Effects of acid sulphate on DOC release in mineral soils: the influence of SO42−retention and Al release

Dissolved organic carbon (DOC) in acid-sensitive upland waters is dominated by allochthonous inputs from organic-rich soils, yet inter-site variability in soil DOC release to changes in acidity has received scant attention in spite of the reported differences between locations in surface water DOC trends over the last few decades. In a previous paper, we demonstrated that pH-related retention of DOC in O horizon soils was influenced by acid-base status, particularly the exchangeable Al content. In the present paper, we investigate the effect of sulphate additions (0–437 μeq l−1) on DOC release in the mineral B horizon soils from the same locations. Dissolved organic carbon release decreased with declining pH in all soils, although the shape of the pH-DOC relationships differed between locations, reflecting the multiple factors controlling DOC mobility. The release of DOC decreased by 32–91% in the treatment with the largest acid input (437 μeq l−1), with the greatest decreases occurring in soils with very small % base saturation (BS, <3%) and/or large capacity for sulphate (SO42−) retention (up to 35% of added SO42−). The greatest DOC release occurred in the soil with the largest initial base status (12% BS). These results support our earlier conclusions that differences in acid-base status between soils alter the sensitivity of DOC release to similar sulphur deposition declines. However,superimposed on this is the capacity of mineral soils to sorb DOC and SO42−, and more work is needed to determine the fate of sorbed DOC under conditions of increasing pH and decreasing SO42−.

A generalised Bayesian instrumental variable approach under student t-distributed errors with application

Bayesian analysis is given of an instrumental variable model that allows for heteroscedasticity in both the structural equation and the instrument equation. Specifically, the approach for dealing with heteroscedastic errors in Geweke (1993) is extended to the Bayesian instrumental variable estimator outlined in Rossi et al. (2005). Heteroscedasticity is treated by modelling the variance for each error using a hierarchical prior that is Gamma distributed. The computation is carried out by using a Markov chain Monte Carlo sampling algorithm with an augmented draw for the heteroscedastic case. An example using real data illustrates the approach and shows that ignoring heteroscedasticity in the instrument equation when it exists may lead to biased estimates.