Background, aim and scope Australia is the largest producer of bauxite in the world, with an annual output of approximately 62 million metric dry tons in 2007. For every tonne of alumina, about 2 tonnes of highly alkaline and highly saline bauxite-processing residue are produced. In Western Australia, Alcoa World Alumina, Australia (Alcoa) produces approximately 15 MT of residue annually from its refineries (Kwinana, Pinjarra and Wagerup). The bauxite-processing residue sand (BRS) fraction represents the primary material for rehabilitating Alcoa's residue disposal areas (RDAs). However, the inherently hostile characteristics (high alkalinity, high salinity and poor nutrient availability) of BRS pose severe limitations for establishing sustainable plant cover systems. Alcoa currently applies 2.7 t ha(-1) of di-ammonium phosphate ((NH4)(2)HPO4; DAP)-based fertiliser as a part of rehabilitation of the outer residue sand embankments of its RDAs. Limited information on the behaviour of the dominant components of this inorganic fertiliser in highly alkaline BRS is currently available, despite the known effects of pH on ammonium (NH4) and phosphorus (P) behaviour. The aim of this study was to quantify the effects of pH on NH3 volatilisation and residual nitrogen (N) in BRS following DAP applications. Methods The sponge-trapping and KCl-extraction method was used for determining NH3 volatilisation from surface-applied DAP in samples of BRS collected from each of Alcoa's three Western Australia Refineries (Kwinana, Pinjarra, Wagerup) under various pH conditions (pH 4, 7, 9 and 11). Following cessation of volatilisation, the residual N was extracted from BRS using 2 M KCl and concentrations of NH4+-N and NO3--N were determined by flow injection analysis. Results The quantities of NH3 volatilised increased dramatically as the pH increased from 4 to 11. Much of the N lost as NH3 (up to 95.2%) occurred within a short period (24 h to 7 days), particularly for the pH 9 and 11 treatments. Concentrations of residual NH4+-N recovered in DAP-treated BRS at the end of the experiment decreased with increasing pH. This finding was consistent with increasing loss of N via volatilisation as pH increased. The concentration of NO3--N was very low due to no nitrification in BRS. Discussion The pH was a key driver for NH3 volatilisation from DAP-treated BRS and primarily controlled N dynamics in BRS. Results indicate that NH4 not adsorbed by BRS was highly susceptible to volatilisation. The likely lack of nitrifying bacteria did not allow conversion of ammonium to nitrate, thereby further exacerbating the potential for loss via volatilisation Conclusions It was demonstrated that the pH is the key factor controlling the loss of inorganic N from BRS. Although volatilisation was considerably lower at pH 4, achieving this pH reduction in the field is not possible at present. Findings from this study highlight the need to better understand which forms of N fertiliser are most suitable for use in highly alkaline BRS. Recommendation and perspectives Although pH reduction is the most likely means of stopping NH3 volatilisation in BRS, it is economically and operationally unfeasible to add sufficient acidity for adequately lowering pH in the BRS for revegetation. More attention on forms of fertilisers more suitable to highly alkaline, microbially inert soil conditions appears to be warranted.
