Although it is commonly accepted that dissolved organic carbon (DOC) of algal origin limits bacterial growth in pelagic systems, there are relatively few empirical studies documenting this effect. Depending on site and season, both organic and inorganic nutrients can limit the growth of freshwater bacteria. By contrast, inorganic nutrients have only recently been implicated as potentially growth-limiting for marine bacteria. At stations in the Gulf Stream, Sargasso, and Caribbean seas, we used a factorial experimental design to examine effects of inorganic (NH4 and PO4) and organic (glucose) nutrients on the uptake of thymidine and leucine and changes in bacterial abundance. Bacterial growth in seawater dilution cultures varied with station and treatment. Growth rates for the unamended controls (mean +/- SD = O.11 +/- 0.02 d(-1)) were not significantly different among the stations or from the in situ rates. In the Caribbean Sea, additions of NH4 and PO4 resulted in a modest increase in growth rate (mu = 0.20-0.35 d(-1)), whereas glucose, either alone or in combination with PO4 and NH4, resulted in the largest increase (mu = 0.50 - 0.55 d(-1)). By contrast, in the Gulf Stream and Sargasso Sea the addition of glucose, either alone or in combination with NH4, resulted in the smallest increase in growth rate (mu = 0.2-0.4 d(-1)), whereas the addition of PO4, either alone or in combination with glucose and NH4, resulted in the largest increase (mu = 0.55-0.60 d(-1)). Growth rates in the PO4-amended seawater culture were 5-6-fold greater than the controls, suggesting that ambient concentrations of labile DOC were sufficient to sustain vigorous growth. We propose that PO4 limitation of bacterial growth may directly influence the accumulation of DOC in the surface layer and thus have a significant impact on carbon cycling in the sea. If bacterial growth were not constrained by inorganic nutrients, more DOC could be assimilated into bacterial biomass and subsequently transferred to protistan and metazoan grazers. At the third trophic step beyond bacteria, >90% of the DOC initially assimilated would be released as CO2 and <5% would be transferred to incorporated into bacterial biomass. Thus, when bacterial growth is P limited, the quantity of biogenic carbon (POC + DOC) exported to depth may be greater than would occur if DOC were first incorporated into bacterial biomass.
