No abstract

No abstract

Twenty sediment traps were deployed for a year in five arrays at three locations in a seasonal anoxic basin in Esthwaite Water, U.K. Accumulated iron and manganese were measured fortnightly, as were vertical profiles of temperature, oxygen, total iron and manganese, polarographically determined Fe(II) and Mn(II), and light attenuance. Most of the iron and manganese flows into the lake in winter and is nearly all caught in sediment traps in deep hypolimnetic water, indicating that both elements are transported to the sediment. Accumulation rates in the sediments show that the iron is retained with minimal loss. Although dissolved iron increases to very high concentrations in the hypolimnion during the period of summer anoxia, this accumulation only accounts for a small fraction of the annual iron loading to the lake. The overall cycling of iron in the lake is consistent with a simple conception of sediment‐water interactions. When a redox boundary exists in the water column there is a separate cycle of iron due to vertical transport of ferrous iron by eddy diffusion, oxidation to Fe(III) by oxygen, sinking of the resultant ferric particles, and redissolution to ferrous iron. The manganese cycle contrasts markedly with that of iron. The manganese which reaches the sediment in winter is rapidly reduced and released to the overlying oxic water so that <10% is permanently accumulated in the sediment. During summer the manganese never reaches the sediment because it is reduced and accumulated in the anoxic hypolimnion. Relatively little manganese is supplied from the sediment during this time because the supply of particles is interrupted. Although most of the manganese which enters the lake is washed out again it will nearly all have undergone a redox cycle.

WHAM, an equilibrium chemical model for soils, waters and sediments, centred on a discrete-site/ electrostatic model of humic substances (HS), has been used to analyse batch titration data for organic and mineral horizons of acid soils. In most cases, tolerable fits were obtained by optimizing the soil contents of HS and aluminium, while keeping the model parameters (site densities, equilibrium constants, electrostatic terms) fixed. The optimized contents agreed reasonably with those estimated by chemical extraction. For some mineral soil samples, low in HS and high in aluminium, fitting of the titration data was improved by assuming the formation and dissolution of AI(OH)3 and adjusting its solubility product. Solid-solution distributions of base cations (Na+, Mg2+, K + , Ca2+, NH: ) could be explained by non-specific counterion accumulation, with a small degree of selectivity. The WHAM sub-model for fulvic acid sorption accounted approximately for observed aqueous-phase concentrations of organic carbon and organically-complexed aluminium.

Summary The pH buffering and aluminium solubility characteristics of acid soil are important in determining the soil's response to changes in precipitation acidity. The chemistry of soil organic matter (humic substances) plays a key role in both processes, yet is complex and still poorly understood. Nevertheless, models of humic substance chemistry have been developed, one of which is WHAM–S, which contains a model (Model V) of proton and metal binding at discrete sites on humic substances and considers electrostatic effects on the binding strength. Here we have tested the ability of WHAM–S to model solution pH and Al using batch titration studies on organic and mineral soil horizons from forested sites in Norway, Germany and Spain, with ambient pH values from 3.73 to 5.73. We optimized the model predictions by adjusting the amounts of soil aluminium and humic substances within defined limits, taking the contents of copper chloride‐extractable Al and the base‐extractable organic matter as starting values. The model simulated both pH and dissolved Al well with optimized amounts of aluminium and humic substances within the defined limits (root mean squared error for pH from 0.01 to 0.22, for p[Al]aq (total dissolved Al) from 0.03 to 0.49, five data points). Control of dissolved Al by dissolved organic matter was important particularly at above‐ambient pH. In two mineral horizons we improved the fits by assuming that Al could precipitate as Al(OH)3. The optimized model also gave reasonable predictions of pH and dissolved Al in supernatants obtained by repeated leaching of the soil horizons. The results show that humic substances dominate the control of pH and dissolved Al in most of the horizons studied. Control by Al(OH)3 occurs but is the exception.