Use of elemental composition to predict bioavailability of dissolved organic matter in a Georgia river

The relationship between the bioavailability of dissolved organic matter (DOM) and its bulk chemical composition was examined on three dates at 10 sites on the Ogeechee River, a blackwater river in Georgia, Samples of riverine DOM were concentrated from filtered river water using reverse osmosis. In addition, particulate organic matter (POM), in the form of leaf litter, algae, and macrophytes, was leached with synthetic rainwater to obtain fresh DOM. Elemental composition, carboxylic acid content, and bacterial growth were measured on all samples. The results of this study indicate that fresh DOM in POM leachates is generally more bioavailable than riverine DOM. The bioavailability of riverine DOM appears to be greater under low discharge conditions and decreases with distance downstream. The bioavailability of all DOM samples is very well predicted (r2 = 0.93, II = 20) by an empirical equation of the form: bioavailability = a,, + a,(H: C) + a,(0 : C) + al(N : C). When compositional data arc plotted on a van Krevelen diagram, it is evident that POM leachates and, to a lesser degree, DOM from headwatcr sites have compositions that differ little from simple mixtures of major components of biomass (lipids, sugars, proteins, and lignins). Father downstream, major diagenetic alteration of organic matter is evident from the compositions of DOM samples, whose H : C and 0 : C ratios are lower and higher, respectively, than for any possible mixture of biomass components. Bioavailability of DOM is closely related to the pcrccntage of aliphatic carbon in a sample, and downstream decreases in bioavailability arc mainly attributable to selective degradation of aliphatic carbon in riverine DOM. Dissolved organic matter (DOM) is an essential component of microbial food webs in natural waters because it is a growth substrate for bacteria that are then consumed by higher trophic levels (Pomeroy 1974). Only a fraction of the DOM present in natural waters supports bacterial growth, and the bioavailability of DOM varies from <1 to >75% for different molecular weight fractions and in different environments (Table I). Because of the significance of DOM in microbial food webs and the great variation in its availability to bacteria, researchers have sought simple physical or chemical indicators of the bioavailability of DOM in natural waters. One approach has been to quantify specific biochemical compounds such as sugars and amino acids, whose concentrations may reflect the bioavailability of DOM. This approach is neither simple nor predictive of bioavailability of the entire DOM pool. Separating riverine DOM by molecular size using ultrafiltration showed that bacteria use DOM fractions in the order small > large > moderate (Meyer et al. 1987); yet high molecular weight oceanic DOM is more available than low (Amon and Benner 1994). Hence one cannot use the size distribution of DOM to predict reliably its bioavailability. A third approach is based on a bioassay of bacterial growth after resins are used to separate Acknowledgments The authors acknowledge the support of the National Science Foundation (award Nos. BSR-87 178 14 and BSR-8705744). We arc also grateful for the excellent technical support provided by Laura Lcff (UGA) and Susan Dougherty (GIT) on this project. The manuscript was improved based on comments from two anonymous reviewers. humic and nonhumic fractions that differ significantly in their ability to support bacterial growth (Moran and Hodson 1990). DOM in aquatic systems consists mainly of an extremely complex mixture of organic acids, the majority of which are operationally classified as humic substances (Rcuter and Perdue 1977). L,acking the means to isolate significant quantities of pure substances from DOM, organic geochemists have traditionally used bulk characteristics such as elemental composition (% wt/wt), acidic functional group content [peq (mg C) I], LJV-visible absorptivities [absorbance (mg C) ‘1, and average molecular weights to describe DOM (Hedges 1990). In this paper, we evaluate the geochemical approach of using bulk indicators of the nature of DOM (e.g. elemental composition, acidic functional group content, proportions of aliphatic and aromatic carbon) to predict its availability to bacteria.