Metabolically active bacteria in Lake Kinneret

Abstract

Three staining methods were used to identify metabolically active bacteria in Lake Kin- neret, northern Israel: CTC, DAPI staining followed by a propanol wash, and the Molecular Probes Live/Dead stain. Positive results from these methods purport to show, respectively, actively respiring bacteria (CTC+), cells with intact nucleoids (NuCC), and cells with intact membranes (MEM+). Con- comitantly, bacterial metabolic activity was measured as electron transport system (ETS) flux, O2 uptake, activities of peptidase, β-glucosidase and lipase, and rate of leucine incorporation in monthly samples taken for 2.5 yr at a pelagic lake station. Laboratory experiments followed changes during 22 or 40 h in the percentages of 'active' bacteria in GF/C-filtered lake water with or without substrate enrichment or antibiotic inhibitors of cell division, or with bacterivorous protists. In lake samples, each of the staining methods detected different aspects of cellular state or metabolic activity but all 3 indicated low percentages of 'active' bacteria relative to total bacterial abundance. CTC+ ranged from 1.0 to 27.3% (average 5.1%), NuCC from 1.4 to 42.9% (average 8.3%) and MEM+ from 1.0 to 29.9% (average 8.8%), with no clear seasonal or spatial patterns. No significant correlations were found between the proportions of 'active' bacteria in lake water as determined by these methods, although such correlations were observed in the laboratory experiments. Significant correlations were obtained between ETS and O2 uptake, peptidase and β-glucosidase, and between leucine incorporation and peptidase. ETS was significantly correlated with CTC+ and NuCC cell abundance, but not with total bacteria (DAPI counts). In contrast, peptidase activity correlated with total bacter- ial counts. Results of time course experiments indicated that some bacteria which initially appear to be inactive can become active when stimulated by substrate addition, even though cell division is inhibited. Grazing by protists increased the percentage of active bacteria, at least during the active predator-prey phase. Our data support the hypothesis that in natural waters usually only a small frac- tion (probably <20%) of the entire bacterial assemblage is strongly active metabolically at any given time. This proportion may increase dramatically with localized substrate inputs. The concept of bac- terial assemblages, heterogeneous not only in terms of phylotype, but also in terms of levels of meta- bolic activity will need to be considered in future aquatic ecosystem models.