of the introduced micro-organisms decreases to the limit of the carrying capacity of the soil. Moreover, the soil represents very heterogeneous environments in which the introduced micro-organisms must find suitable habitats, some of which are very strain specific (Hattori and Hattori, 1976).