Potential Methyl Mercury Contamination in the Three Gorges Reservoir

by Alan Penn, M.Sc.

Background on Methyl Mercury Contamination

The
CYJV study indicates that inorganic mercury is present in the sediment,
soil, and vegetation that would be subject to flooding by the Three
Gorges Project. Canadian experience with the impoundment of rivers to
create reservoirs for hydroelectric dams demonstrates that methyl
mercury, a central nervous system toxin, is formed through bacterial
synthesis in flooded soils and vegetation. The methyl mercury so
produced is accumulated by fish, and is thus a potential health hazard
for consumers of fish.

The process of methyl mercury production occurs
naturally in lakes and rivers but certain new reservoirs in northern
Manitoba and northwestern Quebec in Canada have resulted in a four to
six-fold increase in methyl mercury concentrations in fish. In
particular, methyl mercury contamination has become a significant issue
at the La Grande hydroelectric complex (Phase I of the James Bay
Project) which was built by Hydro-Québec in northwestern Quebec between
1970 and 1984, and will continue to play a role in the Quebec
government’s plans to proceed with additional hydroelectric development
in the same region.

Prior to hydroelectric development in these
regions of Canada, methyl mercury contamination was a concern because
of regionally elevated levels of mercury found in fish. In the 1970s,
methyl mercury concentrations as high as 100 milligrams per kilogram
(mg/kg) were found in the hair of some Cree Indian fishermen. The World
Health Organization sets the tolerance limit for human exposure to
methyl mercury at 6 mg/kg; therefore, the high levels found in the
Crees have prompted questions about long-term toxicity and the possible
effects of fetal exposure.

Since the construction of Phase I of the James
Bay Project, limits have been placed on fish consumption in order to
control exposure to methyl mercury. Recent surveys of adult Crees over
40 years old reveal that approximately 5 percent of this adult
population have in excess of 25 mg/kg in their hair.

In the case of the Three Gorges Project, CYJV has
identified methyl mercury contamination as a potentially major water
quality impact and proposes a monitoring program for soil, vegetation,
and fish – including more scientific studies, post-impoundment
monitoring of fish, and restricting consumption should mercury levels
in fish become a health risk.

CYJV suggests several conditions which would act
to reduce the extent of methyl mercury contamination: high annual flows
through the reservoir; low organic matter in soils and absence of peat
in the reservoir region; removal of vegetation before flooding; and the
water’s resistance to acidification. Factors which would act to
increase the likelihood of contamination are cited as dissolved oxygen
depletion in the reservoir, mercury present in soils and vegetation,
and the transport of mercury with suspended sediment.

Key Criticisms of CYJV’s Discussion of Methyl Mercury

• In the past two decades, methyl mercury
  contamination associated with North American water resource development
  projects has stimulated research on the factors which influence the
  rate of methyl mercury production and breakdown over flooded terrain,
  and its subsequent accumulation by fish. CYJV does little to assess the
  extent to which Canadian or United States research on methyl mercury in
  the aquatic environment could be used in the evaluation of the Three
  Gorges Project. Without further documentation on the nature of
  available research, the CYJV text on this subject must be considered
  inadequate, and potentially misleading.
• In any event, applying experience from
  water resource development in Manitoba and Quebec to the very different
  geological setting of the Three Gorges Project has its own limitations.
  Even in the absence of flooding to create a reservoir, current
  understanding of geochemical factors which influence the
  bioavailability of inorganic mercury and production of methyl mercury
  is quite limited. These technical limitations which make impact
  assessment in this area rather uncertain, were not brought out by CYJV
  in its study.
• The rate of accumulation of methyl
  mercury by fish ultimately depends on the feeding ecology (i.e.,
  feeding patterns and habitat) of the fish species present and the
  nature of the food chain. There is little in the CYJV study to indicate
  how these biological factors would influence the nature of anticipated
  methyl mercury contamination in the case of the Three Gorges Project.
• CYJV did not conduct a risk assessment for
  plausible scenarios of methyl mercury contamination in consumers of
  fish from the Three Gorges reservoir. The current health status of the
  population and the role of fish in the diet (actual or potential) are
  not addressed. Also, the role of fish as a subsistence food within the
  local economy is unclear. In these circumstances, it is especially
  difficult to assess the extent to which methyl mercury contamination
  should be considered a significant public health issue. There appears
  to be no fundamental reason that such a risk assessment could not have
  been undertaken.
• CYJV’s treatment of methyl mercury as an
  environmental issue in the feasibility study, not unexpectedly,
  reflects Hydro-Québec’s approach to, and experience with, Phase I of
  the James Bay Project. Hydro-Québec’s guiding principle* seems to be
  that ecological impacts, by their very nature, are difficult to
  predict. Therefore, the argument runs, rather than rely on prior
  assessments, ecological and social consequences should be monitored as
  they unfold and responded to by the proponents, where necessary, with
  appropriate remedial measures. This leaves a great deal of discretion
  and control in the hands of proponents and places local people, who are
  directly affected by ecological impacts, in a disadvantaged position.
• In this setting, this also means that it
  is difficult to deal directly with ecological concerns, such as methyl
  mercury contamination, in the initial project design. A more
  fundamental concern is that the overall approach to hydroelectric
  development reflects an underlying lack of conviction about the
  relevance of initial environmental impact assessments and about the
  value of acquiring practical experience through those assessments
  before decisions are made.

Sources and Further Commentary

*Hydro
Quebec’s approach is set out in Section 8 of the James Bay and Northern
Quebec Agreement, the aboriginal land claims settlement associated with
the James Bay Project.

Continue to Chapter 6

Back to
Chapter 4