Researchers' preliminary findings show a higher volume than expected of the particles, known as microplastics, even in the cleanest and best-flushed reaches.

A team led by the director of the Sydney Institute of Marine Science Emma Johnston wants to do further research to identify the origin of the microplastics and measure toxic effects on ocean organisms and animals.

Microplastics are less than five millimetres in length. They enter waterways as clothes fibres, facial scrubs and bags or larger items that later break down.

Overseas studies have shown these tiny particles have the capacity to absorb persistent organic pollutants but there is little known of the effects once they are consumed by sea life.

These debris are now everywhere in the aquatic environment, contaminating coastal areas and the open ocean.

Global trends suggest accumulations are increasing in aquatic habitats at a rate consistent with trends in plastic production, increasing 560 fold in more than 60 years.

In November, the prestigious Nature magazine published US research that showed fish were suffering from liver damage after bioaccumulating chemicals attached to microplastics they had swallowed.

"Plastics enter and persist in environments from the poles to the equator and down to the depths of the sea," said Dr Johnston, a professor of marine ecology and ecotoxicology at the University of NSW.

"Slow degradation of plastic debris into ever-smaller particles means that microplastics are accumulating in the environment. Laboratory trials indicate this material is likely to be present in animal tissues and food webs.

"Although larger pieces of plastic can be removed by sewage treatment plants and stormwater filters, no existing filtration methods retain microplastics so they continue to be released.

"Risk analysis cannot be used for microplastics because of a lack of fundamental information about levels of contamination in habitats and the uptake or consequences of this material in natural systems."

One of the researchers Vivian Sim, a University of NSW PhD candidate, said from 27 sites searched in Parramatta River, Lane Cove and Port Jackson, the highest densities of fragments were found in Middle Harbour.

She said the samples were dominated by microplastic threads. Flakes and balls were observed but were too fragile for extraction in most cases, but their abundance was comparatively lower than that of threads.

"We were expecting to find more fragments in samples with fine silt sediments but actually got the opposite," she said. "Middle Harbour generally has a coarse sand grain size but yet we found the highest densities there.

"This suggests that something interesting is going on here, but we're not sure what."