, 2008). In this scenario, the subsequent enhancement in aquatic viral numbers is not caused by lytic success from the inoculation of allochtonous viruses, but rather from the massive activation of prophages from local populations. We thus need more

data to disentangle the complex host specificity paradigm of phage–prokaryotes interactions in aquatic habitats, especially by preventing prokaryotes from being subjected to perceptible changes in environmental conditions. In this study, cross-inoculation assays were conducted between phages and prokaryotes from three aquatic sites of contrasting salinity (freshwater, seawater and hypersaline water). Before incubation, viral concentrates (VC) were resuspended and reconcentrated into ultrafiltered water of the targeted prokaryotes Ipatasertib to avoid potential bias from induction of lysogenic MK-8669 cost phages. Water samples were collected in Senegal (West Africa), on March 5 and 6, 2007 in three ecosystems with contrasting salinity, including (1) a freshwater station (F): Dakar Bango Reservoir, which is the main drinking water supply of St. Louis city, (2) a near-shore seawater station (S) of the Atlantic Ocean located c. 100 m from the Senegalese coast, near the city of St. Louis and (3) a hypersaline (salinity, 310‰) water station (H) located at the center of Lake Retba [more details in Bettarel et al. (2006)] (Table 1). Triplicate

20 L volumes of subsurface water (<0.5 m) were collected at each sampling station and transferred into polycarbonate Nalgene bottles before immediate transfer to the laboratory for processing as follows: Fifteen liters of water from the freshwater and marine site and 4 L from the Retba site were sequentially filtered

onto 3- and 0.2-μm pore-size polycarbonate membranes (Isopore, Millipore, Molsheim, France) to remove larger particles and organisms. The viral filtrates (<0.2 μm) were then ultrafiltered using a Pellicon system (30 kDa) to obtain a solution of concentrated viruses in a final volume of c. 300 mL. This volume was then divided into three replicate VC of 100 mL. To avoid potential bias from nutrient or salinity shifts during the cross inoculations, all the different VCs generated at each site were resuspended in 4 L of ultrafiltered Ribose-5-phosphate isomerase water (<30 kDa) and reconcentrated by ultrafiltration to a final volume of 100 mL. Nine triplicate ‘neoconcentrates’ were thus generated for the cross-inoculation assays, with respect to the different transplantation possibilities (see Fig. 1). The 100 mL neoconcentrates were added to an equivalent volume of 3 μm filtered water from the three different sites in 250-mL polyethylene UV-permeable sterile Whirl-Pack® bags, and incubated for 24 h, at ambient temperature (26 °C) in a large bath (74 × 32 × 18 cm) filled with water corresponding to the incubation type.