Approximately 1500 km3 of wastewater is produced globally every day, according to the UNESCO World Water Development Report. Estimates suggest that around two thirds of the world’s freshwater is used to dilute wastewater discharge, leaving behind a significantly diminished amount of unpolluted freshwater to meet global demands. The need to better understand the impact of wastewater effluents on this vital resource and the ecosystem it supports, has led to numerous studies on freshwater fish, particularly looking at the feminising effects of the wastewater pollutant oestrogen on male fish. Knowledge of the individual effects of oestrogen however must be accompanied with a wider view of the effects on fish populations. Patrick Hamilton and Charles Tyler from the University of Exeter, UK, set out to examine this in wild populations of roach in UK rivers, as published in a recent BMC Biology study and discussed in an associated Commentary. Here Hamilton and Tyler explain how these cyprinid fish may be more resilient when viewed as a population than as individuals.
How widespread are oestrogenic pollutants, and what do we know about their effects on fish?
Nearly all treated wastewater treatment work effluents examined to date are oestrogenic. The oestrogenicity results from the widespread use of ethinylestradiol (EE2), a synthetic hormone that is a component of the female contraceptive pill, and also natural oestrogens from human excretion that are not fully removed during most sewage treatment processes. There are also a range of chemicals, including pesticides, pharmaceuticals, industrial detergents and plasticisers that act as oestrogen mimics and can contribute to this oestrogenic activity. It is therefore not surprising that the impacts on fish have been reported from across the developed world and particularly in densely populated areas with low rainfall, where treated effluents can make up a large proportion of the flow of rivers.
In fish, the effects of exposures of oestrogenic pollution are diverse and they particularly include effects on sexual development, reproduction and behaviour. Male fish exposed to oestrogen produce vitellogenin, a protein that is normally only produced by females for deposition into developing eggs. Prolonged oestrogen exposure during sexual development also feminises the testes in males. In some effluent-contaminated rivers in England all male roach were found to possess feminised testes. In extreme cases, almost half of the testis can be made up of female tissue – developing eggs. These fish have reduced sperm quality and an impaired reproductive performance in competitive breeding scenarios. Thus, at an individual level, the adverse health effects of exposure to oestrogen are reasonably well understood. In contrast effects of oestrogenic chemicals at the population level are less well understood. Findings that have raised the greatest concern are where long-term exposure has resulted in reproductive failure, the best known example being a Canadian study, where after three years of exposure of an entire lake to EE2, the population of resident fathead minnows underwent collapse. It recovered subsequently, after removal of the oestrogen dosing.
Your study involved wild rather than caged populations; what are the benefits and drawbacks of this approach?
Cage fish studies are useful for assessing responses of fish to oestrogens in contaminated water, but they do not necessarily reflect exposures of wild fish, which can be life-long. Equally, wild fish are often highly mobile within a water system and do not necessarily reside in one part of a water system only. Some of the feminised phenotypes seen in roach take several years to develop at environmental concentrations and most roach only start breeding at three years of age, so assessing these impacts through experimental exposures is time consuming and expensive. Studies on wild fish reflect effects in fish that are derived from real world, life-long exposures.
However, there are drawbacks in using wild fish, including incomplete knowledge on the exposure history of the population. We attempted to overcome this in our study by choosing sites where migration is likely to be restricted by obstructions such as weirs and locks. However, until we examined the population structure, we did not know the extent to which these man-made barriers restricted movement in these river systems. Another drawback in studying wild populations is that very few lowland rivers stretches will have no effluent inputs so we instead compared wild roach populations experiencing the upper and lower ranges of exposure regimens. In essence in the wild there are no clean reference sites.
What did you measure, and what did it show?
We used the wild fish roach for this study because feminisation of their males has been well documented in the UK in wild populations, and the species frequently occurs in rivers with a high proportion of effluent. Many of the rivers in southern England that we used have a high proportion of effluent and obstructions such as weirs and locks that are likely to confine fish populations to clean or polluted stretches of river. Our analysis of population structure using DNA microsatellites demonstrated that there are some roach populations in river stretches with a high proportion of effluent exposure that are sustained without substantial immigration from less polluted sites. We also calculated effective population sizes for each roach population from the microsatellite genotypes. [Natural populations often have less genetic diversity than one might expect for their size; to account for this, effective population size refers to the size of a theoretical population that has similar genetic characteristics to the actual population in question – in this case the wild roach].
Effective population size is affected by biased-sex ratios and by skews in reproductive success and these effects have been documented in the laboratory in exposures to oestrogens. Critically this parameter affects the rate at which genetic diversity is lost from a population, thereby influencing sustainability. We found no evidence for a relationship between oestrogen exposure, which was predicted using hydrographical modelling, and effective population sizes. However we could not exclude an influence of oestrogen exposure on effective population sizes.
We’ve heard wild populations are hard to measure accurately. How precise were your estimates, and could there be an effect of pollutants that didn’t show up in this study?
Assessing population sizes of any fish is difficult, as numbers fluctuate over time, and they are not evenly dispersed in any environment. In addition, in fish with high fecundity, a few fish can contribute disproportionally to the next generation, so fish numbers can give a misleading picture of sustainability. In this study, we instead focused on calculating effective population sizes using DNA microsatellite variation. Generally the estimates of effective population size were precise when they were small, as was the case for many of the populations we examined, but were less precise for large estimates. It is possible that we could have missed up to a 65 percent reduction in effective population size at the most impacted sites due to the wide confidence limits in the statistical model. However, river stretches with this level of contamination make up only about three percent of lowland rivers in England.
In light of your results, do you think the dangers associated with oestrogen exposure are not as severe as first thought?
Our finding that populations living in some rivers with a high proportion of effluent are self-sustaining is, in fact, consistent with some of the published work. In particular, life cycle exposures to fathead minnows at concentrations similar to those predicted at impacted sites in this study (0.3-1ng/L EE2), have found impacts on fertilisation success, but not reproductive failure. The studies which have found reproductive failure have used higher concentrations of oestrogen. Our results are also consistent with our previous study on roach where we found the majority of both feminised male and female roach from two impacted rivers were able to produce offspring. Oestrogen exposure is therefore likely to impact populations, but collectively, as yet, there is no substantial evidence that oestrogen exposure at environmental levels will lead to population collapse.
Having said this, we have previously found that feminised gonads are one of the most important determinants of male reproductive performance of roach at an impacted river. Furthermore, in this study we could not exclude a 65 percent reduction in effective population sizes at the most impacted sites. There may therefore be impacts on roach populations that we did not detect or measure.
What are the implications of your study for environmental policy?
There is little doubt that oestrogens and concentrations found in the environment, impact on individual fish in UK rivers, more widely in Europe and indeed globally, but most regulation requires demonstration of adverse impacts on fish populations, rather than at an individuals. Obtaining such results on populations is extremely difficult (and there will always be some degree of uncertainty) but this information will be of most interest to regulators compared with effects seen on individuals. Our results provide a better understanding of the influence of oestrogenic effluents on roach and although we found evidence for self-sustaining populations, it is important to note that our study could not exclude a negative impact on effective population sizes at the most impacted sites.
Roach are not the most sensitive fish to chemical pollutants and even if there is not an adverse population level effect in this species, it does not mean that this will be the case for all fish species. Another important consideration is that these roach populations are likely to have been exposed for many generations, so may well have adapted to some of the adverse effects of oestrogen exposure.
The information we have presented will help decision making on the risk posed by steroid oestrogens in the environment and will be of direct benefit to the water industry, environmental, policy and regulatory agencies both in Europe and worldwide. Despite the uncertainties of the effects of oestrogens at the population level, and in recognition for the need to control the concentrations of steroid oestrogens in surface waters, EE2 and E2 are now under consideration for regulation within legislation of the EU Water Framework Directive.