Does exposure to domestic wastewater effluent (including steroid estrogens) harm fish populations in the UK?
Graphical abstract
Introduction
For thousands of years man's activities have disturbed the river environment. The river can be exploited as a food, drinking water and irrigation resource, used as a highway for goods transport, a generator of energy, and a conduit for our waste products. Rivers are also feared as a source of flooding, so they may be excavated to ensure they act as efficient drains. Many of these human activities have had damaging impacts on the river as a habitat for fish. The fish that live in our rivers are at, or near, the top of a complex food web. Unfortunately, the abundance of fish in rivers have not been consistently recorded through history, but it would appear that serious declines in some major rivers in the UK occurred from the 1930s to 1950s. Inadequate treatment of sewage and industrial waste led to the disappearance of fish in the lower reaches of big rivers like the Trent (Mann, 1989), Mersey (Jones, 2006) and Thames rivers (Wheeler, 1979). Fortunately, an increasing appreciation of the amenity value of rivers, legislation, industrial decline, and more investment in water treatment has largely eliminated the problem of gross organic pollution, at least in the UK, with the exception of occasional combined sewer overflows. However, it has been increasingly recognised that as individuals we now consume many more pharmaceuticals and personal care products (PPCPs) than ever before. Sewage treatment plants (STPs) were never designed to remove all of such micropollutants. Could it be that we are now harming our river environment and fish through this insidious ‘invisible’ pollution (Daughton and Ternes, 1999)?
When we examine the tissue of freshwater wild fish, we can certainly find many hydrophobic pollutants present (Jurgens et al., 2015), but what evidence do we have that chemicals can harm fish individuals and populations? There are, of course, examples of extreme one-off pollution events with industrial, oil and farm waste killing fish (Giger, 2009, Kubach et al., 2011, Kennedy et al., 2012, Eros et al., 2015). But our concern here is with chronic pollution. The strongest evidence seems to be related to metals. Soil acidification thanks to ‘acid rain’ from coal combustion led to the release of the toxic monomeric forms of Al into upland streams and lakes, leading to fish kills in the 70s and 80s (Henriksen et al., 1984). Freshwaters with high metal concentrations associated with mine waste or heavy industry have also had a recorded impact on fish populations (Filipek et al., 1987).
Thus, there are examples of fish kills due to exposure to acutely toxic chemicals at pollution hot-spots. But what of the chemicals routinely discharged in domestic sewage effluent? The chronic sub-lethal phenomena of endocrine disruption, associated with sewage effluent, has had and continues to have a major influence on our thinking regarding PPCPs. There is overwhelming evidence that a ubiquitous component of sewage effluent has led to endocrine disruption effects in resident wild roach (Rutilis rutilis) (Jobling et al., 1998, Jobling et al., 2006). The most likely agents being the natural and synthetic steroid estrogens excreted by humans (Desbrow et al., 1998). Similarly, there is evidence that increasing exposure to wastewater effluent elevates the level of the stress hormone cortisol in fish, at least in stickleback (Pottinger et al., 2016). Recently, a disastrous decline in Asian vultures has been strongly linked to the non-steroidal anti-inflammatory agent diclofenac (Oaks et al., 2004). Given that diclofenac is a common constituent of sewage effluent, this has now risen as a concern for fish in rivers too (Schwaiger et al., 2004, Cuklev et al., 2011). So now both the steroid estrogens and diclofenac have been identified by the European Union as requiring special monitoring, with a view to control at a later stage (COM(2011)876). It is also recognised that freshwater fish will be exposed to a wide range of pharmaceuticals and this chronic exposure is a concern (Fent et al., 2006). Given the fear and uncertainty over this chronic exposure to PPCPs, there are increasing arguments that an end of pipe solution at STPs will be needed to protect aquatic wildlife (Eggen et al., 2014, Oehlmann et al., 2014, Stamm et al., 2015). But is this fear justified? We know that if the synthetic estrogen ethinylestradiol reaches a high enough level some fish populations will collapse (Kidd et al., 2007). It can be presumed that our consumption of PPCPs has been growing steadily since the 1970s (Richardson and Ternes, 2014), so it would seem a reasonable question to ask how fish populations have fared since then? Rather surprisingly, examining responses in the abundance of wildlife populations to chemical or estrogen exposure has not been a frequently asked question in the aquatic environment (Mills and Chichester, 2005, Johnson and Sumpter, 2016). In contrast, such approaches are seen as central in the terrestrial environment, such as with neonicotinoid pesticides and bees (Woodcock et al., 2016).
Unfortunately, until recently there has been little systematic collection of data on fish populations in rivers. However, some species that were relatively common in many UK lowland rivers have declined or disappeared, was this due to chemicals or estrogens even? These include the migrating salmonids (Salmo salar and Salmo trutta) and Barbel (Barbus barbus) but these declines are most closely linked with habitats becoming unsuitable (Johnson and Sumpter, 2014). We are sadly aware that there has been a decline in eel numbers in many parts of the world. But the evidence suggests that the eel decline, which started in the early 1980s, occurred in a period of reduced chemical challenge (Jurgens et al., 2015). Eel populations appeared to have done better in the much more polluted post-war period. There are, however, quite a lot of encouraging information on cyprinid fish, such as bream (Abramis brama), whose average length for 5 year olds increased from 1966 to 1976 in the Dutch Rhine (Slooff and Dezwart, 1983) and whose condition steadily improved in several major German rivers from 1992 to 2014 (Teubner et al., 2015). Data appear to show that UK cyprinid populations have been recovering since reaching a low-point in the 1950–1970s period (Mann, 1989, Robinson et al., 2003). However, although encouraging, the limited information available is too coarse and not sufficiently focused to address whether the chemicals routinely present in domestic sewage effluent are harming wildlife populations.
To begin addressing the question in a more systematic way, we compared routine fish population monitoring data collected in the UK by the Environment Agency of England and Wales with predicted wastewater effluent exposure. This study tested the following hypotheses:
- •
Any fish population (average density) will be severely harmed by average exposure to domestic wastewater
- •
Any roach population will be severely harmed by temporal increases in domestic wastewater exposure
It should be pointed out the intention of this study was not to identify the most important environmental factors that stimulate fish population abundance and aid recruitment in UK rivers. The complex interactions of flow, temperature, habitat, disease, and position of the Gulf Stream in the North Atlantic, amongst others, are all likely to be playing a role together. Nor will simple population data, such as we use here, reveal sub-lethal impacts that could hamper fish performance and well-being. The aim was to see whether it was possible to rule out sewage and estrogen exposure as having a consistent and seriously damaging impact on fish populations.
Section snippets
Fisheries monitoring data
The fisheries data were collected for the National Fisheries Monitoring Programme by the Environment Agency of England and Wales. Only sites where the electro-fishing method was used for counting were examined. The method involves a boom boat applying a 50 Hz pulsed DC current to the water. Downstream runs may be up to 2 km between dividing locks or be of shorter duration, such as around islands or weir pools (Table 1). The sampling runs were mainly carried out in close proximity to the river
Fish density compared to estrogen (effluent) exposure
Depending on the site, the fisheries monitoring records start from 1995 to 2004, and thus the average density of fish per site were calculated based on a minimum of 6 to a maximum of 17 years of fisheries data (Table 1). The predicted mean EEQ exposure at these sites ranged between 0.6 ng/L and 3.2 ng/L, a five-fold difference (Table 2). If we were to assume water use of 200 L per capita per day, then this would represent a wastewater content of 3 to 18% in the river. Whilst the 90%ile exposure the
Conclusions
At 38 sites across England (UK), the density of roach, bleak, dace and perch populations over a period of 6 to 17 years, starting from the early 2000 period, were not obviously linked to estrogen (sewage effluent) exposure. Hence it is possible to conclude that wastewater was not a clearly damaging factor on fish density. As a test case, the temporal rises and falls of roach populations in the middle Thames and Great Ouse were compared over several years with the preceding 12 months of sewage
Supplementary data
Acknowledgements
The authors are grateful for support from the CEH Science budget provided by NERC and fisheries data provided by the Environment Agency of England and Wales. The authors are grateful for helpful advice from Francois Edwards and John Sumpter.
References (63)
The use of modelling to predict levels of estrogens in a river catchment: how does modelled data compare with chemical analysis and in vitro yeast assay results?
Sci. Total Environ.
(2010)Ecotoxicology of human pharmaceuticals
Aquat. Toxicol.
(2006)A catchment-based water resource decision-support tool for the United Kingdom
Environ. Model. Softw.
(2005)Water quality and fisheries in the Mersey estuary, England: a historical perspective
Mar. Pollut. Bull.
(2006)PCB and organochlorine pesticide burden in eels in the lower Thames River (UK)
Chemosphere
(2015)The influence of environmental factors on movements of lowland-river fish in the Yorkshire Ouse system
Sci. Total Environ.
(2000)- et al.
Review of evidence: are endocrine-disrupting chemicals in the aquatic environment impacting fish populations?
Sci. Total Environ.
(2005) A comparison of two methods for the assessment of stress axis activity in wild fish in relation to wastewater effluent exposure
Gen. Comp. Endocrinol.
(2016)Toxic effects of the non-steroidal anti-inflammatory drug diclofenac part 1: histopathological alterations and bioaccumulation in rainbow trout
Aquat. Toxicol.
(2004)- et al.
The growth, fecundity and mortality of bream (Abramis brama) from polluted and less polluted surface waters in the Netherlands
Sci. Total Environ.
(1983)