What did we learn from 24 field releases of GMMs in Italy?

The deliberate release of genetically modified microorganisms (GMMs) stems on the advancements, during the last two decades, of recombinant DNA technology and has shown its potential for environmental applications in bioremediation, biocontrol of plant pathogens and plant growth promotion. The modification of bacterial strains especially designed to allow appropriate monitoring and risk assessment started on the eve of ‘90s in our lab, and their evaluation continued at microcosm and green-house level. The field releases in Italy started in 1994 and involved nitrogen-fixing nodule-forming rhizobia, plant growth promoting azospirilli, and plant protecting pseudomonads. All 24 releases (1994-1999) were carried out by our groups according to existing European legislation and have provided useful information on environmental impact characterization and biosafety-related aspects of GMMs. In some instances, post-release monitoring is still being done after eight years at the site of release. Different marker genes/reporter systems were used for Rhizobium leguminosarum, Azospirillum brasilense and Pseudomonas fluorescens. They include lacZY, gusA, MerRlacZ and luc. No introduced antibiotic resistance genes were used, well in advance on the actual Directive 18/2001. The genetic modifications involved agronomically relevant as well as non-relevant traits, and the crops used included peas, soybean, faba bean, sorghum and sugarbeet. Among the relevant outcome of these studies there are the following: (1) the genetic modification such as the introduction of a marker gene, does not affect the ecological fitness nor the impact of the GMM on culturable microbial populations (2) the genetic modification consisting of agronomically relevant traits may lead to the expected agronomically beneficial effects (3) although an initial expected increase of population density of the GMM has been observed, the introduced seed inoculants remain 102-103 cfu per gram of soil or below detection limit few weeks or few months after release (4) in some cases, the beneficial effects of the GMM observed at lab, microcosm and green-house level were not confirmed when the microbe was released in the field (e.g. as a biocontrol agent). Based on the results appeared in the literature and on existing strict regulations, it appears that the environmental impact characterization represents an important step for a case-by-case evaluation of the biological agent to be used at field level. A contribution towards mathematical modeling of risk assessment of field-released biological agents is finally presented and discussed.