N-compound to be examined. In case of spiking with nitrite, thesmall circulating systems were operated for 3 h prior to spiking toensure that background levels of urea and ammonia were oxidizedto nitrite and nitrate concomitantly. Total volumes of the stock solu-tions used for the separate spiking with ammonia, nitrite and urea were 50, 50 and 25 ml, respectively, yielding bulk concentrations of approximately 5.0, 5.0 and 2.5 mg/l, respectively in the circulating biofilter systems. The stock solutions of ammonia, nitrite and ureawere added in steps of 10, 10 and 5 ml, respectively, every 2 min tothe reservoirs in order to distribute the compound evenly withinthe system.The first water sample was taken from the reservoir 2 min afterthe last addition of stock solution using a 25 ml syringe. Subsequent water samples were obtained at 15 min intervals after spiking with ammonia and nitrite, and at 10 min intervals after spiking with urea. Sampling in all cases continued for approximately 6 h(350 min). Nitrate-N concentrations were measured prior to clos-ing the loop and after the final sampling in the ammonia and nitrite spiking trials, while it was measured every 20 min in the urea spik-ing trials. Water samples were sterile filtered (Filtropur S 0.2 m,Sarstedt, Germany) and stored in 15 ml tubes (Sarstedt, Germany)at 3◦C in the dark until further analyses in the laboratory.Spiking and sampling was repeated once for each of the threeN-compounds with three days interval. The nitrite spiking trial was initiated 3 days after the ammonium trial, while the ureatrial started one month after the nitrite trial. Biofilters were oper-ated in end-of-pipe configuration modes in between measuring periods