An indigenous marine Achromobacter sp. was isolated from coastal Georgia seawater and modified in the laboratory by introduction of a plasmid with a phoA hybrid gene that directed constitutive overproduction of alkaline phosphatase. The effects of this "indigenous" genetically engineered microorganism (GEM) on phosphorus cycling were determined in seawater microcosms following the addition of a model dissolved organic phosphorus compound, glycerol 3-phosphate, at a concentration of 1 or 10 (mu)M. Within 48 h, a 2- to 10-fold increase in the concentration of inorganic phosphate occurred in microcosms containing the GEM (added at an initial density equivalent to 8% of the total bacterial population) relative to controls containing only natural microbial populations, natural populations with the unmodified Achromobacter sp., or natural populations with the Achromobacter sp. containing the plasmid but not the phoA gene. Secondary effects of the GEM on the phytoplankton community were observed after several days, evident as sustained increases in phytoplankton biomass (up to 14-fold) over that in controls. Even in the absence of added glycerol 3-phosphate, a numerically stable GEM population (averaging 3 to 5% of culturable bacteria) was established within 2 to 3 weeks of introduction into seawater. Moreover, alkaline phosphatase activity in microcosms with the GEM was substantially higher than that in controls for up to 25 days, and microcosms containing the GEM maintained the potential for net phosphate accumulation above control levels for longer than 1 month.