The outbreak of food poisoning that killed 18 in Scotland was contained locally, but the war against nasties such as E coli must be fought everywhere. Hugh Aldersley-Williams explains why
Later this month the full report into the outbreak in Lanarkshire, Scotland, of infections caused by the bacterium E coli 0157 will be published. But already, two apparently contradictory facts are clear from the incident, which has claimed 18 lives since November. One is that the bacterium is newly observed, and was unknown before 1982. The other is that although the outbreak was local, precautions against future outbreaks must now be taken globally.

In his interim report, announced on 15 January, Professor Hugh Pennington, the head of the team investigating the food poisoning outbreak, recommended that central government be notified of new outbreaks even if they are contained within a single local authority area.

The puzzle is, how can a species unknown until 15 years ago already pose such a wide risk?

In most cases when a new species is discovered, we would expect to see it in a confined habitat. But small organisms - including bacteria and even viruses - seem not to follow the rules of evolution and biodiversity used by large plants and animals.

Many such micro-organisms are waterborne. Dr Bland Finlay, at the Institute of Freshwater Ecology, based on the shores of Windermere, has studied a particular type of these creatures, known as ciliated protozoa. His field data have led him to challenge scientists who suggest that total numbers of microbe species can be estimated in the same way as for insects and larger species.

The debate is significant because the calculations performed by each side lead to very different estimates of global microbe biodiversity.

The importance of biodiversity in sustaining life on earth is gaining greater public understanding. But aquatic habitats remain a largely unknown quantity even to scientists. They are thought to support the great mass of living organisms; do they also host the greatest diversity?

Extrapolation from conventional population models used for large species would say they do. These models relate the richness of species to the numbers of all species present in a given habitat. The richness varies as the square root of this abundance, and the abundance in turn is directly proportional to the area of the habitat as well as being related to climate. In other words, in locations at about the same geographical latitude, one would expect a habitat four times larger than another to support a population four times greater, with twice the diversity of species. The model can be used to evaluate the prospects of species in isolated habitats or to calculate viable habitat sizes, and has proved valuable in conservation efforts for large species.

But Dr Finlay's results suggest that for many aquatic micro-organisms the diversity of a habitat no longer increases in this way as a function of its size and climate. "Estimates of the number of species globally run from 2 to 200 million," he explains. "The aquatic fraction is anyone's guess. Many organisms, such as bacteria, are aquatic, but we have no idea how to define them as species. In one pond, you can find 10 `species' never seen before for every one that is known."

But this does not mean that the total number of species of these organisms globally is necessarily very large. "Our model tells you that the total number of different types of microbial species is actually rather modest."

In the first experiment of its kind, Dr Finlay and co-workers set out to determine the number of ciliated protozoa in a sample of one square metre of sandy river bed sediment in a location near Seplveda in Spain. Ciliated protozoa are single-celled organisms covered in hair-like cilia that provide a means both of locomotion and of trapping food. They are common species, representing a significant proportion of the biological mass in these environments.

They found 65 species, eight of them never before observed, which between them exhibited a wide variety of sizes and shapes. (In organisms this small, where reproduction is often asexual, reproductive selectivity is no longer a useful way of defining species. Distinctive size, shape and structure become the criteria.)

The results meant that even smaller samples could be taken as representative. Recent studies of small samples from both freshwater and marine environments confirm Dr Finlay's thinking. The samples were treated in various ways, for example by varying the temperature or by adding particular nutrients, to reveal hitherto unrecognised species. Even in these small samples, it is possible to find up to 10 per cent of all ciliated protozoa ever recorded, confirming that the great majority of species must be globally distributed.

"Microbial diversity has unique features, because most microbes are probably cosmopolitan," says Dr Finlay. Compared to the typical pattern in larger species, the protozoa seem disproportionately diverse in small habitats and impoverished in large ones. "You can find all microbes anywhere, because their abundance is great and dispersion is easy. The implication is that any ecosystem, if studied properly, will reveal a great diversity of organisms, but a diversity that is also found in other regions."

Because of the cosmopolitan nature of protozoan species, the researchers believe this highly diverse community can be taken as representative of communities in other such habitats, wherever they may be. The data can be used to calculate the likely numbers of protozoan species in the entire biosphere.

Writing recently in the science journal Nature, he challenged researchers who have blithely assumed that what goes for large species also goes for micro-organisms. "Many species may be observed in a small natural sample, although the global number of species may be modest." For example, he estimates that there are only about 3,000 species of ciliated protozoa, although "the global number of individuals is unimaginably high".

The same goes for E coli. The danger that these and other bacteria pose is indeed global. It is perhaps small comfort to know that there are probably fairly few species that we need to worry about.