The Sea Water Microbiology project
was a large-scale collaborative effort involving 34 laboratories from
12 Member States. Their goal was to evaluate the wide variety of techniques
currently used to test the quality of bathing water and improve the
accuracy, reliability and comparability of the data available to regulatory
authorities and to the public.
The project's conclusions, unanimously supported by all its participants,
provide a rigorous scientific basis for the standardisation of testing
methods. Its recommendations are likely to be adopted in full in the
next European Directive on water quality, and will provide the framework
for the progressive improvement of bathing water throughout the Union.
Every summer, millions of Europeans and
large numbers of visitors from overseas swim in coastal waters off
the beaches of the Aegean, the Adriatic, the Mediterranean, the
Baltic and North Seas, and the Atlantic Ocean.
It has long been accepted that the pollution of sea water by human
and animal sewage effluents constitutes a serious threat to human
health, both directly to swimmers and indirectly to those eating
contaminated shellfish. As long ago as 1976, the European Community
issued a Directive which requires all Member States to carry out
regular monitoring of bathing water quality, and sets out both desirable
and minimum standards. Since 1991, the Commission has published
an annual report which brings this data together and maps the quality
of water at beaches across Europe.
The map receives widespread publicity, and is known to be a key
factor in many holiday-makers' choice of destination. A poor showing
on the water quality map guarantees that an area will suffer a significant
loss of tourist income.
Unfortunately, the 1976 Directive is flawed. It permits, and even
encourages, the use of difficult testing methods, and it does nothing
to ensure the comparability of the analytical results submitted
by different countries or regions. In fact, according to senior
staff at the Institut Pasteur de Lille, which coordinated the Sea
Water Microbiology project, until very recently the published maps
were scientifically worthless.
The project itself, which involved 34 laboratories from 12 EU Member
States, was financed by the EC's Standards, Measurements and Testing
Programme, and represents an important first step towards improving
the accuracy, reliability and comparability of the data available
both to regulatory authorities and to members of the public.
Clouding the waters
The shortcomings of the Directive on the testing of bathing water
have long been recognised by the scientific community. Test results
from a single stretch of beach spanning the French-Belgian border,
for example, had consistently shown water quality on one side to
be better than that on the other. It was clear to both sides that
the discrepancy arose solely from differences in the measuring techniques
Results from different Member States, regions or localities were
simply not comparable, because they had been produced using different
methods. It was perfectly possible for water which failed to meet
European standards when measured by one widely used method to achieve
compliance when measured by another.
Universal demand for a standard measurement technique is not readily
translated into universal enthusiasm for any particular method,
however. Predictably, each laboratory involved in the routine monitoring
of sea water wanted its own technique to be adopted by all the others.
Testing test methods
A collaborative, pan-European study was therefore needed to evaluate
the range of current testing methods. Each participating laboratory
would test the same samples, using both its own techniques and common
methods agreed with other participants.
Sea water, unfortunately, is highly unstable, so distributing centrally
prepared samples to the participating laboratories would have made
it impossible to guarantee uniform microbiological composition.
It was therefore agreed to bring all 34 participants to work together
in the same laboratory, testing the various techniques on the same
samples on a single bench. Two groups of participants travelled
to Lille for a series of week-long trials, where the Institut Pasteur
was able to provide appropriate materials and apparatus, as well
as transport and accommodation for the visiting technicians.
Despite initial language difficulties, both test groups rapidly
gelled into friendly and constructive teams. Expertise was freely
and enthusiastically shared, and clear results were produced, supported
by all participants.
The study generated three main sets of conclusions. First, it found
that experienced microbiologists obtain equivalent results when
using a common test method, and significantly different results
when using their own methods.
Some test methods in regular use - the three and five-tube Most
Probable Number (MPN) tests - were imprecise and unreliable and
should be abandoned altogether. They should be replaced by more
recently developed Membrane Filtration (MF) or MPN 96-well techniques,
which were found to give much better performance.
There were a number of conclusions regarding which measurement parameters
to use. E. coli, for example - a dangerous faecal bacterium found
in sewage and other wastes - was confirmed to be the only well-defined
and easily measurable component of its group (the faecal coliform).
Measurement of the larger but poorly defined faecal coliform and
thermo-resistant coliform groups, therefore, should be discontinued.
Third, a reference method should be developed to serve as a benchmark
for the evaluation of existing and newly developed test methods.
For use in the routine monitoring of sea water quality, test methods
must be relatively quick and easy, as well as accurate and reliable.
A reference method, which need be neither quick nor easy, but whose
accuracy is known, is essential if the equivalence of different
routine methods is to be determined.
Development of this reference method fell outside the scope of the
project itself, but the report indicates the participants' firm
conviction that it should consist of two complementary techniques
applied in parallel, in order to ensure complete accuracy.
A framework for improved performance
Some of these conclusions were published as early as 1992, following
the project's first trial. Since then, very largely as a result
of these results, there has been a steady improvement in the homogeneity
of testing techniques, and therefore in the reliability of the published
water quality maps.
Different techniques are still in use, but the project has at least
provided a basis for assessing the comparability of results produced
by different methods, and its recommendations have been progressively
adopted by individual laboratories or countries. France, for example,
has now specified a single measurement technique for all its bathing
The full set of recommendations, which are supported by all the
participating laboratories, are not national but European. Indeed,
they accord closely with the latest thinking of the International
Standards Organisation (ISO), which has made extensive use of the
project's findings, and has already adopted E. coli as the target
measurement paramater for coliforms.
The Institut Pasteur is confident that the next European Directive
on bathing water quality will also adopt the recommendations. When
this will happen is, however, less certain. Individual Member States
are, naturally, as protective of their own testing methods as individual
The actual quality of the European bathing water depends not on
the methods used to test it, but on the discharge of sewage and
an effective, comprehensive management policy which includes a waste
water collection network, storm run-off management, regulations
not to build too close to the coastline and avoiding waterproofing
nearby surfaces with concrete and tarmac. The project's achievement
has been to provide a framework for the accurate, reliable and comparable
measurement which is the prerequisite for effective regulation of