By Richard Lovett

Pharmaceuticals and household chemicals in rivers and streams may be affecting how fish mate and spawn, scientists warn, even when the substances are not present at levels high enough to cause visible damage.

The laboratory studies, reported at the 7-11 November meeting of the Society of Environmental Toxicology and Chemistry in Portland, Oregon, are part of a growing effort to determine how pollution affects fish behaviour, rather than just their anatomy or physiology.

Drugs and chemicals that are flushed down toilets and drains have long been reported in urban waste water and the streams into which it flows. But it has been difficult to work out how this is affecting wildlife.

"Subtle effects are the issue," says Melissa Schultz, a chemist at the College of Wooster in Ohio. "It's easy to tell if a fish suffers from obvious anatomical changes such as being intersex or not having mature secondary sexual characteristics," she says. But determining effects on mating behaviour "takes more meticulous work".

Schultz has been studying the effects of low concentrations of triclosan and triclocarban -- antibacterial agents commonly added to soaps, disinfectants and increasingly to household products such as toys, bedding, socks and rubbish bags. Her team placed adult fathead minnows (Pimephales promelas) -- fish commonly used in aquatic toxicity studies -- in aquaria containing 0.01-0.5 micrograms per litre (0.01-0.5 parts per billion) of various mixtures of these chemicals. "These are levels we have found in the environment," she says.

In the wild, fathead minnows establish nests, which males defend from rivals. But in Schultz's laboratory tests, exposed males were less interested than unexposed controls in trying to drive off artificial "bait" minnows placed near their nests.

The study shows that even at the low concentrations found in the wild, these chemicals can alter reproductive behaviour, says Schultz's collaborator, Heiko Schoenfuss, an aquatic toxicologist from St Cloud State University in Minnesota. "That's the take-home message."

Anti-inflammatory apathy

A second study by Dalma Martinovi, an environmental toxicologist from the University of St Thomas in St Paul, Minnesota, found equally disturbing effects in zebrafish (Danio rerio) housed in tanks containing 50 micrograms per litre (50 parts per billion) of ibuprofen.

The drug has been found in environmental samples at levels up to 20 micrograms per litre. That's a little lower than the levels used in her study, admits Martinovi, "but the same order of magnitude."

Ibuprofen and related drugs work by inhibiting two enzymes called COX-1 and COX-2. In humans, these enzymes are part of the pathway that produces prostaglandins, which have a role in inflammation. But in fish, says Martinovi, prostaglandins also serve as pheromones, and play a part in initiating sexual behaviour and spawning.

Martinovi found a significant reduction in courtship behaviour, such as "chasing and touching noses", in males exposed to ibuprofen, she says. And water collected from the aquaria of ibuprofen-exposed females elicited less of a reaction in male fish than water from the tanks of females that had not been exposed to the drug.

She concludes that even though there were no observed anatomical or physiological changes in the fish, they were nonetheless affected by exposure to levels of ibuprofen similar to those in municipal waste water. "Physiological endpoints are less sensitive than behavioural ones," she says. "I think effects on behaviour might be more common than we realize."

"We are finding subtle effects at environmentally relevant concentrations," adds Schultz. Both researchers point out that the affected behaviours could impact the ability of fish species to thrive in the wild.

James Lazorchak, an aquatic biologist from the US Environmental Protection Agency in Cincinnati, Ohio, who was not involved with either study, agrees. "We need to be paying more attention to behaviour," he says, "especially with pharmaceuticals".