Phosphorus Geochemistry in the Sediment-Water Column of a Hypereutrophic Lake

The role of sediment biogeochemistry on diffusive flux of inorganic P was examined under laboratory conditions using intact sediment cores obtained from a subtropical hypereutrophic lake. The effect of light and dark conditions in the water column on soluble P flux from sediments was measured over a period of 1 yr. Soluble reactive P (SRP) level in the overlying water column was low (<0.01 mg -I) under light conditions, while values under dark conditions teadily increased. This suggests algal uptake under light conditions. Diffusive flux of soluble P from sediments as calculated by measuring the increase in concentration f the overlying water in the dark was 2.71 mg P m-2 d-1. Flux of this magnitude could increase lake water SRP levels by approximately 0.5 mg P L-I yr1. Soluble reactive P concentrations were roughly equivalent o dissolved P concentrations, both for the water column and for the sediment purewater. Purewater SRP increased with depth to a maximum of 6.0 mg P L -~ at 40 cm. Phosphorus flux as calculated from prewater SRP gradients averaged 1.69 mg P m-2 d-a. Sediment P fractionation indicated that water soluble P, KCI extractable P, and Fe and Al-bound P decreased with depth, whereas the amount of Ca-bound P increased. Calcium-bound P was the dominant fraction, comprising over 50% of the total P content. Ion activity products (IAP) calculated using GEOCHEM indicated that the sediments were supersaturated with respect o apatite, beta tricalcium phosphate and whitlockite, with the latter expected to be the phase controlling PO~ ~ activities. L SEDIMENTS can act as either a source or a sink for P. Under normal conditions, sedimentation of P via particulates exceeds the amount released by sediments, causing a net accumulation over time. However, under certain conditions, release of inorganic P from sediments may be large enough to cause, or at least perpetuate, the eutrophication process (Lennox, 1984). For example, Bengtsson (1975) demonstrated that although 98% of the external P load had been diverted from Lake S6dra Bergundasj/)n in Sweden, it remained eutrophic due to the high rate of P regeneration from its sediments. Bostrom et al. (1982) suggested that the most important processes involving P release from lake sediments include desorption, dissolution, ligand exchange and enzymatic hydrolysis. Once the P is in dissolved inorganic form it can be transported from the sediment to the water via diffusion, wind-induced sediment resuspension, bioturbation and gas ebullition. A number of approaches have been used to quantify P availability in lake sediments. Many studies have been conducted on P release from intact sediment columns to the overlying water (Andersen, 1975; Banoub, 1975; Wildung et al., 1977; Theis and McCabe, 1978; Holdren and Armstrong, 1980; Fowler et al., 1987). These studies have usually shown that Eh, pH, ternSoil Science Dep., 106 Newell Hall, Univ. of Florida, Inst. of Food and Agricultural Science, Gainesville, FL 32611-0313. Senior author is currently with the Southeast Research and Extension Center, Univ. of Arkansas, P.O. Box 3508, Monticello, AR 71655. Florida Agricultural Experiment Station Journal Series No. R-01605. *Corresponding author. Received 20 Aug. 1990. Published in J. Environ. Qual. 20:869-875 (1991). perature, microbial zctivity and porewater soluble P levels strongly affect P release or retention by sediments. Early work by Mortimer (1941, 1942) demonstrated that sediment P release was greater when the overlying water was anoxic, due to the reduction of relatively insoluble f~rric phosphates. Later work by Holdren and Armstrong (1980) showed that this was true only for noncalcareous lakes. Several researchers have shown that the rate of P release from intact cores increases with pH (Andersen, 1975; Drake and Heaney, 1987). Andersen (1975) showed that P release from sediments increased as pH increased until about pH 9.5, followed by decrease in P release at pH levels >9.5. This was attributed to two processes: (i) P desorption from Fe oxides and hydroxides due to exchange with OH-, and (ii) precipitation of P as hydroxyapatite. Phosphorus release from sediments was also shown to be highly correlated to the concentration gradient of P in the interstitial waters (Kamp-Nielsen, 1974) and to temperature (Holdren and Armstrong, 1980). Sequential chemical extraction procedures have been widely used to characterize sediment P. Chang and Jackson (1957) tirst developed a P-fractionation scheme for soils. Williams et al. (1971) modified the procedure so that it ~liscriminated Feand Al-bound P from that associated with Ca; however, their fractionation scheme included the use of CDB, which also extracts about one-third of the P from C phosphate minerals such as apatite (Hieltjes and Lijklema, 1980). Modifications of this procedure by Hieltjes and Lijklema (1980) and by Van Eck (1982) suggested using NH4C1 for loosely bound P, NaOH for Feand A1bound P, and HC1 tbr Ca-bound P. Bostrom et al. (1982) reported that fi~w papers are available that combine studies on fractional distribution of P with studies on associated P release from sediments. Solubility measur,~ments provide an alternative method for characterizing the stability of P minerals in lake sediments. In Fe-dominated systems that are oxidized, ferrosoferric hydroxyphosphates have been reported to be the most stable minerals (Nriagu and Dell, 1974), whereas vivianite is the dominant phase under reduced conditions (Nriagu and Dell, 1974; Emerson and Widmer, 1978). Golterman (1982) dicated that apatite equilibrium may also influence P concentrations in calcareous lakes. However, apatite formation may be hindered by the presence of Mg (Martens and Harris,;, 1970), carbonate (Stumm Leckie, 1970) and organic acids (Inskeep and Silvertooth, 1988), or by the absence of calcite surfaces or seed crystals (Griffin and Jurinak, 1973, 1974). Lake Apopka, located in central Florida, is a shallow lake (mean water depth of 2 m) that is currently hypereutrophic as a result of nutrient loadings from external (agricultural d:rainage from adjacent vegetable farms and runoff from nearby citrus groves) and internal (sediment) sources (U.S. Environmental Protection Agency, 1979a). Heavy N and P loadings from