GRAZING OF NONINDIGENOUS BACTERIA BY NANO-SIZED PROTOZOA IN A NATURALCOASTAL SYSTEM

Mesocosms (similar to 4.5 m(3)) situated in a closed bay area were use d to investigate the effect of protozoan predation on nonindigenous ba cteria. Pseudomonas fluorescens strain Ag1 was released into mesocosms as a single inoculum of 1 x 10(5) cells ml(-1) (final concentration) or as four inocula (same concentration each) at intervals of 3 days. M esocosms that had received growth media corresponding to the inoculum served as controls. Numbers of P. fluorescens Ag1 decreased rapidly wh ether released as single or multiple inocula. Direct estimation of pro tozoan predation using fluorescently labeled P. fluorescens from log p hase and starved cultures, respectively, revealed that natural populat ions of heterotrophic nanoflagellates consumed substantial amounts of the nonindigenous bacterial strain. The volume of fluorescently labele d cells prepared from starved cells was 68% of log phase cell volume, but the individual clearance of the small cells was five to seven time s higher than that of the log phase bacteria. The natural populations of nanoflagellates consumed 34-62% of P. fluorescens Ag1 daily if star ved bacteria were offered as food, and 3-13% if the cells were in the logarithmic growth phase. This suggests that the effect of protozoan p redation on nonindigenous bacterial. strains is substantial because cu ltured bacteria are likely to starve in natural environments. The addi tion of P. fluorescens Ag1 and the growth medium enhanced the abundanc e of natural bacteria, chlorophyll a, heterotrophic nanoflagellates, a nd ciliates, but it did not improve the growth conditions for the rele ased strain. The effects on the indigenous populations were more prono unced after addition of fresh medium than following inoculation with c ells, which possibly was due to the lower nutrient content of spent me dium. However, these results, based on direct estimation of protozoan predation on log phase and starved nonindigenous bacteria, point to th e conclusion that mortality induced by bacterivorous predators is the key factor determining removal of nonindigenous bacteria introduced in natural aquatic systems.