Report No FR/D0022

JULY 1995

EXECUTIVE SUMMARY

1. The work reported here follows on from an earlier DoE-funded study which had shown that treated sewage effluent discharges were oestrogenic to fish. The report describes research conducted in 1992-95 to assess whether oestrogenic effects could be detected in male fish held in cages at selected sites in rivers below sewage treatment works, and at downstream points either where raw water is abstracted for treatment before public supply, or in raw water storage reservoirs1. The basic survey technique involved holding caged male rainbow trout in the river or raw water storage reservoir for 3-6 weeks and measuring production of a protein (vitellogenin or VTG) which is a specific marker for the presence of oestrogens. Production of vitellogenin is confined to egg-laying vertebrates and is described as an oestrogenic effect because in female egg-laying vertebrates it is usually controlled by hormones known as oestrogens. The research also set out to determine possible causes of the observed oestrogenic responses; it seemed probable that some substances in sewage effluents could be acting as oestrogens in the male fish.
   
2. Of the 15 raw water storage reservoirs1 that were surveyed, it was shown that none produced a vitellogenin response in caged fish. On the basis that biologically significant amounts of substances, which are responsible for the oestrogenic response (in fish), were not detected in all 15 raw water storage reservoirs most likely to be affected, there is no evidence of a risk to drinking water supplies, and for the present there is no need for further research.
   
3. The research covered stretches of six rivers: the Sussex Arun, the Kent Stour, the Hertfordshire Lea, the Essex Chelmer, the Suffolk Stour and the West Yorkshire Aire. In five of the rivers, work focused on a single stretch below a particular sewage treatment works (STW) outfall, while in the Lea surveys were made downstream of several STW discharges. The Aire was chosen because it is known to contain alkyl phenol ethoxylates from wool scouring; it is not used for the abstraction of raw water2. In all, 10 river stretches were surveyed.
   
4. For four of the 10 surveyed stretches (in the Lea, the Kent Stour and the Chelmer), the oestrogenic response in caged fish was confined to the undiluted sewage effluent discharges and was not observed downstream. For five of the 10 river stretches, in the Arun, Lea and Aire, the response was observed close to the sewage effluent discharges and downstream; on the Arun the response was observed up to 1.5 km downstream of the discharge and was not detected further downstream; on the Lea and the Aire the response was detected up to 5 km below the particular discharges, although the response in the Lea declined rapidly with distance downstream. The magnitude of the response in these five river stretches varied widely, from relatively small responses in the Arun through moderate responses in the Lea, to large responses in the Aire. For the remaining surveyed stretch on the Suffolk Stour, neither the sewage effluent discharge nor the river downstream was oestrogenic to caged fish.
   
5. Of the four sites that were surveyed close to points on the rivers Kent Stour, Chelmer, Essex Stour and Lea where raw water is abstracted for treatment prior to entering public supply, it was shown that none produced a vitellogenin response in caged fish.
   
6. In the case of the Aire, heavy stimulation of vitellogenin synthesis in caged male trout to levels similar to those in gravid females occurred at all sites tested in summer 1994 over a 5 km stretch downstream of Marley STW. This synthesis of vitellogenin was accompanied by reduced testicular development. A similar effect on the testes was seen in fish deployed at seven sites on the Lea during summer 1992. These observations show that exposure to oestrogenic substances can have effects on fish with potential implications for their ability to breed normally. However, the implications of these short term observations of reduced testicular growth for fish reproduction and populations are not yet known.
   
7. Dilution experiments at a STW on the Lea which had strongly oestrogenic effluent showed that two or more times dilution of unfiltered effluent was sufficient to prevent the vitellogenin response in fish. Because STWs are often sited along river stretches where the effluent is diluted at least two times, the results from the dilution study could suggest that the majority of STW discharges are unlikely to cause oestrogenic effects in fish held in their receiving waters. However, additional dilution experiments with different types of discharge would be required to confirm this point. These experiments also suggest that oestrogenic effects are likely to be greater in summer when dilution is less than in winter. Indeed, the smaller responses seen in the River Lea in winter support this prediction, although other influences such as temperature could have contributed to these effects.
   
8. The likely ecological effects of oestrogenic river water have not yet been established, although it is clear that oestrogens have the potential to disrupt reproduction in vertebrates such as fish, perhaps leading to long term declines in population size. Limited evidence presented in this report shows that some free-living wild fish (roach) are responding to oestrogenic substances in the environment so the possibility exists of localised ecological impacts.
   
9. Whilst the identity of the oestrogens in sewage effluent, responsible for these effects, is still largely unknown, this project provided useful data about the oestrogenicity of a number of substances known to be present in sewage effluent. Alkyl phenol ethoxylates, which are used as surfactants in industrial detergents, and their degradation products, were shown to be weakly oestrogenic to fish in laboratory experiments (in vitro and in vivo), possessing around one ten-thousandth the potency of the natural oestrogen, 17B-oestradiol. Of the 20 chemicals selected on the basis of their likely occurrence in sewage effluent and examined for oestrogenic activity using a number of screening methods, nine were shown to be weakly oestrogenic. This further supports the hypothesis that a number of substances in sewage effluent could be acting as oestrogens in the male fish in this project.
   
10. The hypothesis that ethinyl estradiol (EE2), the synthetic oestrogen present in the contraceptive pill, is partly responsible for our observations was not resolved by this project, although women taking oral contraceptives excrete the conjugated form of EE2 which is not oestrogenically active. Some preliminary laboratory work (MAFF unpublished data) has shown that this inactive form of EE2 may be activated by enzymes present in sewage. Brief and unsuccessful attempts were made in this project to detect EE2 in sewage effluent using a sensitive radioimmunoassay, but the technique was found to be of limited success because of interferences present in the sewage effluent.
    
11. On the basis of the laboratory work and residue measurements in the field, it seems likely that the majority of the effect seen in the surveyed stretch of the River Aire can be explained by the presence of high levels of a degradation product of nonylphenol ethoxylates. Alkyl phenols and their parent ethoxylates may also be responsible for the effects seen in the other rivers surveyed because this group of chemicals is in wide use in industrial detergents and is known to be present in rivers. However, it is likely that a wide range of unrelated substances is contributing to the effects, and that different substances will dominate in different rivers according to local inputs. Concentrations of nonylphenol and alkylphenol ethoxylates vary (within and) between rivers but are generally much lower than levels seen on the Aire.
    
12. The report concludes with a series of recommendations for further research to answer more comprehensively the question of which substances are responsible for the effects, which habitats are the most affected, what the ecological implications may be.

Copies of the report are available from FWR, price £25.00, less 20% to FWR Members.