Microb Ecol (2003) 45:119–127 DOI: 10.1007/s00248-002-1016-1
2003 Springer-Verlag New York Inc.

Viral Lysis, Flagellate Grazing Potential, and Bacterial Production in Lake Pavin Y. Bettarel,1 C. Amblard,1 T. Sime-Ngando,1 J.-F. Carrias,1 D. Sargos,1 F. Garabe´tian,2 P. Lavandier2 1

Laboratoire de Biologie des Protistes, UMR CNRS 6023, Universite´ Blaise Pascal (Clermont-Ferrand II), F-63177 Aubie`re Cedex, France 2 Centre d’Ecologie des Syte`mes Aquatiques Continentaux, Universite´ Paul Sabatier, F-31062 Toulouse Cedex 4, France Received: 23 October 2001; Accepted: 7 October 2002; Online publication: 28 January 2003

A

B S T R A C T

Abundances of different compartments of the microbial loop (i.e., viruses, heterotrophic bacteria, nonpigmented nanoflagellates, and pigmented nanoflagellates), bacterial heterotrophic production (BHP), viral lysis, and potential flagellate grazing impacts on the bacterial assemblages were estimated during a short-term study (24 h) conducted in June 1998 in the epilimnion (5 m) and metalimnion (10 m) of a moderate-altitude oligomesotrophic lake (Lake Pavin, France). Viral and bacterial abundances were higher in the metalimnion than in the epilimnion, whereas pigmented and nonpigmented nanoflagellates were more numerous in the epilimnion. The control of the BHP due to viral lysis (determined by examination of viral-containing bacteria using a transmission electron microscope) was significantly higher in the meta- (range = 6.0– 33.7%, mean = 15.6%) than in the epilimnion (3.5–10.3%, 6.4%). The same was for the losses of BHP from the potential predation by nanoflagellates which ranged from 0.5 to 115.4% (mean = 38.7%) in the epilimnion, and from 0.7 to 97.5% (mean = 66.7%) in the metalimnion. Finally, estimated viral mediated mortality rates from the percentage of visibly infected cells and potential nanoflagellate grazing rates based on assumed clearance rates suggest that flagellates consumed a larger proportion of bacterial production than was lost to viral lysis.

Introduction For a long time it was assumed that the regulation of bacterial abundance and production in freshwaters and marine environment was mainly related to the availability and/or quality of organic substrates [9]. However, studies Correspondence to: T. Sime-Ngando; E-mail: [email protected]

conducted in recent decades have tended to show that limitation by inorganic nutrients, in lakes [29] and in the marine environment [33], predation by phagotrophic protists [36], and by metazoan zooplankton [24], and viral lysis [17] are also factors that are liable to be significantly involved in this regulation. Other processes, such as sedimentation, autolysis and allelopathic interactions exist but are usually thought to be of minor importance [49]. It is therefore possible to draw up a relatively exhaustive

120

inventory of the factors regulating bacterial communities. However, the relative importance of these factors and their changes over the annual cycle or in the short term are still poorly known. The high quantitative abundance and diversity of viruses in the plankton [26] has prompted planktonologists in their efforts to understand the role of viruses in the process of losses affecting bacterioplankton. Seasonal studies conducted in marine environments [17, 43, 46] and in freshwaters [21, 27] have demonstrated that between 0 and 30% of bacterial production was removed by viral lysis. Based on the observation that the ratio between bacteria and protozoan flagellates was relatively constant (1000/1), Sanders et al. [36] suggested that bacterial abundance was regulated by heterotrophic flagellates in the marine and lacustrine pelagic zone. Indeed, in view of their predation rates (10 to 250 bacteria per protozoan per day [15, 37]) and their abundance (102 to 105 mL)1), heterotrophic flagellates are potentially capable of regulating bacterial abundance. Mixotrophic flagellates [35] and ciliates [38] could also, at times, exert a significant predation on the bacterioplankton. A few attempts have been made to analyze the relative importance of viral lysis and of flagellate grazing as potential regulating factors for natural bacterioplankton, in both marine [6, 18–20, 42, 44] and freshwaters [40, 48]. Most of these studies were thus conducted in marine systems, and in experimental conditions (micro- or mesocosms) that were relatively different from the natural ones. The main objective of this work was therefore to study the relative bacterial mortality rates inferred from estimated viral infected cells and assumed flagellate grazing rates in the oligomesotrophic Lake Pavin, at a time scale (24 h) that is closer to the generation times of microorganisms in this lake [12] and in other pelagic systems as well (cf. [13]). The choice of short time scale eliminated, to a large extent, the temptation to falsely interpret the correlations that frequently result at a seasonal scale from the covariation of two variables with respect to a third factor, usually the temperature [33].

Materials and Methods Sampling The study was conducted in Lake Pavin, situated in the Massif Central of France (256’E, 4529’N) at an altitude of 1197 m. The

Y. Bettarel et al. lake is of recent volcanic origin, occupying an almost circular caldera (area = 44 · 104 m2). It is a deep (maximum depth: 92 m) dimictic (with partial overturn), meromictic (limit between mixo and monimolimnion: 55 m), and oligomesotrophic lake. The lake is apparently sheltered from the wind and, during the study, the water column was well stratified and no wind-driven mixing occurred, since the weather was calm and sunny. For more details on the morphometric characteristics of Lake Pavin, see Amblard et al. [2]. Nine series of samples were collected over 24 h, on 15 and 16 June 1998, at the following local times: 10 h, 14 h, 18 h, 22 h, and 24 h (on 15 June), 2 h, 4 h, 6 h, and 10 h (on 16 June), except for bacterial production which was measured at 10 h, 15 h, 17 h (on 15 June), 2 h, and 6 h (on 16 June). Samples were taken at depths of 5 and 10 m, which at this time of year were situated in the epilimnion and metalimnion, respectively. All the samples were collected at the central point in the lake using a PVC Van Dorn bottle. Water temperature and dissolved oxygen concentrations were measured in situ with a YSI-GRANT/3000 multiparametric probe.

Counts of Viruses, Bacteria, and Nanoflagellates Samples were fixed with formaldehyde (2% final concentration) for viral and bacterial abundance and in ice-cold glutaraldehyde (2% final concentration) for the abundance of nanoflagellates. Virus-like particles (VLP) were counted according to the protocol of Xenopoulos and Bird [53] using epifluorescence microscopy and the fluorochrome Yo-Pro. This protocol, which is a modification of the original protocol proposed by Hennes and Suttle [22], reduces the time needed to stain viruses from 48 h to 4 min of heating, works with samples fixed by aldehydes, and produces reliable counts of free viruses in aquatic ecosystems [5]. A stock solution of Yo-Pro-1 (Molecular Probes Europe, Leiden, The Netherlands) was diluted to 50 mM in an aqueous solution of 2 mM NaCN. VLP in 1 mL of lake samples were gently filtered (