Phosphorus in groundwater discharge – A potential source for lake eutrophication
Introduction
Phosphorus (P) overloads are still a major threat to lake ecosystems worldwide. As a limiting nutrient P often controls the trophic state of temperate freshwater systems (Heathwaite et al., 2005, Sondergaard and Jeppesen, 2007). After the significant reduction of P from point sources to improve freshwater quality it became more and more obvious that diffuse transport of P also has a critical ecological relevance. Nowadays, many studies claim agriculture is the main source of diffuse P in freshwater systems (Heathwaite et al., 2005, Orderud and Vogt, 2013, Withers and Haygarth, 2007), especially since sewage discharges from point sources have been eliminated to a large extent (Orderud and Vogt, 2013). Depending on site conditions (i.e. inclination, sediment retention capacity, etc.) diffuse P transport occurs as particulate or dissolved P in overland flow, channelized surface runoff, drainage, or groundwater. In groundwater natural dissolved P concentrations are usually low, since potentially mobile P (i.e. in general orthophosphate) is adsorbed in the soil and sediment matrix either in the vadose or the saturated zone. As a consequence groundwater was evaluated to be of “low source strength” by Edwards and Withers (2007). However, it needs to be accepted that dissolved P concentrations can increase largely above natural background conditions in groundwater. Interestingly, studies have again found wastewater to cause heavy groundwater P contaminations (Ptacek, 1998, McCobb et al., 2003, Robertson, 2008, Roy et al., 2009), although this was considered to be eliminated as a nutrient source with the reduction of point sources. However, especially among practitioners it still is a common paradigm that P is completely immobile in groundwater. This might also be supported by a generally low data basis on this issue. Since P is non-hazardous for human health it is often not regularly monitored, neither in drinking water nor in groundwater. This is one of the reasons why lacustrine groundwater discharge (LGD) is often dismissed as a major source of external P inputs to lakes. In recent years the awareness of groundwater P is growing and it is becoming more and more accepted that groundwater can indeed have P concentrations exceeding thresholds of ecological relevance (e.g. Burkart et al., 2004, Holman et al., 2010, Kidmose et al., 2013).
Studies on groundwater P often deal with the determination of P concentrations on the catchment scale in order to determine natural background concentrations and to separate them from contamination-derived concentrations. Based on these findings thresholds are raised and discussed for groundwater discharging into surface waters (Burkart et al., 2004, Lewandowski et al., 2015). So far only a few studies tried to actually quantify groundwater-borne P loads to lakes and rivers and to evaluate the impact on their trophic state (Ala-aho et al., 2013, Jarosiewicz and Witek, 2014, McCobb et al., 2003, Ouyang, 2012, Shaw et al., 1990). However, the quantification of LGD-derived P loads is difficult. Usually LGD volume fluxes and nutrient concentrations are determined separately and are subsequently multiplied. Both, hydrological (i.e. LGD volume fluxes) and geochemical (i.e. nutrient concentrations) factors may be affected by spatial and temporal heterogeneities, which impede the empirical determination of representative values. Simplification and upscaling of point measurements are often necessary to approximate nutrient loads. In many studies the groundwater path is simply considered as the residual in budget calculations (Rosenberry et al., 2015) or even is completely neglected.
With the present study we aim to provoke an intensified discussion on the potentially harmful contribution of groundwater to lake nutrient budgets and to demonstrate that groundwater P can fuel eutrophication of lakes. The study site is Lake Arendsee in Northern Germany where the mean total phosphorus (TP) concentrations in the lake water showed a gradual increase in the past decades to more than 150 μg l−1. First investigations indicated a large spatial variability of near-shore groundwater P concentrations, with concentrations of soluble reactive P (SRP) higher than 1000 μg l−1 at one site. These results enforced the effort to better understand and quantify LGD and its contribution to the nutrient budget of Lake Arendsee.
As a first step towards the determination of groundwater-borne P loads detailed investigations on LGD volume fluxes and patterns were conducted (Lewandowski et al., 2013, Meinikmann et al., 2013). To incorporate spatial heterogeneity of LGD, the shoreline was subdivided into sections of about 200 m length, for which individual volume fluxes of LGD were calculated. Based on these results P loads were calculated by applying P concentrations of four near-shore groundwater observation wells (sites 1–4 in Fig. 1). This resulted in a groundwater-derived P load of 425 kg yr−1. However, it was hypothesized that detailed spatial information on groundwater P concentrations increases the accuracy of P load calculations. The present study focuses on groundwater P concentrations as the second factor of groundwater-borne P loads (volume fluxes × concentration) to Lake Arendsee. We hereby aim to (1) localize crucial areas for P input by detailed measurements of P concentrations in near-shore groundwater, (2) calculate LGD-derived P loads and evaluate them within the context of overall P inputs to the lake, as well as (3) localize the origin of the P contamination in the catchment of the lake.
Section snippets
Study site
Lake Arendsee in Northern Germany is 5.14 km2 in size. As already described in previous studies (e.g. Hupfer and Lewandowski, 2005, Meinikmann et al., 2013) it is a deep stratified lake (max. depth 49 m, mean depth 29 m) which was originally solely groundwater-fed. Currently, four ditches draining adjacent agricultural fields discharge into the lake and an artificial runoff channel transports water out of the lake (Fig. 1). Since the middle of the last century the lake is eutrophied. The annual
Groundwater observation wells
Mean SRP concentrations at the eight monitoring sites (Fig. 1) vary by orders of magnitudes (Fig. 3). Most remarkable are the results of sites 3 and (upgradient to it) site 5 where the shallow wells have mean SRP concentrations of 1600 and 3900 μg l−1, respectively. The deeper wells at these sites show concentrations of 650 and 610 μg SRP l−1, respectively. At near-shore sites 2 and 6 SRP concentrations also decrease with increasing aquifer depths. At both sites the shallowest wells still have SRP
The role of LGD at Lake Arendsee
Groundwater P concentrations found in the catchment of Lake Arendsee partly exceed ecological thresholds discussed in literature by far (Burkart et al., 2004, Holman et al., 2010, Lewandowski et al., 2015). Moreover, main LGD takes place where the groundwater is most contaminated. As a result, LGD-derived P loads from the segmented approach account for 53% of all quantified external P inputs to Lake Arendsee, compared to only 9% based on natural background concentrations of P. Temporary
Summary and conclusions
- 1.
Completing the segmented approach introduced by Meinikmann et al. (2013) with near-shore groundwater P concentrations from temporary piezometers enabled us to quantify LGD-derived P loads and to evaluate their impact on the lake’s trophic state.
- 2.
Groundwater can be a main cause of lake eutrophication, especially when a contamination leads to high nutrient concentrations in those parts of the shoreline where main lacustrine groundwater discharge (LGD) takes place.
- 3.
Despite a large number of
Acknowledgements
Sincere thanks are given to Christine Sturm for groundwater sampling as well as the team of the chemical lab at the Leibniz-Institute of Freshwater Ecology and Inland Fisheries Berlin (IGB). This study was funded by the State Agency for Flood Protection and Water Management Saxony-Anhalt (LHW).
References (56)
- et al.
Interaction of esker groundwater with headwater lakes and streams
J. Hydrol.
(2013) - et al.
Groundwater seepage nutrient loading in a Florida lake
Water Res.
(1985) - et al.
Groundwater seepage as a nutrient source to a drainage lake – Lake Mendota, Wisconsin
Water Res.
(1982) A tiered risk-based approach for predicting diffuse and point source phosphorus losses in agricultural areas
Sci. Total Environ.
(2005)An assessment of the risk to surface water ecosystems of groundwater P in the UK and Ireland
Sci. Total Environ.
(2010)- et al.
Using Cl/Br ratios and other indicators to assess potential impacts on groundwater quality from septic systems: a review and examples from principal aquifers in the United States
J. Hydrol.
(2011) - et al.
Localization of lacustrine groundwater discharge (LGD) by airborne measurement of thermal infrared radiation
Remote Sens. Environ.
(2013) - et al.
Lacustrine groundwater discharge: combined determination of volumes and spatial patterns
J. Hydrol.
(2013) - et al.
Ground-water inflow of nutrients to a lake from differently utilized catchments
Water Res.
(1992) Estimation of shallow groundwater discharge and nutrient load into a river
Ecol. Eng.
(2012)