AME 41:209-219 (2005)  -  doi:10.3354/ame041209

ABSTRACT: Assessment of the role of viral lysis in aquatic microbial communities requires a robust means of estimating viral production (VP). Here, 3 incubation-based VP methods (fluorescently labeled viruses [FLV], dilution [DIL], and thymidine incorporation [TdR]) were evaluated in water samples from the Delaware and Chesapeake Bays. In Chesapeake Bay samples, average VP rates were 10-fold higher for FLV and DIL (~3 to 25 × 10¹⁰ viruses l^–1 d^–1) than for TdR (~0.2 to 3 × 10¹⁰ viruses l^–1 d^–1). Estimates of viral-mediated bacterial mortality (VMM) indicate that FLV overestimates VP in eutrophic waters, since >100% of bacterial production (BP) would have been consumed through viral lysis. DIL and TdR VP methods gave more realistic estimates of VMM with respect to BP. The FLV method provides estimates of both VP and virus removal rates; however, it requires preparation of a viral tracer and twice as many microscopic enumerations as the DIL method. DIL was the least difficult and most efficient method; however, bacterial loss during vacuum diafiltration resulted in poor replicability. TdR gave similar VP estimates to DIL, but requires use of a large and poorly constrained conversion factor. With methodological improvements in bacterial cell recovery, DIL should be the most widely applicable method for estimation of VP in highly productive estuarine waters.

KEY WORDS: Viral production · Bacteriophage · Marine viruses · Marine microbial ecology