Fate of organic micropollutants in the hyporheic zone of a eutrophic lowland stream: Results of a preliminary field study

Abstract

Many rivers and streams worldwide are impacted by pharmaceuticals originating from sewage. The hyporheic zone underlying streams is often regarded as reactive bioreactor with the potential for eliminating such sewage-born micropollutants. The present study aims at checking the elimination potential and analyzing the coupling of hydrodynamics, biogeochemistry and micropollutant processing. To this end, two sites at the lowland stream Erpe, which receives a high sewage burden, were equipped and sampled with nested piezometers. From temperature depth profiles we determined that at one of the sites infiltration of surface water into the aquifer occurs while exfiltration dominates at the other site. Biogeochemical data reveal intense mineralization processes and strictly anoxic conditions in the streambed sediments at both sites. Concentrations of the pharmaceuticals indomethacin, diclofenac, ibuprofen, bezafibrate, ketoprofen, naproxen and clofibric acid were high in the surface water and also in the subsurface at the infiltrating site. The evaluation of the depth profiles indicates some attenuation but due to varying surface water composition the evaluation of subsurface processes is quite complex. Borate and non-geogenic gadolinium were measured as conservative wastewater indicators. To eliminate the influence of fluctuating sewage proportions in the surface water, micropollutant concentrations are related to these indicators. The indicators can cope with different dilutions of the sewage but not with temporally varying sewage composition.

Introduction

The hyporheic zone—the transition zone between surface water in streams and groundwater (Runkel et al., 2003)—is a key compartment of the hydrosphere. It is of utmost importance for maintaining the ecological function of running waters and a natural reactor taking main responsibility for the impressive self-purification capacity of lotic systems, from small brooks to large rivers. The hyporheic zone is also a barrier against contamination of near-surface aquifers, which are essential for the production of drinking water. Its ecological service is provided and sustained by the interaction of physical (e.g., transport of water and solutes), chemical (e.g., chemical reactions, sorption), and biotic processes (e.g., microbial transformation, bioturbation) by diverse and active hyporheic communities (Krause et al., 2009).

Numerous investigations on nitrogen, phosphorus and organic carbon processing in rivers highlight the high intrinsic potential of hyporheic zones as efficient bioreactors. As in other environmental systems, the cycling of nutrients and other major water compounds is closely coupled. For example, phosphorus cycling is impacted by the cycling of iron, aluminium, calcium, and sulphur (Hendricks and White, 2000, Reddy et al., 1999, Roden and Edmonds, 1997, House, 2003). While the reducing milieu in the hyporheic zone favours the unwanted mobilization of phosphate, it also favours the elimination of nitrate (Lewandowski and Nützmann, 2010). Denitrification may occur even in well-oxygenated environments due to the presence of anaerobic microhabitats (Birgand et al., 2007, Hendricks and White, 2000, Mulholland and DeAngelis, 2000). Over a range of diverse headwater streams, for example, 70–80% of ammonium removal from the flowing water was attributed to uptake and subsequent consumption in the streambed sediments (Peterson et al., 2001).

Beside nutrients, organic micropollutants such as pharmaceuticals and personal care products are present in streams and rivers. The concern about their presence is mainly related to potential adverse effects on environmental systems, their bioaccumulation potential, and to human toxicology when surface water is used for the production of drinking water. In spite of large efforts to minimize the emission of such compounds into the environment, they are ubiquitous in surface waters. Their major sources to streams are emissions from wastewater treatment plants (WWTPs). Other sources contributing to the environmental burden are for example sewer overflows, and diffuse sources like runoff from agricultural or industrial areas. In contrast to non-polar organic contaminants like polycyclic aromatic hydrocarbons (PAH) or polychlorinated biphenyls (PCB), sorption of these emerging micropollutants to sewage sludge is less efficient due to their higher polarity (Kalsch, 1999). Biodegradation of many micropollutants in WWTPs is comparatively inefficient (Zwiener and Frimmel, 2003). The assessment of pharmaceuticals by standardized tests indicated low biodegradation (Al-Ahmad et al., 1999), and thus these emerging micropollutants are frequently referred to as pseudo-persistent in the environment (Daughton, 2003). However, studies performed under experimental conditions more resembling streams or studies on bank filtration and artificial groundwater recharge show that the same compounds considered persistent in WWTPs are transformed in river sediments or other porous matrices (Schittko et al., 2004, Löffler et al., 2005, Gruenheid et al., 2008, Kunkel and Radke, 2008, Schulz et al., 2008, Radke et al., 2009). This apparent contradiction is mainly due to two reasons: i) the residence time in WWTPs is too short to allow efficient biodegradation while in the porous space of river sediments or aquifers the residence time can be much longer, and ii) the microbial community in environmental systems is much more diverse than in WWTP. In river water, transformation processes in the absence of sediment are comparatively inefficient for many micropollutants (Kalsch, 1999, Radke et al., 2009). Thus, the hyporheic zone is hypothesized to be a key compartment with major influence on the environmental fate of organic micropollutants in lotic systems.

To study the fate of sewage-borne organic micropollutants and nutrients in the stream and the hyporheic zone, conservative sewage indicators are helpful. Borate was frequently used as such an indicator due to its predominantly anthropogenic origin in many streams. However, the amount of perborate in detergents is decreasing for more than a decade now and thus borate is nowadays less suitable (Neal et al., 2010). In recent years, the use of the rare earth element gadolinium (Gd) as a sewage indicator in hydrology has been discussed (Verplanck et al., 2005). Gd compounds have been used as contrasting agent in clinical diagnosis (magnetic resonance imaging) since 1988 (Kümmerer and Helmers, 2000) and are currently used in about 150 million annual applications worldwide. Gd is usually administered as a complex (up to 0.3 mmol kg⁻¹ body weight) and excreted unmetabolised within a few hours (Kümmerer and Helmers, 2000). Such organic Gd complexes are able to reach the surface water systems because they are stable enough to pass nearly unaffected through common WWTPs (Möller et al., 2003, Künnemeyer et al., 2009). During the last decade, a number of studies from various countries have reported anthropogenic Gd in rivers (Bau and Dulski, 1996, Elbaz-Poulichet et al., 2002, Möller et al., 2003, Morteani et al., 2006). Due to its stability, low sorption tendency and low geogenic background, the use of anthropogenic Gd as a conservative indicator of urban wastewater in rivers has been proposed (Verplanck et al., 2005). Furthermore, even in the case of geogenic Gd present, the anthropogenic fraction of Gd can be calculated from the concentration pattern of the other rare earth elements (REE) in a water sample.

Most previous studies on hydrology and biogeochemistry of the hyporheic zone were conducted in small headwater streams. Water quality in headwaters is usually more pristine than in lowland streams. Due to lower flow velocities and more eutrophic conditions, bed sediments of lowland streams are usually much finer with higher organic matter content than those occurring in headwaters. Often the redox milieu in streambed sediments of lowland streams is anoxic while headwater streambed sediments are usually aerobic. Due to the texture of the streambed sediment in headwaters there is an intense exchange between surface water and the hyporheic zone. Flow velocities in the hyporheic zone are high compared to lowland streams. It is widely unknown whether unfavourable hydraulic conditions spoil the importance of the hyporheic zone in lowland streams. On the one hand, lowered exchange rates and lower flow velocities in the hyporheic zone decrease the percentage of water passing through the hyporheic zone. On the other hand, this increases the reaction time in the hyporheic zone. The underlying question is whether transformation of micropollutants in the hyporheic zone is limited by transport (of micropollutants and/or reaction partners) or kinetics. Gücker and Pusch (2006) investigated whether the paradigm of effective nutrient retention, also known as self-purification concept, derived from pristine headwater streams (e. g. Peterson et al., 2001), holds true for lowland streams with excessive nutrient loads like the stream Erpe. Haggard et al., 2001, Marti et al., 2004 reported low load-specific nutrient retention efficiencies in such streams with high sewage burden and the study of Gücker and Pusch (2006) at the lowland stream Erpe confirms this conclusion. However, an overload of the system hypothesized as cause for the low relative efficiency of nutrient retention cannot be assumed for micropollutants. Thus, we want to check whether the hyporheic zone is a compartment with paramount responsibility for transformation of micropollutants in lowland rivers.

The aims of the present preliminary study are (1) to analyze the coupling of hydrodynamics, biogeochemistry and micropollutant processing in the hyporheic zone of a lowland river, (2) to evaluate whether borate and anthropogenic gadolinium are suitable indicators for the proportion of wastewater in the hyporheic zone, and (3) to investigate the hypothesis that the hyporheic zone is an important compartment for transformation of micropollutants in the stream Erpe. Previous research of nutrient retention in the hyporheic zone (e.g. Gücker and Pusch, 2006) was usually based on tracer tests conducted in the overlying water. In tracer tests surface storage and hyporheic exchange are lumped together in a single storage zone, so it is often difficult to distinguish between surface and subsurface storage (Cardenas, 2006, Runkel et al., 2003, Wörman et al., 2007). Furthermore, hyporheic processes and their interaction cannot be studied in detail on the scale of tracer tests, and downscaling from larger scales is impossible without prior knowledge on the individual processes, especially due to the large heterogeneity of the hyporheic zone (McClain et al., 2003). Therefore, in the present study we investigate the hyporheic processes on a decimeter scale directly in the hyporheic zone. Knowledge on the aforementioned aims are of major importance since many streams and rivers in densely populated areas are sewage-polluted lowland rivers while most studies on hyporheic zone processes were conducted in pristine headwater streams so far. We have chosen the stream Erpe for our investigations because it represents a typical lowland stream with high sewage burden, and previous investigations of nutrient retention by Gücker and Pusch, 2006, Gücker et al., 2006 provide a useful basis for our study.