Aluminium speciation in streams and lakes of the UK Acid Waters Monitoring Network, modelled with WHAM

Abstract

The Windermere Humic Aqueous Model (WHAM) incorporating Humic Ion-Binding Model VI was applied to analytical data from the United Kingdom Acid Waters Monitoring Network, collected for 22 streams and lakes over the period 1988–2007, to calculate the chemical speciation of monomeric aluminium (Al_mon) in 3087 water samples. Model outputs were compared with analytical measurements of labile and non-labile Al_mon concentrations, the former being equated with inorganic forms of Al_mon and the latter with organically-complexed metal. Raw analytical data were used, and also data produced by applying a correction for the possible dissociation of organically-complexed Al_mon, and therefore its underestimation, during passage through the analytical cation-exchange column. Model calibration was performed by finding the conversion factor, F_FADOC, between the concentration of isolated fulvic acid, with default ion-binding properties, required by the model, and the measured concentration of dissolved organic carbon, [DOC]. For both uncorrected and corrected data, the value of F_FADOC for streams was greater than for lakes, indicating greater binding activity towards aluminium. Model fits were better using uncorrected analytical data, but the values of F_FADOC obtained from corrected data agreed more closely with previous estimates. The model provided reasonably good explanations of differences in aluminium speciation between sampling sites, and of temporal variations at individual sites. With total monomeric concentration as input, WHAM calculations might substitute for analytical speciation measurements, or aid analytical quality control. Calculated Al³⁺ activities, a_Al3+, showed a pH-dependence similar to that previously found for other surface waters, and the modelling exercise identified differences between waters of up to two orders of magnitude in the value of a_Al3+ at a given pH. The model gives the net charge of dissolved organic matter, which is calculated to have risen significantly at 15 of the AWMN sites, due to increases in pH and decreases in aluminium concentration.

Introduction

Aluminium is mobilised from catchment minerals by atmospherically-deposited sulphur and nitrogen (Johnson et al., 1981) and its concentration is a key variable in the monitoring and assessment of acidified surface waters. In particular, inorganic monomeric aluminium is widely regarded as a measure of potential toxicity, because organically-complexed metal is considered not to be bioavailable, or at least less bioavailable (Gensemer and Playle, 1999). Furthermore, Al³⁺ and its cationic hydrolysis products are important as competitors with trace metals for binding to natural organic matter and mineral surfaces (Tipping et al., 2002, Tipping, 2005). Thus information about Al speciation in surface waters, and quantitative modelling of the reactions involved, are desirable. An important aspect of research efforts in this area is the testing of models with field data, which is the subject of the present paper.

The sampling programme of UK Acid Waters Monitoring Network (AWMN) was started in 1988, and involves the regular collection and analysis of lake and stream samples at sites throughout Great Britain and Northern Ireland (Table 1, Fig. 1). The monitoring sites were chosen because of their perceived susceptibility to acidification and therefore the waters are generally acidic and soft. Monthly measurements are made of chemical and biological variables for the streams, and quarterly measurements are made for the lakes. Information and data summaries are given in a special journal issue (see Monteith and Evans, 2005, and following papers), and in reports available from the network's website (http://www.ukawmn.ucl.ac.uk). The routine analyses include the fractionation of monomeric aluminium by the cation exchange column method of Driscoll (1984), which distinguishes labile and non-labile fractions, the former being a measure of Al³⁺ together with its inorganic complexes with OH⁻, SO₄²⁻ and F⁻. For the period to 2007, the monitoring and analyses generated measured Al speciation data, suitable for comparison with modelled speciation, for 3087 surface water samples (Table 1).

The main purpose of the present work was to compare the AWMN Al speciation results with calculated values from the Windermere Humic Aqueous Model (WHAM; Tipping, 1994, Tipping, 1998, Tipping, 2002), which computes the full chemical speciation of a water sample, including Al species, using as inputs pH and the total dissolved concentrations of base cations, strong acid anions and dissolved organic matter. First, the exercise can be considered a test of the model. Second, it allows us to assess whether modelled values might (a) substitute for future routine measurements within the AWMN, in order to save effort and cost, and (b) extend the application of speciation information to other samples for which only total aluminium concentrations are measured. A third application of the results is to estimate Al³⁺ activity and its dependence on pH, in order to compare the results with those from an earlier study by Tipping (2005) using published Al speciation data from a number of other surface waters. As well as using the reported concentrations of labile and non-labile Al, we explored the use of the correction procedure given by Backes and Tipping (1987) that takes account of the tendency for the cation-exchange speciation method to underestimate [Al_mon,NL] because of partial dissociation of Al-DOM complexes during passage through the fractionation column.

Of related interest is the net charge of dissolved organic matter (DOM), which is considered an indicator of its mobility in soil–water systems (de Wit et al., 2007). To date, charge on DOM in natural waters has been estimated from ionic balance (Kopáček et al., 2000, de Wit et al., 2007), or by simple modelling that takes into account the binding of protons, but not metals, to DOM (Oliver et al., 1983, Lawrence et al., 2007). By estimating the binding of all cationic species, including the hydrolysis products of Al³⁺, WHAM provides more complete estimates of the DOM charge, and this is done as part of the main speciation calculation.

In this paper, we refer to total, labile and non-labile monomeric aluminium as Al_mon,T, Al_mon,L and Al_mon,NL respectively. Square brackets – [] – indicate concentrations, which are either in μg l⁻¹, as reported analytically, or μmol l⁻¹ which is more relevant to speciation calculations. The activity of Al³⁺ is denoted by a_Al3+ Activities or concentrations of Al³⁺ are given in mol l⁻¹. The abbreviations DOC and DOM refer to dissolved organic carbon and dissolved organic matter. In comparing measured and modelled concentrations, we explicitly equate Al_mon,NL with Al complexed to DOM, and Al_mon,L with inorganic forms.