In a recent study, Christianson and several colleagues looked at whether they could also remove phosphorus by adding a special "P-filter" designed to trap the fertilizer-derived pollutant. The team tested two types of industrial waste products in the P-filters: acid mine drainage treatment residual (MDR) and steel slag. Phosphorus binds to elements such as iron, calcium, and aluminum contained in these products, removing it from the water. Rather than mixing MDR or steel slag with woodchips in one big nitrate- and phosphorus-removing machine, the team placed a separate P-filter upstream or downstream of a lab-scale bioreactor. They ran wastewater from an aquaculture tank through the system and measured the amount of nitrate and phosphorus at various points along the way. Nitrate removal was consistent, regardless of P-filter type and whether the P-filter was upstream or downstream of the bioreactor. But MDR was far superior as a phosphorus filter. "It removed 80 to 90 percent of the phosphorus at our medium flow rate," Christianson says. "That was really, really good. Amazing."Steel slag, on the other hand, only removed about 25 percent of the phosphorus. "But steel slag is a lot easier to find in the Midwest. And according to the Illinois Nutrient Loss Reduction Strategy, we're only trying to remove 45 percent of the phosphorus we send downstream. Since agriculture is only responsible for half of that, 25 percent would be pretty good," Christianson says.The system clearly shows potential, but several unknowns remain. Paired bioreactors and P-filters have yet to be tested in real-world conditions, although a handful have been installed in the United States. Perhaps more importantly, researchers don't have a good handle on how much phosphorus is running off agricultural fields in tile drainage.