Column studies to assess the effects of climate variables on redox processes during riverbank filtration.

Riverbank filtration is an established technique used world-wide to produce clean drinking water in a reliable and cost-efficient way. This practice is, however, facing new challenges posed by climate change, as already observed during past heat waves with the local occurrence of anoxic conditions. In this study we investigated the effect of direct (temperature) and indirect (dissolved organic matter (DOM) concentration and composition, flow rate) climate change variables on redox processes (aerobic respiration, denitrification and Mn(III/IV)/Fe(III) reduction) by means of column experiments. Natural river water, modified river water and river water mixed with treated wastewater effluent were used as feed waters for the columns filled with natural sand from a river-infiltration system in Switzerland. Biodegradable dissolved organic matter was mainly removed immediately at the column inlet and particulate organic matter (POM) associated with the natural sand was the main electron donor for aerobic respiration throughout the column. Low infiltration rates (≤0.01 m/h) enhanced the oxygen consumption leading to anoxic conditions. DOM consumption did not seem to be sensitive to temperature, although oxygen consumption (i.e., associated with POM degradation) showed a strong temperature dependence with an activation energy of ∼70 kJmol(-1). Anoxic conditions developed at 30 °C with partial denitrification and formation of nitrite and ammonium. In absence of oxygen and nitrate, Mn(II) was mobilized at 20 °C, highlighting the importance of nitrate acting as a redox buffer under anoxic conditions preventing the reductive dissolution of Mn(III/IV)(hydr)oxides. Reductive dissolution of Fe(III)(hydr)oxides was not observed under these conditions.

NOM degradation during river infiltration: effects of the climate variables temperature and discharge.

Most peri-alpine shallow aquifers fed by rivers are oxic and the drinking water derived by riverbank filtration is generally of excellent quality. However, observations during past heat waves suggest that water quality may be affected by climate change due to effects on redox processes such as aerobic respiration, denitrification, reductive dissolution of manganese(III/IV)- and iron(III)(hydr)oxides that occur during river infiltration. To assess the dependence of these redox processes on the climate-related variables temperature and discharge, we performed periodic and targeted (summer and winter) field sampling campaigns at the Thur River, Switzerland, and laboratory column experiments simulating the field conditions. Typical summer and winter field conditions could be successfully simulated by the column experiments. Dissolved organic matter (DOM) was found not to be a major electron donor for aerobic respiration in summer and the DOM consumption did not reveal a significant correlation with temperature and discharge. It is hypothesized that under summer conditions, organic matter associated with the aquifer material (particulate organic matter, POM) is responsible for most of the consumption of dissolved oxygen (DO), which was the most important electron acceptor in both the field and the column system. For typical summer conditions at temperatures >20 °C, complete depletion of DO was observed in the column system and in a piezometer located only a few metres from the river. Both in the field system and the column experiments, nitrate acted as a redox buffer preventing the release of manganese(II) and iron(II). For periodic field observations over five years, DO consumption showed a pronounced temperature dependence (correlation coefficient r = 0.74) and therefore a seasonal pattern, which seemed to be mostly explained by the temperature dependence of the calculated POM consumption (r = 0.7). The river discharge was found to be highly and positively correlated with DO consumption (r = 0.85), suggesting an enhanced POM input during flood events. This high correlation could only be observed for the low-temperature range (T < 15 °C). For temperatures >15 °C, DO consumption was already high (almost complete) and the impact of discharge could not be resolved. Based on our results, we estimate the risk for similar river-infiltration systems to release manganese(II) and iron(II) to be low during future average summer conditions. However, long-lasting heat waves might lead to a consumption of the nitrate buffer, inducing a mobilization of manganese and iron.

Effect of iron and phosphate on bacterial cyanide formation determined by methemoglobin in two-dimensional gradient microcultivations.

Cyanide formation by microorganisms is typically observed during early stationary growth phase and is facilitated by the presence of iron(III) and inorganic phosphate. Extracellular free cyanide in aqueous solutions might readily react with methemoglobin added and can be determined by UV/VIS spectroscopy. As alternative to existing methods, this provided the basis for an analytical method which has not been used previously for the determination of cyanide in bacterial cultivations. We successfully applied the technique to study the combined effect of both iron(III) and phosphate on the cyanide formation by Pseudomonas fluorescens, which we used as model organism known for its ability to form HCN. Information on the combined effect of iron and phosphate was obtained by using commercially available 24-well microtiter plates as two-dimensional gradient systems. After its reaction with methemoglobin, cyanide was measured reproducibly at the wavelength of 427 nm. We found a combined effect of both iron and phosphate. In the absence of inorganic phosphate, cyanide formation was stimulated considerably by increasing concentrations of iron(III), although the effect on the bacterial growth of P. fluorescens was almost insignificant. This suggests that iron is more important than inorganic phosphate for the cyanogenesis by P. fluorescens.