Using bulk deposit samplers to evaluate pollutant loads from atmospheric deposition

The aim of this study was to evaluate the number of bulk deposit samplers needed to measure atmospheric deposition of pollutants with an acceptable level of variability. A set of 10 bulk deposit samplers was placed at an experimental site near Lake Biwa in the Shiga Prefecture in Japan, and water samples were collected monthly from each sampler. First, the effects of the intrusion of insects and litter were examined by placing intrusion-prevention devices on the samplers. All measures of variation (coefficient of variation [CV], skewness, and kurtosis) were smaller when the device was used than when it was not. Total nitrogen (TN) loads started to increase in April and peaked in June, with a secondary small peak in February. Levels of dissolved nitrogen, NO3-N, and NH4-N showed similar trends. Total phosphorus (TP) reached a maximum in May, and stayed at a relatively high level until it reached a second peak in August and September. The annual precipitation data obtained from the bulk deposit samplers were normally distributed and had an average and median value of 1570 mm yr-1 (CV, 1.0%). Annual TN loads were also distributed normally, and their average and median values were 12.1 (CV, 2.7%) and 12.0 kg ha-1 yr-1, respectively. Annual TP load, however, apparently deviated from a normal distribution, and the average (0.417 kg ha-1 yr-1; CV, 24%) was larger than median (0.395 kg ha-1 yr-1). At least 8 bulk deposit samplers were needed to obtain TP loads of atmospheric deposition with a CV of less than 5%. The TN loads from the atmosphere to Lake Biwa were estimated to range from 780 to 854 t y-1, whereas the TP load showed a much larger relative range of 20.8 to 42.9 t y-1, indicating that more studies are needed to measure and evaluate the atmospheric deposition of phosphorus.