Viral activity in two contrasting lake ecosystems

Abstract

For aquatic systems, especially freshwaters, there is little data on the long-term (i.e., >6-month period) and depth-related variability of viruses. In this study, we examined virus-induced mortality of heterotrophic bacteria over a 10-month period and throughout the water column in two lakes of the French Massif Central, the oligomesotrophic Lake Pavin and the eutrophic Lake Aydat. Concurrently, we estimated nonviral mortality through heterotrophic nanoflagellate and ciliate bacterivory. Overall, viral infection parameters were much less variable than bacterial production. We found that the frequency of visibly infected cells (FVIC), estimated using transmission electron microscopy, peaked in both lakes at the end of spring (May to June) and in early autumn (September to October). FVIC values were significantly higher in Lake Pavin (mean [M] = 1.6%) than in Lake Aydat (M = 1.1%), whereas the opposite trend was observed for burst sizes, which averaged 25.7 and 30.2 virus particles bacterium(-1), respectively. We detected no significant depth-related differences in FVIC or burst size. We found that in both lakes the removal of bacterial production by flagellate grazing (M(Pavin) = 37.7%, M(Aydat) = 18.5%) was nearly always more than the production removed by viral lysis (M(Pavin) = 16.2%, M(Aydat) = 19%) or ciliate grazing (M(Pavin) = 2.7%, M(Aydat) = 8.8%). However, at specific times and locations, viral lysis prevailed over protistan grazing, for example, in the anoxic hypolimnion of Lake Aydat. In addition, viral mortality represented a relatively constant mortality source in a bacterial community showing large variations in growth rate and subject to large variations in loss rates from grazers. Finally, although viruses did not represent the main agent of bacterial mortality, our data seem to show that their relative importance was higher in the less productive system.

FIG. 1.

Spatiotemporal variations of temperature and oxygen concentration in the water column of Lake Pavin (A and B) and Lake Aydat (C and D).

FIG. 2.

Temporal and vertical variations of bacterial production, FVIC, and burst size (virus particles per bacterium) in the two lakes under study. Errors bars represent standard deviations (n = 3), and letters represent the months.

FIG. 3.

Temporal and vertical variations of bacterivory from HNFs and ciliates in the two lakes under study. Errors bars represent standard deviations (n = 3), and letters represent the months.

FIG. 4.

Temporal and vertical impacts of viral lysis and of flagellate and ciliate grazing on bacterial production (expressed in percentage of BP) in the two lakes under study. E, epilimnion; M, metalimnion; H, hypolimnion. Letters represent the months.

FIG. 5.

Frequencies of visibly infected bacteria in freshwaters, along a trophic gradient. Open triangles correspond to short-term studies (<5 days), and filled triangles correspond to long-term studies (>3 months).