Abstract
Using quantitative PCR, the abundances of six phytoplankton viruses DNA polymerase (polB) gene fragments were estimated in water samples collected from Lake Ontario, Canada over 26 months. Four of the polB fragments were most related to marine prasinoviruses, while the other two were most closely related to cultivated chloroviruses. Two Prasinovirus-related genes reached peak abundances of >1000 copies ml(-1) and were considered 'high abundance', whereas the other two Prasinovirus-related genes peaked at abundances <1000 copies ml(-1) and were considered 'low abundance'. Of the genes related to chloroviruses, one peaked at ca 1600 copies ml(-1), whereas the other reached only ca 300 copies ml(-1). Despite these differences in peak abundance, the abundances of all genes monitored were lowest during the late fall, winter and early spring; during these months the high abundance genes persisted at 100-1000 copies ml(-1) while the low abundance Prasinovirus- and Chlorovirus-related genes persisted at fewer than ca 100 copies ml(-1). Clone libraries of psbA genes from Lake Ontario revealed numerous Chlorella-like algae and two prasinophytes demonstrating the presence of candidate hosts for all types of viruses monitored. Our results corroborate recent metagenomic analyses that suggest that aquatic virus communities are composed of only a few abundant populations and many low abundance populations. Thus, we speculate that an ecologically important characteristic of phycodnavirus communities is seed-bank populations with members that can become numerically dominant when their host abundances reach appropriate levels.