Assessing sources and ages of organic matter supporting river and estuarine bacterial production : A multiple-isotope ( D 14 C , d 13 C , and d 15 N ) approach

We used radiocarbon (D14C) and stable isotopic (d13C, d15N) signatures of bacterial nucleic acids to estimate the sources and ages of organic matter (OM) assimilated by bacteria in the Hudson River and York River estuary. Dualisotope plots of D14C and d13C coupled with a three-source mixing model resolved the major OM sources supporting bacterial biomass production (BBP). However, overlap in the stable isotopic (d13C and d15N) values of potential source end members (i.e., terrestrial, freshwater phytoplankton, and marsh-derived) prohibited unequivocal source assignments for certain samples. In freshwater regions of the York, terrigenous material of relatively recent origin (i.e., decadal in age) accounted for the majority of OM assimilated by bacteria (49–83%). Marsh and freshwater planktonic material made up the other major source of OM, with 5–33% and 6–25% assimilated, respectively. In the mesohaline York, BBP was supported primarily by estuarine phytoplankton–derived OM during spring and summer (53–87%) and by marsh-derived OM during fall (as much as 83%). Isotopic signatures from higher salinity regions of the York suggested that BBP there was fueled predominantly by either estuarine phytoplankton-derived OM (July and November) or by material advected in from the Chesapeake Bay proper (October). In contrast to the York, BBP in the Hudson River estuary was subsidized by a greater portion (up to ;25%) of old (;24,000 yr BP) allochthonous OM, which was presumably derived from soils. These findings collectively suggest that bacterial metabolism and degradation in rivers and estuaries may profoundly alter the mean composition and age of OM during transport within these systems and before its export to the coastal ocean. The fate of organic matter (OM) in aquatic systems is controlled primarily by heterotrophic bacterial respiration and biomass production (Findlay et al. 1992; Williams 2000). Sources and sinks of OM in river and estuarine systems in particular are often difficult to establish quantitatively because of such factors as spatial and temporal variability in the simultaneous inputs and turnover of autochthonous and allochthonous forms and the subsequent homogenization of OM source signatures (Canuel et al. 1 Corresponding author (leigh@vims.edu). Present address: Dèpt des Sciences Biologiques, Universitè du Quèbec à Montrèal, CP 8888, Succ. Centre Ville, Montrèal, Quèbec, Canada H3C 3P8. Acknowledgments We are grateful to the G. G. Hatch Isotope Laboratories, University of Ottawa, for stable isotope analyses and to Brian Frantz and Paula Zermeno at the Lawrence Livermore National Lab for their invaluable help with radiocarbon analyses. We thank Nina Caraco and Jon Cole for logistical support and the use of unpublished data. We extend special thanks to Stuart Findlay for the collection of water samples for humic extraction. This manuscript was greatly improved through insightful comments from two anonymous reviewers. This work was supported by grants from the National Science Foundation (OCE-9810669) and the U.S. Department of Energy Ocean Margins Program (FG05-94ER61833) to J.E.B. and from the Hudson River Foundation and National Science Foundation–Division of Environmental Biology to S.L.M. 1995; Cloern et al. 2002). Although bioassays are frequently used to evaluate the reactivity of bulk pools such as dissolved organic matter (DOM; del Giorgio and Davis 2003), the information they provide about the biochemical composition and age structure of potential sources of OM supporting heterotrophic production is often limited. Globally, rivers transport ;0.25 Pg of dissolved organic carbon (DOC) per year toward the ocean (Hedges et al. 1997). This typically occurs via estuaries or similar mixing zones, where a number of biogeochemical and physical processes may modify the quantities and characteristics of the OM delivered to the ocean. Some studies have reported conservative transport of DOC through estuaries (e.g., Mantoura and Woodward 1983; Ittekot 1989), which suggests insignificant removal by bacteria. However, others have indicated that DOC processing in estuaries is more complex and may include both internal sources and sinks of DOC (Mannino and Harvey 2000; Raymond and Bauer 2000a). Thus, there may be no general pattern governing OM transport through estuaries as a whole. Instead, different river-estuary systems may possess unique physical, hydrological, and biogeochemical features that result in distinct OM dynamics. Biogeochemical processing in estuaries is a primary control on the transfer of terrigenous OM from land to the coastal sea. Allochthonous OM delivered from watersheds to rivers and estuaries has traditionally been classified as refractory. However, net system heterotrophy in coastal eco1688 McCallister et al. Fig. 1. (A) The York River estuary. Inset shows the York’s location relative to the Chesapeake Bay proper. Sampling locations are designated by an arrow and the approximate salinity. (B) The Hudson River and associated watershed. The map shows the tidal Hudson River (heavy line) formed by the confluence of the Upper Hudson River and the Mohawk River and running from river 240 km south to New York City. Stars denote sampling locations. systems (Smith and Hollibaugh 1993; Frankignoulle et al. 1998) requires the de facto utilization of some portion of this allochthonous material by microheterotrophs. At present, the relative importance of autochthonous versus allochthonous OM to heterotrophic pathways of energy flow in most river-estuary systems remains largely unknown. Furthermore, the relative susceptibility of allochthonous OM sources (e.g., eroded agricultural soils vs. forest runoff) to heterotrophic decomposition is not readily predictable using current approaches. Thus, a better understanding of the quantitative and qualitative processing of OM in estuarine systems may be key to reconciling the biogeochemical fate of terrigenous OM as it is transported to coastal seas. Stable isotopes (d13C, d15N, d34S, etc.) have been used previously to infer OM inputs and cycling in freshwater and marine systems (e.g., Lajtha and Michener 1994), although the relative contributions of multiple sources to bulk OM pools and trophic levels can be difficult to ascertain because of overlap in the isotopic signatures of different components (Cloern et al. 2002). The simultaneous use of multiple isotopic tracers may, however, help overcome some of these limitations (Peterson et al. 1985; Bauer et al. 2002). Both d13C and d15N have been applied, with different degrees of success, for the identification of the sources of OM assimilated by bacteria (Coffin et al. 1989, 1990; Coffin and Cifuentes 1999). Natural abundance 14C measurements also have the potential to provide additional resolution in discerning the relative importance of allochthonous and autochthonous OM sources to bacterial production (Cherrier et al. 1999). The greater sensitivity and potential dynamic range of D14C (approximately 21,000‰ to 1435‰) compared with d13C sources (approximately 235‰ to 212‰) or d15N sources (approximately 22‰ to 140‰) may permit even greater resolution of multiple OM sources in rivers and estuaries. In addition, autochthonous and allochthonous forms of OM may be better differentiated and more accurately quantified by using simultaneous D14C and stable isotope signatures (Raymond and Bauer 2001a,b; Bauer et al. 2002). The objectives of the present study were to evaluate the sources and ages of DOM supporting bacterial production in two distinct temperate systems, the Hudson River and York River estuary, using a novel natural radiocarbon (D14C) and stable isotopic (d13C and d15N) approach. Previous findings have suggested that a substantial portion of heterotrophic bacterial biomass production (BBP) in both the York and Hudson rivers must be supported by allochthonous (i.e., terrigenous) sources of OM (Findlay et al. 1991; Howarth et al. 1996; Raymond et al. 2000; Schultz et al. 2003). The large difference in the mean DOM ages of these two geochemically distinct river-estuary systems (modern age in the York, 102–103 yr BP in the Hudson; Raymond and Bauer 2001b,c) therefore provides a unique opportunity to evaluate these isotopes for tracing the natural sources and ages of OM that fuel bacterial metabolism in both. Materials and methods Study sites and sampling locations—The York River estuary is a moderately stratified subestuary of the Chesapeake Bay that is encompassed by a watershed size of ;4,350 km2 and has an average (50-yr) annual mean flow rate (Pamunkey River) of 28.5 m3 s21 (Fig. 1A). The York is formed by the convergence of the Pamunkey and Mattaponi Rivers, which account for 80% and 20% of the freshwater inputs, respec1689 Sources and ages of organic matter Table 1. Water characteristics of the York River estuary and Hudson River. Site and date Stream flow* (m3 s21) Salinity Water volume (L) (method of concentration) Water temperature (8C) Chl a (mg L21)