Environmental Microbiology (2004) 6(7), 707–715 doi:10.1111/j.1462-2920.2004.00602.x © 2004 Blackwell Publishing Ltd Blackwell Science, LtdOxford, UKEMIEnvironmental Microbiology1462-2912Society for Applied Microbiology and Blackwell Publishing Ltd, 20046 7707715Original ArticleBacteria and viruses in tropical freshwaterP. Peduzzi and F. Schiemer Received 15 September, 2003; revised 12 January, 2004; accepted 12 January, 2004. *For correspondence. E-mail peter.peduzzi@ univie.ac.at; Tel. (+43) 1 4277 54342; Fax (+43) 1 4277 9542. Bacteria and viruses in the water column of tropical freshwater reservoirs Peter Peduzzi* and Fritz Schiemer University of Vienna, Institute of Ecology and Conservation Biology, Division of Limnology, Althanstr. 14, A-1090 Vienna, Austria. Summary In tropical freshwater reservoirs of Sri Lanka, which are linked in an aquatic network, bacterial abundance and production as well as virus abundance, fre- quency of viral infection and virus production were investigated together with a set of nutrient species (Kjeldahl-N, NO 3 -N, total P, soluble P, PO 4 -P). At two characteristic seasons (wet season, dry season), samples were taken from two types of reservoirs (new upland impoundment and ancient, shallow low- land reservoir), each during 4 days at various depths of the entire water columns. Kjeldahl-N and total P were greatly elevated in the wind-mixed water body of the shallow impoundment during the dry season, whereas the deeper reservoir type exhibited no obvi- ous seasonality. In SYBR green™-stained samples, bacterial abundance showed no seasonal pattern in either reservoir type. Bacterial secondary production, however, was significantly elevated in the entire water column of the shallow impoundment under wind-mixed conditions in the dry season. Highest abundance of virus particles and elevated frequency of bacteria containing mature phages were also observed in the shallow reservoir during the dry sea- son indicating favourable conditions for virus propa- gation. Data from this aquatic network show that most virus parameters, such as abundance or fre- quency of visibly infected cells, were positively linked to bacterial abundance and production, but also to organic nitrogen or some phosphorus species. We calculated that between 13.2% and 46.1% of the bac- terial standing stocks would be subjected to virus- mediated mortality. Estimates of bacteriophage pro- duction revealed that from 10 ¥ 10 9 up to 98 ¥ 10 9 phages were produced per litre and day. Bacteria and viruses in the studied tropical freshwater system appear to be linked to various environmental condi- tions and may affect processes at the ecosystem scale. Introduction A vast number of studies in aquatic systems have treated the interdependencies between environmental parame- ters, e.g. such as nutrients or grazing, and the distribution and activity of phytoplankton, zooplankton and even bac- terioplankton. In recent years, it has become obvious that viruses (mainly bacteriophages) are abundant in all aquatic environments, and several studies have investi- gated the ‘virioplankton’ distribution and ecology in marine and temperate inland waters (for a review, consult, e.g. Wommack and Colwell, 2000). Less is known, however, on the inter-relation between viruses and ambient nutrient conditions or bacterial host parameters such as abun- dance and production. Moreover, there is a complete lack of data regarding virus ecology in tropical freshwater systems. In tropical Sri Lanka, inland waters are prominent ecosystems and are considered as an important compo- nent of the country’s socio-economic demands (Duncan et al., 1993). A widespread limnetic system of rivers, channels and reservoirs stretches across the island. This inland water network includes ancient lowland and new upland reservoirs, which differ fundamentally in their basin and catchment morphology, hydrology, nutri- ent regime and productivity. Climatic factors (monsoonal rains, wind) and management practices for, e.g. fisher- ies, irrigation or hydroelectric power affect the limnology of the two types of reservoirs differently. The generation of hydroelectric power in the upland reservoirs requires a continuous and high flowthrough of water, whereas the ancient, lowland dry zone reservoirs, permitting the cultivation of rice crops, exhibit two sets of distinctive flowthrough conditions. From May to October these low- land impoundments feature low water, high wind and a strong water imbalance, resulting in erosion of bottom sediments and high internal nutrient loading. The other characteristic phase takes place between November and April with prevailing windlessness and high water levels with high flowthrough rates. A detailed description of the conditions and general limnology is given by Dun- can et al. (1993). For microbial processes, such