From the wellspring of disaster rise faint hopes

Efforts to provide fresh water to West Bengal and Bangladesh led to an environmental and human catastrophe. But, reports Steve Farrar, scientists are closing in on the cause of the disaster.

A widow stares at the ugly lesions scarring her hands. The same silent killer that devastated her family is now visibly gnawing at her body, too. Her future is bleak, and she is not alone. The poison that is ruining her life lurks in water drunk by millions in Bangladesh and West Bengal. Naturally occurring arsenic has been found in tube wells drilled in the past two decades across a region already ravaged by poverty, political instability and seemingly relentless natural disasters. A humanitarian effort to provide the population with clean water has produced the largest mass poisoning in history.

An understanding of the exact chemical processes that lie behind the arsenic's presence - elusive for many years - is crucial for any long-term solution. Now, with several pioneering studies producing significant results, it seems a definitive answer should be confirmed next year.

The first signs of the epidemic emerged in the 1980s. It gradually became clear that the problems were linked to a well-building programme, started by the United Nations Children's Fund and other international agencies. Some 4 million wells have been to provide an alternative supply of water to pools that were often rife with disease-causing microbes. Two water quality surveys by the British Geological Survey gave the project a clean bill of health after assessing the region's geology. When their results were published, arsenic had simply been overlooked.

Arsenic is a common substance. Typically, you will find tiny quantities of it everywhere, even in food. The quantities consumed every day in the United Kingdom pose little threat and are comfortably below the World Health Organisation's guideline exposure limit of 10 micrograms per litre of water.

A BGS assessment in 1998 found that a third of Bangladesh's 2,022 wells tested exceeded this limit by five times, almost one in ten was above 300 micrograms per litre, and a few reached 1,600 micrograms per litre. Overall, it was calculated that at least 42 million Bangladeshis drink water that contains levels of arsenic above WHO safety recommendations.

The scale of their suffering is only beginning to become clear in what one researcher has described as a "silent epidemic". The most obvious symptoms of arsenicosis are skin blemishes and painful keratoses of thickened and cracked flesh. Arsenicosis can also affect the nervous and circulatory systems and can cause gangrene and organ damage. Scientists are growing increasingly nervous that what has been uncovered so far is a mere prelude to further horrors. There are suggestions that there could be an impact on reproductive health, possibly leading to miscarriages and malformed babies.

Richard Wilson, a nuclear physicist turned epidemiologist at Harvard University, observes that it can take skin cancer ten years to appear after exposure, while lung and bladder cancer can take double that time. "I expect that within 50 years, we will get nearly a million cancers from past arsenic exposures and as many keratoses," he says.

Allan Smith, professor of epidemiology at the University of California at Berkeley, warned the WHO that one in ten of those drinking water contaminated with 500 or more micrograms of arsenic per litre could ultimately die from cancer. "The scale of this environmental disaster is greater than any seen before; it is beyond the accidents at Bhopal, India, in 1984, and Chernobyl, Ukraine, in 1986," he said.

Measures to clean up the water, from filters to chemical treatments, have been proposed, but little has been achieved to alleviate the suffering on a scale to match that of the catastrophe. The 31 national and international organisations committed to tackling the problem have had a patchy start.

Kazi Ahmed, associate professor of geology at Dhaka University, complains:

"The mitigation action taken so far is not enough to address the enormity of the problem. Actions have not been prompt and efficient enough." Beyond attempts to mitigate the poisoning, knowing the precise cause would enable a permanent solution to be found.

The Ganges, Brahmaputra and Meghna rivers make Bangladesh a well-watered, fertile country. Sweeping in from the Himalayas, they also seed the country with minute quantities of arsenic once locked in the rocks. Many of the silt particles in the rivers become coated with ubiquitous iron oxyhydroxide. This coating acts like a chemical sponge, its positive charge mopping up negatively charged arsenic and phosphorus ions. Nevertheless, the arsenic levels found in the Bengal Basin are hardly exceptional. The question is what has caused so much of this arsenic to dissolve in the water.

It has been suggested that the source is iron pyrites, which release arsenic on contact with air drawn into the aquifers by human water extraction. Others have proposed that fertiliser phosphates are causing the iron oxide sponges to release their arsenic. According to David Kinniburgh, who leads the BGS's Bangladeshi research effort, the speculation is caused by insufficient field and laboratory data to pin down the cause.

Nevertheless, increasing numbers of scientists - including Ahmed and Kinniburgh - are coming round to another view, first put forward two years ago by a team of experts from University College London.

The team argued that the process was probably chemical reduction of the iron oxyhydroxides. Now John McArthur and Peter Ravenscroft, two of the original team, and Syed Saffiullah, a scientist at Jahangirnagor University near Dhaka, Bangladesh, believe they have found the final piece of the jigsaw. In a paper to be published in the journal Water Resources Research, McArthur and colleagues identify buried peat deposits and bacteria as the culprits.

Over the millennia, vast amounts of sediment have been deposited in the Bengal Basin. Throughout this time, swings in the climate have caused the region's vegetation to change periodically. The vast, lush swamps thrive and then die, ultimately creating deposits of peat, sandwiched between layers of sediment. Anaerobic microbes that live off organic material trapped in these subsurface deposits can create conditions that make iron oxyhydroxides unstable. The bacteria steal the oxygen they need to metabolise the peat from the iron oxides. When this happens, the oxide sponges dissolve and lose their ability to hold the arsenic. The toxin dissolves into the water.

McArthur believes no other source of organic matter could have so powerful an effect. Furthermore, his team have found a match between the distribution of known peat deposits and arsenic pollution hot spots. They also say that the variation of toxin levels at different depths could correspond to the peat layers. McArthur predicts: "You will find this problem in all delta environments in the tropics such as the Hanoi delta, the Mekong and China's Red River." The solution might be to drill wells in locations known to be peat-free.

But was it an accident waiting to happen? Perhaps not. Although research by Charles Harvey, assistant professor of civil and environmental engineering at the Massachusetts Institute of Technology, paints a similar picture of microbe-driven iron oxyhydroxide reduction, his preliminary results presented to the American Geophysical Union earlier this week suggest that man may have played a critical role.

Harvey proposes that the pumping of water from the wells may have drawn organic-rich surface water into the aquifers. Its unnatural presence would have an affect on the microbes very similar to that of McArthur's peat.

When Harvey was in Bangladesh earlier this year, he was able to measure chemical conditions at different depths of a purpose-bored, 165m well in a highly contaminated district of Bangladesh. He found arsenic throughout, but the quantity dissolved in the water peaked at 30m and then fell off below that depth.

It is a result noted previously by other teams. Harvey suggests that 30m is the typical depth to which irrigation wells are drilled. Harvey has estimated that if no action is taken, half a million cases of keratoses, 100,000 skin cancers and several thousand other cancers could occur in Bangladesh each year.

However, he claims that if he is correct then drilling deeper wells to bypass the arsenic peak created by organic-rich waters drawn down by irrigation could cut average lifetime exposure by two-thirds. Indeed, the BGS survey noted that just 1 per cent of wells extending more than 150m had levels of arsenic exceeding 50 micrograms per litre. Carbon-dating data, currently being analysed, could determine whether peat or surface water is the more significant player.

Meanwhile, both McArthur and Harvey are planning trips to Bangladesh in the spring to gather the chemical evidence that could nail the mystery once and for all and help bring the poisoning to an end.