Effect of grazers and viruses on bacterial community structure and production in two contrasting trophic lakes

Abstract

Background: Over the last 30 years, extensive studies have revealed the crucial roles played by microbes in aquatic ecosystems. It has been shown that bacteria, viruses and protozoan grazers are dominant in terms of abundance and biomass. The frequent interactions between these microbiological compartments are responsible for strong trophic links from dissolved organic matter to higher trophic levels, via heterotrophic bacteria, which form the basis for the important biogeochemical roles of microbial food webs in aquatic ecosystems. To gain a better understanding of the interactions between bacteria, viruses and flagellates in lacustrine ecosystems, we investigated the effect of protistan bacterivory on bacterial abundance, production and structure [determined by 16S rRNA PCR-DGGE], and viral abundance and activity of two lakes of contrasting trophic status. Four experiments were conducted in the oligotrophic Lake Annecy and the mesotrophic Lake Bourget over two seasons (early spring vs. summer) using a fractionation approach. In situ dark vs. light incubations were performed to consider the effects of the different treatments in the presence and absence of phototrophic activity.

Results: The presence of grazers (i.e. <5-μm small eukaryotes) affected viral production positively in all experiments, and the stimulation of viral production (compared to the treatment with no eukaryotic predators) was more variable between lakes than between seasons, with the highest value having been recorded in the mesotrophic lake (+30%). Viral lysis and grazing activities acted additively to sustain high bacterial production in all experiments. Nevertheless, the stimulation of bacterial production was more variable between seasons than between lakes, with the highest values obtained in summer (+33.5% and +37.5% in Lakes Bourget and Annecy, respectively). The presence of both predators (nanoflagellates and viruses) did not seem to have a clear influence upon bacterial community structure according to the four experiments.

Conclusions: Our results highlight the importance of a synergistic effect, i.e. the positive influence of grazers on viral activities in sustaining (directly and indirectly) bacterial production and affecting composition, in both oligotrophic and mesotrophic lakes.

Figure 1

Time-course of viral abundance (10⁷ virus ml^-1) and bacterial abundance (10⁶ cell ml^-1) in the four experiments during the incubation period. Values are given as mean ± standard deviation of triplicate incubations. Asterisks indicate sampling time point for which the VFA and VF treatments were not significantly different from the V treatment (ANOVA, P > 0.05, n = 9). Note that the panels have different scales. LA1, LA2, LB1, LB2: abbreviations as in Table 1.

Figure 2

Stimulation of bacterial abundance (A), production (B) and lysis mortality (C), and viral abundance (D) and production (E). Values are given as mean ± standard deviation of triplicats incubations from VF and VFA treatments and average values from the V treatment. ANOVA were preformed between treatment V and the two other treatments (VF and VFA), and the comparison without significance (P > 0.05) was indicated with asterisks.

Figure 3

Time-course of viral production (10⁵ virus ml^-1 h^-1) and bacterial production (μgC l^-1 h^-1) in the four experiments during the incubation period. Values are given as mean ± standard deviation of triplicate incubations. Asterisks indicate sampling time point for which VFA and VF treatments were not significantly different from the V treatment (P > 0.05, n = 9, ANOVA). Note that the panels have different scales. LA1, LA2, LB1, LB2: abbreviations as in Table 1.

Figure 4

Bacterial community structure at the beginning (referred to as '0') and at the end (96 h, referred as 'final') of the incubation, visualized by DGGE of PCR-amplified partial 16S rRNA genes, and the position of the different bands excised and sequenced. (B1 to B8, see Table 5). V₀ and V_Final, treatment Viruses+Bacteria at the beginning and the end of experiments; VF₀ and VF_final, treatment Viruses+Bacteria+Flagellates at the beginning and the end of experiments, VFA_final, treatment Viruses+Bacteria+Flagellates+Autotrophs at the end of experiments

Figure 5

Cluster analysis of DGGE profiles based on band position and intensity. Scale bars indicated the Bray-Curtis similarity index in Lake Annecy (A) and Lake Bourget (B). V₀ and V_Final, treatment Viruses+Bacteria at the beginning and the end of experiments; VF₀ and VF_final, treatment Viruses+Bacteria+Flagellates at the beginning and the end of experiments, VFA_final, treatment Viruses+Bacteria+Flagellates+Autotrophs at the end of experiments.