Limnol. Oceanogr., 44(3), 1999, 730–736

It is generally accepted that marine bacteria utilize labile, recently produced components of bulk dissolved organic matter. This interpretation is based largely on indirect measurements using model compounds and plankton-derived organic matter. Here, we present an assessment of the relative proportions of modern and older dissolved organic carbon (DOC) utilized by marine bacteria. Bacterial nucleic acids were collected from both estuarine (Santa Rosa Sound, FL) and open-ocean (eastern North Pacific) sites, and the natural radiocarbon signatures of the nucleic acid carbon in both systems were determined. Bacterial nucleic acids from Santa Rosa Sound were significantly enriched in radiocarbon with respect to the bulk DOC and were similar to the radiocarbon signature of atmospheric CO2 at the time of sampling, indicating that these bacteria exclusively assimilate a modern component of the estuarine bulk DOC. In contrast, bacterial nucleic acids from the oceanic site were enriched in 14C relative to the bulk DOC but depleted in 14C with respect to modern surface dissolved inorganic carbon (DIC) and suspended particulate organic carbon (POCsusp). This suggests that open-ocean bacteria assimilate both modern and older components of DOC. The distinct radiocarbon signatures of the nucleic acids at these two sites (i.e., 1120 6 17‰ estuarine vs. 234 6 24‰ oceanic) demonstrate that natural 14C abundance measurements of bacterial biomarkers are a powerful tool for investigations of carbon cycling through microbial communities in different aquatic systems. Marine DOC represents one of the largest exchangeable reservoirs of organic carbon (;0.6 3 1018 g C) at the earth’s surface (Druffel et al. 1992; Hedges 1992). Heterotrophic bacteria are the primary consumers of marine DOC and influence its persistence in seawater through preferential utilization of specific components of the bulk DOC pool. However, despite extensive research examining bacterial DOC utilization in different marine systems (e.g., Rakestraw 1947; Barber 1968; Coffin et al. 1990; Kirchman et al. 1991; Peterson et al. 1994; Carlson and Ducklow 1996; Cherrier et al. 1996), there remains uncertainty in our current understanding as to the relative ages of the specific bulk DOC components supporting bacterial growth. On the one hand, short-term incubation studies (i.e., days to months) indicate that bacteria consume labile, recently produced DOC (Coffin et al. 1993; Carlson and Ducklow 1995, 1996; Cherrier et al. 1996; Coffin and Connolly 1997). On the other hand, as the concentration of oceanic bulk DOC is assumed to be at steady state (Norrman et al. 1995; Carlson and Ducklow 1996) and has an average conventional age of ;4,000–6,000 yr B.P. (Williams and Druffel 1987; Bauer et al. 1992; Druffel et al. 1992), it is generally thought that bacterial remineralization must also be one of the ultimate sinks for older, presumably more refractory bulk DOC constituents (Williams and Carlucci 1976). The latter, however, has yet to be directly demonstrated. The objective of this study was to examine and contrast the relative proportions of modern and older bulk DOC constituents assimilated by estuarine and oceanic bacteria using natural radiocarbon abundances of a specific biomarker, nucleic acids. The carbon isotopic signatures of heterotrophic organisms, including bacteria, reflect the isotopic composition of the organic carbon sources they assimilate (Peterson and Fry 1987; Coffin et al. 1990). It is thus possible to determine what substrates are being consumed in situ by indigenous bacterial populations by measuring their isotopic compositions. Differential filtration cannot effectively isolate natural bacterial assemblages from inorganic and detrital particles of the same size (0.2–1.0-mm effective diameter) for whole-cell isotopic analysis (Coffin et al. 1989). An approach to separate bacteria from the detrital background is to isolate and analyze a biomarker that is specific to bacteria. Bacterial nucleic acids have been shown to have an isotopic composition similar to that of whole cells (Blair et al. 1985; Coffin et al. 1990; Kelly et al. 1998). The specificity of nucleic acids as a bacterial biomarker was confirmed by Coffin et al. (1990), who found that 95% of the 16S rRNA extracted from the 0.2to 0.8-mm size fraction of estuarine water was bacterial. Comparison of the natural 14C signatures of bacterial nucleic acids extracted from free-living bacteria with the 14C signatures of bulk DOC constituents would additionally allow us to evaluate which age fractions of bulk DOC