Grazing rates on bacteria: the significance of methodology and ecological factors

Protists and viruses are recognized as the main predators on bacteria in a variety of environments. Several factors may potentially influence the grazing rates of protists on bacteria: temperature, bacterial abundance and production, and protist abundance. We inspected the relative impbrtance of these variables in determining the specific grazing rates of flagellates (GRF), and total commun~ty grazing rates (GT) by analyzing a data set extracted from the literature. We Included freshwater and marine environments, ranging from oligo-to eutrophic systems. As different methods to measure grazing rates (GRF and GT) were used by different authors, the data were also analyzed for possible differences among methods. As there were significant differences among certain methods, we classified them into 2 groups: (1) M1 methods, those that measured grazing rates (GRFand GT) through the uptake of fluorescent particles; and (2) M2 methods, those determining community grazing rates (GT) by dilution, inhibition, filtration, or the uptake of genetically marked bacteria. GTmeasured with M2 methods were systematically higher than those measured by M1. For studies with M1 methods, both GRF and GTwere positively correlated to temperature, bacterial abundance and production. GT were also correlated to heterotrophic nanoflagellate abundance. GT and GRF showed a different pattern of change with temperature, GRFincreased with temperature within the range 0 to 3 0 C while GT increased only up to 18OC. We did not find any satisfactory multiple regression model that explained the variation in the GT measurements with M2 methods. For M1 methods, flagellate abundance and temperatures below 18 "C were the main factors affecting GT (the model explained 78 % of the vanance). While bacterial and flagellate abundances were the main factors at higher temperatures, the model In this case explained only 19% of the variance. Grazing rates could be affected by other sources of variability not considered here (e.g. prey and predator size, chlorophyll). These effects could become much more conspicuous above certain temperatures (18C in our study).