Significance of bacterivory and viral lysis in bottom waters of Franklin Bay, Canadian Arctic, during winter

Little information is currently available about water column microbial processes or mortality during Arctic winter. To address this paucity, we used epifluorescence microscopy and dilution experiments to determine the abundance of flagellates, bacteria and virus-like particles (VLP) and the rates of bacterial growth, bacterivory and virus-induced mortality in subzero-temperature bottom waters (< 230 m) of Franklin Bay during February and March 2004, when ice-covered surface waters were highly oligotrophic (maximum chlorophyll a value of 0.09 mu g l super(-1)). We focused on bottom waters due to the possible importance of sediment resuspension as a source of organic matter. While flagellates were present at low densities (1.5 to 3.1 x 10 super(2) ml super(-1)), bacterial concentrations resembled those from other seasons in the region and increased over the 5 wk sampling period, from 1.4 x 10 super(5) to 3.0 x 10 super(5) ml super(-1). VLPs were typically an order of magnitude more abundant than bacteria (range of 1.4 to 4.5 x 10 super(6) VLP ml super(-1)) and, like the fraction of particle- associated bacteria (but not total bacteria), correlated with particulate organic carbon concentration (r sub(s) = 0.82, p < 0.04, n = 7). Grazing rates, whether measured in dilution experiments or calculated from flagellate abundance, were low or undetectable (maximum of -0.004 h super(-1)). Of 3 parallel experiments, 2 yielded substantial virus-induced mortality (-0.006 to -0.015 h super(-1)), comparable to or exceeding the intrinsic bacterial growth rate (0.010 h super(-1) in both experiments) and suggesting viruses were the more important agents of bacterial mortality under these conditions. Using a viral reduction approach, VLP production measured in the water column or ice-moored sediment traps was commonly low (0.3 to 7.7 x 10 super(4) VLP ml super(-1) h super(-1)) or undetectable, highly variable among replicates and, when measurable, implied viral turnover times between 0.9 and 12 d. In general, our results show that, despite the oligotrophy of Arctic winter, bottom water bacterial communities can remain active and subject to viral predation.