Bacterial biomass and cell size distributions in lakes: More and larger cells in anoxic waters

We measured bacterial abundance, cell size distribution, and related microbiological and geochemical parameters in 20 stratified lakes from three regions in the U.S. Both largest cell size and greatest bacterial abundance occurred in anoxic waters. Bacterial size distributions from the hypolimnia of lakes that were oxic were identical to those in surface waters. Bacteria from anoxic hypolimnia were 2-10 times larger than those from oxic water and cell size was independent of temperature under either oxic or anoxic conditions. For all lakes and strata, bacterial abundance was strongly related to the concentration of total phosphorus (TP). Cell size was not related to TP or to bacterial abundance, suggesting that abundance and size may be regulated by different processes. Bacterial biomass (calculated from the product of mean cell size and bacterial abundance) was typically 4 times greater in anoxic than in oxic waters of lakes. The abundance, size, and biomass of bacteria are important variables in aquatic ecosystems. The quantity of carbon, nitrogen, and phosphorus in bacteria can be a significant fraction of the total in oceanic (Fuhrman et al. 1989) and lake (Cole and Caraco 1993) water columns. Bacteria can account for a substantial fraction of the cycling of C and other elements. In addition, bacteria may be an important resource for protozoans in pelagic food webs (e.g. Sherr and Sherr 1988). Where comparative studies have been conducted, bacterial abundance has been shown to be related to phytoplankton biomass as measured by chlorophyll (Bird and Kalff 1984; Cole et al. 1988) or total P (TP) (Currie 1990). One remarkable feature of these comparisons is that over a wide spectrum of ecosystem conditions bacterial abundance varies by only one to two orders of magnitude. In contrast, phytoplankton and zooplankton biomass vary by more than three orders of magnitude over similarly large gradients (e.g. Hobbie and Cole 1984; Cole et al. 1988; McCauley and Kalff 1981). Bacterial biomass is not usually measured directly. Rather, bacteria are counted, cell volAcknowledgments This study was supported by the National Scicncc Foundation (BSR 8917962 and BSR 9019873) and is a contribution to the Institute of Ecosystem Studies. We thank the researchers of the NTL-LTER for assistance with sampling and for use of some unpublished chemical data. umes are estimated, and the product of these quantities is used to calculate biomass. Thus, one can consider bacterial biomass (pg C liter I) as containing an “abundance” component and a “size” component. Most comparative work has focused on abundance. Where size has been considered, it has been noted that size varies either inversely or independently of trophic conditions (Van Es and Meyer-Reil 1982; Bird and Kalff 1984). Based on a recent intercomparison of cell size measurements, it would appear that much of the reported variation in size is related to methodology which varies among investigators. This variation seriously compromises quantitative comparison of cell size from published studies (H. Ducklow pers. comm.). A further limitation of earlier comparative studies of bacteria has been that the data were primarily derived from the surface mixed layer (e.g. Bird and Kalff 1984; Cole et al. 1988) despite the strong physical and chemical gradients that occur in stratified water columns and strongly influence microbial activity. In this study we measured bacterial abundance and cell size distributions and computed biomass in the water columns of 20 lakes.