Development And Application of Functional Assays For Freshwater Dissolved Organic Matter

Dissolved organic matter (DOM) in natural waters participates in many important ecological and geochemical reactions, including acid-base buffering, light absorption, proton binding, binding of heavy metals, organic contaminants, aluminium and radionuclides, adsorption at surfaces, aggregation and photochemical reactivity. We are studying DOM in order to understand and quantify these functional properties, so we can use the knowledge to predict the influence of DOM on the natural freshwater environment. As DOM has no readily identifiable structure, our approach is to measure what it does, rather than what it is. Thus, we have developed a series of 12 standardised, reproducible assays of physico-chemical functions of dissolved organic matter (DOM) in freshwaters. The assays provide quantitative information on light absorption, fluorescence, photochemical fading, pH buffering, copper binding, benzo(a)pyrene binding, hydrophilicity and adsorption to alumina. We have collected twenty DOM samples in total, ten samples from a eutrophic lake (Esthwaite Water) and ten samples from three stream waters. A mild isolation method was then used to concentrate the DOM samples for the assay work. When assaying the concentrates, parallel assays were also preformed with Suwannee River Fulvic Acid (SRFA), as a quality control standard. Our results showed that; (i) for eleven of the assays, the variability among the twenty DOM samples was significantly (p<0.001) greater than can be explained by analytical error, i.e. by comparison with results from the SRFA quality control; (ii) the functional properties of the DOM from Esthwaite Water are strongly influenced by the seasonally-dependent input of autochthonous DOM, derived from phytoplankton. The autochthonous DOM is less fluorescent, light absorbing, hydrophobic and has a lower acid group content and capacity to be adsorbed onto alumina than terrestrially derived allochthonous DOM; (iii) significant correlations were found between some functional properties of DOM suggesting that several functional properties of DOM may be predicted from one comparatively simple measurement.