Changes in phosphorus biogeochemistry along an estuarine salinity gradient: The iron conveyer belt

We used sequential extractions to quantify different forms of particulate phosphorus (PP) in sediments along the salinity gradient of the Patuxent River estuary. About 50–90% of the PP was phosphate bound to iron oxides (Fe-P), and 8–30% was organic P (org-P). Loosely sorbed phosphate (sorb-P), detrital apatite, and authigenic plus biogenic apatite each made up ,10% of the PP. Suspended sediments from the watershed and deposited sediments in tidal freshwater had the highest concentrations of Fe-P, ranging about 30–55 mmol g21 sediment. As pore-water salinity increased to 7 along the estuarine gradient, Fe-P declined to 15–25 mmol g21, org-P increased from 4 to 10 mmol g21, sorb-P increased from 0.5 to 2.5 mmol g21, and total sediment PP declined from 60 to 40 mmol g21. Concentrations of pore-water solutes also changed with salinity. As salinity increased, dissolved Fe and ammonium decreased, while dissolved phosphate increased. Near the freshwater end of the gradient, the molar ratio of pore-water ammonium : phosphate was generally .16 (the Redfield ratio) and ranged up to .700, while at the saline end of the gradient the ratio was generally ,16 and ranged down to ,1.5. Our observations are consistent with the hypothesis that phosphate is released from terrigenous sediments when they are deposited in saline portions of the estuary where sulfide may enhance dissolution of Fe-P and form Fe sulfide precipitates. Such phosphate release may contribute to the generally observed switch from phosphorus limitation in freshwater to nitrogen limitation in coastal marine water. Driven by large-scale anthropogenic alterations of N and P cycles, eutrophication has become a major environmental problem in freshwater, estuarine, and coastal water throughout the world (e.g., Cloern 2001). Primary production is usually limited by P in freshwater and N in seawater (e.g., Howarth and Marino 2006), although there are exceptions to this paradigm. In some parts of the ocean, P or Fe may be limiting (e.g., Wu et al. 2000), and P supply may set the long-term limit on oceanic production (Tyrrell 1999). However, N supply usually sets the short-term limit on production in coastal seawater. In estuaries, where freshwater and seawater mix, spatial and temporal changes in the relative availabilities of N and P cause shifts in nutrient limitation (e.g., Fisher et al. 1999), presenting difficulties for prioritizing nutrient management. More than 90% of the P carried by rivers to estuaries and coastal waters is associated with suspended solids (Follmi 1996). Particulate P (PP) may be bound to Fe, Mn, Al, Ca, or organic C. Often, operationally defined PP fractions are analyzed by sequential extractions and digestions (e.g., Ruttenberg 1992). Such analyses have rarely been applied to suspended particles in streams and rivers (Berner and Rao 1994; Pacini and Gächter 1999; Subramanian 2000). Moreover, riverine fluxes of total PP are poorly quantified because a large proportion of the flux of particulate matter generally occurs during brief, unpredictable episodes of high water flow that often go unobserved (Correll et al. 1999). Thus, the amount of PP entering estuaries, its chemical composition, and its eventual fate are not well known. Much of the fluvial PP may be converted to biologically available dissolved inorganic P (DIP) in estuaries through processes that are enhanced by increasing salinity. Interactions of the Fe, S, and P cycles may have important effects on PO 3{ 4 availability along salinity gradients. More than half the fluvial PP may be bound to iron oxyhydroxides (FeOOH; Compton et al. 2000). This P can be released into solution when FeOOH is reduced in anoxic layers of sediments. When this happens under aerobic freshwater, the resulting Fe(II) can diffuse upward into aerobic 1 Present address: Morgan State University Estuarine Research Center, St. Leonard, Maryland 20685. Acknowledgments Funding was provided by National Science Foundation grant DEB-0235884. Technical assistance was provided by Nancy Goff, Joseph Miklas, Nicole Mitchell, Wendy Milonovich, Marc Sigrist, Erica Kiss, Mike Owens, and Eva Bailey. Help also came from graduate students Sarah Greene, Jennifer O’Keefe, Rebecca Holyoke, and Jeanne Hartzell and Smithsonian undergraduate interns Tori Ziemann, Jessica Soule, Sheena Young, and Jenna Jandorf. Intern support was provided by National Science Foundation REU grants DBI-0097216 and DBI-0353759 and by a NOAA Sea Grant Minority Internship. Dr. Federica Tamburini provided helpful advice on SEDEX sequential extraction methods. This paper was improved by the suggestions of two anonymous reviewers. Limnol. Oceanogr., 53(1), 2008, 172–184 E 2008, by the American Society of Limnology and Oceanography, Inc.