A specially designed field chamber has proved to be a good tool when used together with a computer model to evaluate how effectively riparian buffers filter out pollutants before they can reach streams or other bodies of water.

That is the finding from a series of systematic studies of a riparian grass buffer zone by scientists at the Agricultural Research Service's (ARS) Henry A. Wallace Beltsville (Md.) Agricultural Research Center (BARC).

After success with a prototype chamber in the laboratory, BARC soil scientists Jim Starr, Ali Sadeghi and Yakov Pachepsky installed a field version of the chamber in a tall fescue grass buffer near a forested stream and wetland area. The chamber has no top or bottom and encases the four sides of a 3-by-3-foot block of soil down to 4 feet.

The scientists injected water with dissolved nitrate into one side of the soil chamber. Then, as water flowed horizontally and out the other side, they monitored rates of lateral water flow and loss of nitrate due to its breakdown by soil microbes.

Riparian buffers are wooded or grassy streambanks or wetlands that filter out pollutants such as nitrate from fertilizer and other chemicals, as well as sediment. The slow movement of nitrate through highly organic riparian areas provides ideal conditions for soil microbes to break down or transform the nitrate into safer compounds.

The scientists used the two-dimensional computer model "HYDRUS-2D" to simulate water flow and transport of chemicals within the riparian zone soil. Overall, the model-chamber combination provided good results.

The experimental chamber is essential for the accurate use of the growing number of computer models being developed to assess the effectiveness of riparian buffers. Once a model gets this information for a particular location, it can predict nitrate loss rates.

A paper on this study will appear in the November-December issue of the Soil Science Society of America Journal.

ARS is the U.S. Department of Agriculture's chief in-house scientific research agency.