Dissolved and particulate organic matter source-age characterization in the upper and lower Chesapeake Bay: A combined isotope and biochemical approach

In order to characterize the sources and ages of organic matter contributing to river and estuarine outflow waters, the present study investigated D14C and d13C signatures of the major operationally defined biochemical classes of ultrafiltered dissolved organic matter (UDOM) in conjunction with lipid biomarker and elemental compositions of UDOM and suspended particulate organic matter (POM) in the Chesapeake Bay. Freshwater (Susquehanna River) UDOM was dominated by a molecularly uncharacterized (MUC) fraction, followed by total carbohydrate (TCHO), total hydrolysable amino acid (THAA) and total lipid (TLE) components. In contrast, UDOM at the bay mouth (salinity ,22–24) was comprised mainly of TCHO, followed by MUC, THAA, and TLE. The D14C and d13C signatures of both UDOM and its major biochemical classes indicate that Susquehanna DOM is derived in part from old allochthonous terrestrial sources, whereas young marine sources dominate at the bay mouth. In contrast to the other biochemical classes, lipophilic DOM at both sites was very old (,5,000– 7,000 years B.P.). In addition, factor analysis of lipid biomarker compounds revealed unique signatures for the UDOM and POM pools that imply disparate source and/or recycling properties as well as potential influences due to physical partitioning. Lipid biomarker compounds showed that although autochthonous riverine/estuarine sources dominated both the UDOM and POM pools, terrigenous lipids were elevated in the Susquehanna during high flow conditions. The presence of lipid biomarkers diagnostic of ‘‘fresh’’ algal material in UDOM further suggested its greater reactivity than POM. The observed biochemical and lipid biomarker compositions and isotopic signatures of UDOM and POM are consistent with previous findings suggesting that these two major organic matter pools have dissimilar reactivities and cycling times, and they derive from comparatively unique source-age materials in rivers and estuaries. Global riverine discharge of dissolved (DOM) and particulate organic matter (POM) is on the order of ,0.4 3 1015 g C yr–1 (Meybeck 1982; Hedges 1992), making it a potentially major source of organic carbon (C) to the world’s oceans. However, much of this carbon is transported through, and modified within, riverine and estuarine environments before entering the coastal ocean. Organic matter in these systems is comprised of variable amounts of identifiable biomolecules such as carbohydrates, proteins, and lipids (Mannino and Harvey 2000; Minor et al. 2001), as well as operationally defined (e.g., humic and fulvic substances in the dissolved phase) and other noncharacterizable, long-lived geomacromolecules. Radiocarbon (14C) studies indicate that DOM and POM in rivers and estuaries range from modern to thousands of years in age (Hedges et al. 1986; Masiello and Druffel 2001; Raymond and Bauer 2001). Previous studies have also shown, however, that young subcomponents of estuarine DOM can be reactive and cycle on timescales of days to weeks (Guo and Santschi 1997; McCallister et al. 2004). Consequently, the short cycling time of a portion of this DOM may result in even older, more refractory forms of organic matter remaining undegraded in some rivers and estuaries (Raymond and Bauer 2001) and ultimately being exported to the coastal ocean. Radiocarbon studies of riverine and estuarine systems have found that the average age of POM is often much older than co-occurring DOM (Masiello and Druffel 2001; Raymond and Bauer 2001). Paradoxically, the size-reactivity continuum model predicts that the bioreactivity of organic matter decreases along a continuum of size (from large to small; Amon and Benner 1996). These seemingly incongruous characteristics may be due to the complex processes governing the sources and sinks of bulk organic matter pools within riverine/estuarine systems (Hedges and Keil 1999) as well as to the unique physical and hydrological attributes of individual estuarine systems (e.g., small, mountainous river vs. coastal plain estuaries) and their associated watersheds. Although measurement of natural 14C in bulk organic matter provides an estimate of average 14C age for the total pool, it does not allow for assessment of the turnover times 1 To whom correspondence should be addressed. Present address: College of Arts and Sciences, Florida Gulf Coast University, 10501 FGCU Boulevard South, Fort Myers, Florida 33965 ([email protected]). Acknowledgments We thank A. Volety and L. McCallister for field assistance; S. Griffin and M. Ederington-Hagy for guidance in the laboratory; J. Hwang and J. Southon for discussion of methodology; E. Keesee and E. Waterson for technical support; A. McNichol and colleagues at NOSAMS for D14C measurements; E. Franks for d13C measurements; and K. Arzayus, L. McCallister, R. McDaniel, and V. Pilon for assistance in the laboratory. Funding was provided by the National Science Foundation (OCE-9810669, OCE-0327423, and EAR-0403949) to J.E.B. and (OCE-9521170) to E.A.C.; the U.S. Department of Energy Ocean Margins Program (FG05-94ER61833) to J.E.B.; and SMS/VIMS Student Research and College of William and Mary Minor Research Grants to A.N.L. This paper is contribution 2712 of the Virginia Institute of Marine Science, The College of William and Mary. Limnol. Oceanogr., 51(3), 2006, 1421–1431 E 2006, by the American Society of Limnology and Oceanography, Inc.